1998-03-25 07:18:34 +08:00
|
|
|
|
/*
|
|
|
|
|
* Copyright (C) 1998 NCSA
|
|
|
|
|
* All rights reserved.
|
|
|
|
|
*
|
|
|
|
|
* Programmer: Robb Matzke <matzke@llnl.gov>
|
|
|
|
|
* Monday, March 23, 1998
|
|
|
|
|
*/
|
|
|
|
|
|
[svn-r876] Changes since 19981102
----------------------
./bin/snapshot
Made same fix as for the release script yesterday.
./src/H5D.c
./src/H5Dprivate.h
./src/H5G.c
./src/H5Gprivate.h
./src/H5Gpublic.h
./src/H5O.c
./src/H5Oprivate.h
./src/H5RA.c
./src/H5RAprivate.h
./src/H5T.c
./src/H5Tprivate.h
Improved object type checking. Instead of determining the
object type by trying to open each of the possible types, we
keep a table of associations between object type number (like
H5G_GROUP, H5G_DATASET, H5D_TYPE, and H5D_RAGGED) and an `isa'
function that returns true if the object header has the right
messages to make the object a particular type. This mechanism
also allows specialization of object types by permitting an
object to satisfy more than one `isa' function.
Added `isa' functions for groups, datasets, ragged arrays, and
committed data types.
./src/H5config.h.in
Added HAVE_STAT_ST_BLOCKS. I thought this had already been
added, but apparently not.
./tools/h5ls.c
Removed system include files since they're already included by
H5private.h and since I wasn't including them portably anyway.
By default, 1-byte integer types are printed as integer values
instead of ASCII characters. However, the `-s' or `--string'
command-line switch causes the data to be interpretted as
ASCII. String data types are always printed as character
data.
Ragged arrays are now identified as ragged arrays and h5ls
doesn't descend into the group automatically. This uses the
new object type specialization stuff.
./tools/h5tools.c
./tools/h5tools.h
Added the ability to print 1-byte integer types as either
ASCII or numeric data instead of always ASCII. The default is
to print as numeric data.
1998-11-06 04:28:34 +08:00
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* We include the private header file so we can get to the uniform
|
1998-11-20 02:52:56 +08:00
|
|
|
|
* programming environment it declares. Other than that, h5ls only calls
|
1999-01-22 02:33:39 +08:00
|
|
|
|
* HDF5 API functions (except for H5G_basename())
|
[svn-r876] Changes since 19981102
----------------------
./bin/snapshot
Made same fix as for the release script yesterday.
./src/H5D.c
./src/H5Dprivate.h
./src/H5G.c
./src/H5Gprivate.h
./src/H5Gpublic.h
./src/H5O.c
./src/H5Oprivate.h
./src/H5RA.c
./src/H5RAprivate.h
./src/H5T.c
./src/H5Tprivate.h
Improved object type checking. Instead of determining the
object type by trying to open each of the possible types, we
keep a table of associations between object type number (like
H5G_GROUP, H5G_DATASET, H5D_TYPE, and H5D_RAGGED) and an `isa'
function that returns true if the object header has the right
messages to make the object a particular type. This mechanism
also allows specialization of object types by permitting an
object to satisfy more than one `isa' function.
Added `isa' functions for groups, datasets, ragged arrays, and
committed data types.
./src/H5config.h.in
Added HAVE_STAT_ST_BLOCKS. I thought this had already been
added, but apparently not.
./tools/h5ls.c
Removed system include files since they're already included by
H5private.h and since I wasn't including them portably anyway.
By default, 1-byte integer types are printed as integer values
instead of ASCII characters. However, the `-s' or `--string'
command-line switch causes the data to be interpretted as
ASCII. String data types are always printed as character
data.
Ragged arrays are now identified as ragged arrays and h5ls
doesn't descend into the group automatically. This uses the
new object type specialization stuff.
./tools/h5tools.c
./tools/h5tools.h
Added the ability to print 1-byte integer types as either
ASCII or numeric data instead of always ASCII. The default is
to print as numeric data.
1998-11-06 04:28:34 +08:00
|
|
|
|
*/
|
[svn-r339] Changes since 19980408
----------------------
./src/H5Osdspace.c
./html/H5.format.html
In the past we were allowed to have >2GB files on a 32-bit
machine as long as no dataset within the file was larger than
4GB (or whatever sizeof(size_t) is). That's been fixed now.
All dataset size calculations are done with `hsize_t' which is
normally defined as `unsigned long long'.
./src/H5F.c
./src/H5Ffamily.c
./src/H5Fprivate.h
./src/H5P.c
./src/H5Ppublic.h
The file family member size can now be set/queried. The
default is still 64MB, but it can be set to 1GB by saying:
H5Pset_family (plist, 30, H5P_DEFAULT);
When opening an existing file family the specified
bits-per-member is ignored and the first member of the family
determines the bits-per-member, which can be retrieved with
H5Pget_family().
./acconfig.h
./configure.in
./src/H5config.h
./src/H5public.h
Added `--disable-hsizet' so that those with old GCC compilers
(<2.8.1) can still compile the code.
./src/H5.c
./src/H5private.h
Added HDfprintf() which works just like fprintf() except you
can give `H' as a size modifier for the integer conversions
and supply an `hsize_t' or `hssize_t' argument without casting
it. For instance:
hsize_t npoints = H5Sget_npoints(space);
HDfprintf(stdout,"Dataset has %Hd (%#018Hx) points\n",
npoints, npoints);
You can now give `%a' as a format to print an address, but all
formating flags are ignored and it causes the return value of
HDfprintf() to not include the characters in the address (but
who uses the return value anyway :-). Example:
H5G_t *grp;
HDfprintf(stdout, "Group object header at %a\n",
&(grp->ent.header));
Added HDstrtoll() which works exactly like [HD]strtol() except
the result is an int64.
./src/debug.c
Large addresses can now be entered from the command-line. Use
either decimal, octal (leading `0') or hexadecimal (leading
`0x') when giving the address.
./src/h5ls.c
The printf format for dataset dimensions was changed to `%Hu'
to support large datasets.
./test/big.c [NEW]
A test for big datasets on 32-bit machines. This test is not
run by default. Don't try to run it on an nfs-mounted file
system or other file system that doesn't support holes because
it creates two 32GB datasets of all zero.
1998-04-10 04:22:11 +08:00
|
|
|
|
#include <H5private.h>
|
[svn-r876] Changes since 19981102
----------------------
./bin/snapshot
Made same fix as for the release script yesterday.
./src/H5D.c
./src/H5Dprivate.h
./src/H5G.c
./src/H5Gprivate.h
./src/H5Gpublic.h
./src/H5O.c
./src/H5Oprivate.h
./src/H5RA.c
./src/H5RAprivate.h
./src/H5T.c
./src/H5Tprivate.h
Improved object type checking. Instead of determining the
object type by trying to open each of the possible types, we
keep a table of associations between object type number (like
H5G_GROUP, H5G_DATASET, H5D_TYPE, and H5D_RAGGED) and an `isa'
function that returns true if the object header has the right
messages to make the object a particular type. This mechanism
also allows specialization of object types by permitting an
object to satisfy more than one `isa' function.
Added `isa' functions for groups, datasets, ragged arrays, and
committed data types.
./src/H5config.h.in
Added HAVE_STAT_ST_BLOCKS. I thought this had already been
added, but apparently not.
./tools/h5ls.c
Removed system include files since they're already included by
H5private.h and since I wasn't including them portably anyway.
By default, 1-byte integer types are printed as integer values
instead of ASCII characters. However, the `-s' or `--string'
command-line switch causes the data to be interpretted as
ASCII. String data types are always printed as character
data.
Ragged arrays are now identified as ragged arrays and h5ls
doesn't descend into the group automatically. This uses the
new object type specialization stuff.
./tools/h5tools.c
./tools/h5tools.h
Added the ability to print 1-byte integer types as either
ASCII or numeric data instead of always ASCII. The default is
to print as numeric data.
1998-11-06 04:28:34 +08:00
|
|
|
|
#include <h5tools.h>
|
|
|
|
|
|
1999-08-11 04:21:32 +08:00
|
|
|
|
/*
|
|
|
|
|
* File drivers
|
|
|
|
|
*/
|
2000-05-03 23:48:20 +08:00
|
|
|
|
#if defined VERSION13
|
1999-08-24 20:52:10 +08:00
|
|
|
|
#include <H5FDsec2.h>
|
|
|
|
|
#include <H5FDmulti.h>
|
1999-08-11 04:21:32 +08:00
|
|
|
|
#include <H5FDfamily.h>
|
2000-05-03 23:48:20 +08:00
|
|
|
|
#elif defined VERSION12
|
|
|
|
|
#include <H5Fpublic.h>
|
|
|
|
|
#endif
|
1999-08-11 04:21:32 +08:00
|
|
|
|
|
2000-06-07 02:07:20 +08:00
|
|
|
|
|
1999-06-14 23:07:58 +08:00
|
|
|
|
/*
|
|
|
|
|
* If defined then include the file name as part of the object name when
|
|
|
|
|
* printing full object names. Otherwise leave the file name off.
|
|
|
|
|
*/
|
|
|
|
|
#define H5LS_PREPEND_FILENAME
|
|
|
|
|
|
[svn-r537] Changes since 19980722
----------------------
./src/H5A.c
./src/H5Apublic.h
./test/tattr.c
Switched the order of the second and third argument of
H5Aget_name() to make it consistent with other functions that
take buffers and buffer sizes.
./src/H5G.c
./src/H5Gpublic.h
./src/H5Gprivate.h
The H5Gget_comment() function returns the size of the comment
including the null terminator. If the object has no comment
then zero is returned. If an error occurs then a negative
value is returned.
./MANIFEST
./tools/Makefile.in
./tools/h5tools.h [NEW]
./tools/h5dump.c [NEW]
Created a library for printing values of datasets in a way
that looks nice. It's not done yet, but I needed it for
debugging the contents of files from Jim Reus.
./tools/h5ls.c
Added the `-d' and `--dump' options which cause the contents
of a dataset to be printed. Added `-w N' and `--width=N'
options to control how wide the raw data output should be. If
you want single-column output then say `-w1'.
Printing dataset values can now handle datasets of any integer
or floating point atomic type. As a special case, integers
which are one byte wide are treated a character strings for
now.
Sample output:
$ h5ls --dump --width=60 banana.hdf
ARCHIVE 0:0:0:744 Dataset {52/Inf}
Data:
(0) "U struct complex { double R; double I; };\012V"
(43) " double;\012"
U 0:0:0:2500 Dataset {256/512}
Data: printing of compound data types is not implemented yet
V 0:0:0:3928 Dataset {256/512}
Data:
(0) 0, 0.015625, 0.03125, 0.046875, 0.0625,
(5) 0.078125, 0.09375, 0.109375, 0.125, 0.140625,
(10) 0.15625, 0.171875, 0.1875, 0.203125, 0.21875,
(15) 0.234375, 0.25, 0.265625, 0.28125, 0.296875,
...
1998-07-24 05:19:17 +08:00
|
|
|
|
/* Command-line switches */
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
static int verbose_g = 0; /*lots of extra output */
|
|
|
|
|
static int width_g = 80; /*output width in characters */
|
1999-04-30 23:54:52 +08:00
|
|
|
|
static hbool_t address_g = FALSE; /*print raw data addresses */
|
|
|
|
|
static hbool_t data_g = FALSE; /*display dataset values? */
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
static hbool_t label_g = FALSE; /*label compound values? */
|
|
|
|
|
static hbool_t string_g = FALSE; /*print 1-byte numbers as ASCII? */
|
1999-01-22 02:33:39 +08:00
|
|
|
|
static hbool_t fullname_g = FALSE; /*print full path names */
|
|
|
|
|
static hbool_t recursive_g = FALSE; /*recursive descent listing */
|
1999-04-16 03:57:50 +08:00
|
|
|
|
static hbool_t grp_literal_g = FALSE; /*list group, not contents */
|
1999-06-07 23:05:02 +08:00
|
|
|
|
static hbool_t hexdump_g = FALSE; /*show data as raw hexadecimal */
|
1999-08-24 20:52:10 +08:00
|
|
|
|
static hbool_t show_errors_g = FALSE; /*print HDF5 error messages */
|
1999-10-01 00:14:14 +08:00
|
|
|
|
static hbool_t simple_output_g = FALSE; /*make output more machine-readable */
|
1999-01-22 02:33:39 +08:00
|
|
|
|
|
|
|
|
|
/* Info to pass to the iteration functions */
|
|
|
|
|
typedef struct iter_t {
|
|
|
|
|
const char *container; /*full name of the container object */
|
|
|
|
|
} iter_t;
|
|
|
|
|
|
|
|
|
|
/* Table containing object id and object name */
|
|
|
|
|
static struct {
|
|
|
|
|
int nalloc; /*number of slots allocated */
|
|
|
|
|
int nobjs; /*number of objects */
|
|
|
|
|
struct {
|
|
|
|
|
unsigned long id[2]; /*object number */
|
|
|
|
|
char *name; /*full object name */
|
|
|
|
|
} *obj;
|
|
|
|
|
} idtab_g;
|
1998-07-18 03:03:43 +08:00
|
|
|
|
|
1998-08-28 23:24:12 +08:00
|
|
|
|
/* Information about how to display each type of object */
|
|
|
|
|
static struct dispatch_t {
|
|
|
|
|
const char *name;
|
|
|
|
|
hid_t (*open)(hid_t loc, const char *name);
|
|
|
|
|
herr_t (*close)(hid_t obj);
|
|
|
|
|
herr_t (*list1)(hid_t obj);
|
1999-01-22 02:33:39 +08:00
|
|
|
|
herr_t (*list2)(hid_t obj, const char *name);
|
1998-08-28 23:24:12 +08:00
|
|
|
|
} dispatch_g[H5G_NTYPES];
|
|
|
|
|
|
|
|
|
|
#define DISPATCH(TYPE,NAME,OPEN,CLOSE,LIST1,LIST2) { \
|
|
|
|
|
dispatch_g[TYPE].name = (NAME); \
|
|
|
|
|
dispatch_g[TYPE].open = (OPEN); \
|
|
|
|
|
dispatch_g[TYPE].close = (CLOSE); \
|
|
|
|
|
dispatch_g[TYPE].list1 = (LIST1); \
|
|
|
|
|
dispatch_g[TYPE].list2 = (LIST2); \
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static herr_t list (hid_t group, const char *name, void *cd);
|
1999-06-15 22:58:25 +08:00
|
|
|
|
static void display_type(hid_t type, int ind);
|
1999-01-22 02:33:39 +08:00
|
|
|
|
static char *fix_name(const char *path, const char *base);
|
1998-08-28 23:24:12 +08:00
|
|
|
|
|
1998-07-18 03:03:43 +08:00
|
|
|
|
|
|
|
|
|
/*-------------------------------------------------------------------------
|
|
|
|
|
* Function: usage
|
|
|
|
|
*
|
|
|
|
|
* Purpose: Prints a usage message on stderr and then returns.
|
|
|
|
|
*
|
|
|
|
|
* Return: void
|
|
|
|
|
*
|
|
|
|
|
* Programmer: Robb Matzke
|
|
|
|
|
* Thursday, July 16, 1998
|
|
|
|
|
*
|
|
|
|
|
* Modifications:
|
|
|
|
|
*
|
|
|
|
|
*-------------------------------------------------------------------------
|
|
|
|
|
*/
|
1998-05-14 01:58:24 +08:00
|
|
|
|
static void
|
|
|
|
|
usage (const char *progname)
|
|
|
|
|
{
|
1998-07-18 03:03:43 +08:00
|
|
|
|
fprintf(stderr, "\
|
1999-07-14 01:59:21 +08:00
|
|
|
|
usage: %s [OPTIONS] [OBJECTS...]\n\
|
1998-07-18 03:03:43 +08:00
|
|
|
|
OPTIONS\n\
|
|
|
|
|
-h, -?, --help Print a usage message and exit\n\
|
1999-04-30 23:54:52 +08:00
|
|
|
|
-a, --address Print addresses for raw data\n\
|
|
|
|
|
-d, --data Print the values of datasets\n\
|
1999-08-24 20:52:10 +08:00
|
|
|
|
-e, --errors Show all HDF5 error reporting\n\
|
1999-01-22 02:33:39 +08:00
|
|
|
|
-f, --full Print full path names instead of base names\n\
|
1999-04-16 03:57:50 +08:00
|
|
|
|
-g, --group Show information about a group, not its contents\n\
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
-l, --label Label members of compound datasets\n\
|
1999-01-22 02:33:39 +08:00
|
|
|
|
-r, --recursive List all groups recursively, avoiding cycles\n\
|
[svn-r876] Changes since 19981102
----------------------
./bin/snapshot
Made same fix as for the release script yesterday.
./src/H5D.c
./src/H5Dprivate.h
./src/H5G.c
./src/H5Gprivate.h
./src/H5Gpublic.h
./src/H5O.c
./src/H5Oprivate.h
./src/H5RA.c
./src/H5RAprivate.h
./src/H5T.c
./src/H5Tprivate.h
Improved object type checking. Instead of determining the
object type by trying to open each of the possible types, we
keep a table of associations between object type number (like
H5G_GROUP, H5G_DATASET, H5D_TYPE, and H5D_RAGGED) and an `isa'
function that returns true if the object header has the right
messages to make the object a particular type. This mechanism
also allows specialization of object types by permitting an
object to satisfy more than one `isa' function.
Added `isa' functions for groups, datasets, ragged arrays, and
committed data types.
./src/H5config.h.in
Added HAVE_STAT_ST_BLOCKS. I thought this had already been
added, but apparently not.
./tools/h5ls.c
Removed system include files since they're already included by
H5private.h and since I wasn't including them portably anyway.
By default, 1-byte integer types are printed as integer values
instead of ASCII characters. However, the `-s' or `--string'
command-line switch causes the data to be interpretted as
ASCII. String data types are always printed as character
data.
Ragged arrays are now identified as ragged arrays and h5ls
doesn't descend into the group automatically. This uses the
new object type specialization stuff.
./tools/h5tools.c
./tools/h5tools.h
Added the ability to print 1-byte integer types as either
ASCII or numeric data instead of always ASCII. The default is
to print as numeric data.
1998-11-06 04:28:34 +08:00
|
|
|
|
-s, --string Print 1-byte integer datasets as ASCII\n\
|
1999-10-01 00:14:14 +08:00
|
|
|
|
-S, --simple Use a machine-readable output format\n\
|
[svn-r537] Changes since 19980722
----------------------
./src/H5A.c
./src/H5Apublic.h
./test/tattr.c
Switched the order of the second and third argument of
H5Aget_name() to make it consistent with other functions that
take buffers and buffer sizes.
./src/H5G.c
./src/H5Gpublic.h
./src/H5Gprivate.h
The H5Gget_comment() function returns the size of the comment
including the null terminator. If the object has no comment
then zero is returned. If an error occurs then a negative
value is returned.
./MANIFEST
./tools/Makefile.in
./tools/h5tools.h [NEW]
./tools/h5dump.c [NEW]
Created a library for printing values of datasets in a way
that looks nice. It's not done yet, but I needed it for
debugging the contents of files from Jim Reus.
./tools/h5ls.c
Added the `-d' and `--dump' options which cause the contents
of a dataset to be printed. Added `-w N' and `--width=N'
options to control how wide the raw data output should be. If
you want single-column output then say `-w1'.
Printing dataset values can now handle datasets of any integer
or floating point atomic type. As a special case, integers
which are one byte wide are treated a character strings for
now.
Sample output:
$ h5ls --dump --width=60 banana.hdf
ARCHIVE 0:0:0:744 Dataset {52/Inf}
Data:
(0) "U struct complex { double R; double I; };\012V"
(43) " double;\012"
U 0:0:0:2500 Dataset {256/512}
Data: printing of compound data types is not implemented yet
V 0:0:0:3928 Dataset {256/512}
Data:
(0) 0, 0.015625, 0.03125, 0.046875, 0.0625,
(5) 0.078125, 0.09375, 0.109375, 0.125, 0.140625,
(10) 0.15625, 0.171875, 0.1875, 0.203125, 0.21875,
(15) 0.234375, 0.25, 0.265625, 0.28125, 0.296875,
...
1998-07-24 05:19:17 +08:00
|
|
|
|
-wN, --width=N Set the number of columns of output\n\
|
1998-07-18 03:03:43 +08:00
|
|
|
|
-v, --verbose Generate more verbose output\n\
|
1998-07-21 00:05:43 +08:00
|
|
|
|
-V, --version Print version number and exit\n\
|
1999-06-07 23:05:02 +08:00
|
|
|
|
-x, --hexdump Show raw data in hexadecimal format\n\
|
1999-07-14 01:59:21 +08:00
|
|
|
|
\n\
|
1998-09-10 22:18:17 +08:00
|
|
|
|
OBJECTS\n\
|
1999-07-14 01:59:21 +08:00
|
|
|
|
Each object consists of an HDF5 file name optionally followed by a\n\
|
|
|
|
|
slash and an object name within the file (if no object is specified\n\
|
|
|
|
|
within the file then the contents of the root group are dispalyed).\n\
|
|
|
|
|
The file name may include a printf(3C) integer format such as\n\
|
|
|
|
|
\"%%05d\" to open a file family.\n",
|
1998-09-10 22:18:17 +08:00
|
|
|
|
progname);
|
1998-05-14 01:58:24 +08:00
|
|
|
|
}
|
|
|
|
|
|
1999-01-22 02:33:39 +08:00
|
|
|
|
|
|
|
|
|
/*-------------------------------------------------------------------------
|
|
|
|
|
* Function: sym_insert
|
|
|
|
|
*
|
|
|
|
|
* Purpose: Add a symbol to the table.
|
|
|
|
|
*
|
|
|
|
|
* Return: void
|
|
|
|
|
*
|
|
|
|
|
* Programmer: Robb Matzke
|
|
|
|
|
* Thursday, January 21, 1999
|
|
|
|
|
*
|
|
|
|
|
* Modifications:
|
|
|
|
|
*
|
|
|
|
|
*-------------------------------------------------------------------------
|
|
|
|
|
*/
|
|
|
|
|
static void
|
|
|
|
|
sym_insert(H5G_stat_t *sb, const char *name)
|
|
|
|
|
{
|
|
|
|
|
int n;
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* Don't add it if the link count is 1 because such an object can only
|
|
|
|
|
* have one name.
|
|
|
|
|
*/
|
|
|
|
|
if (sb->nlink<2) return;
|
|
|
|
|
|
|
|
|
|
/* Extend the table */
|
|
|
|
|
if (idtab_g.nobjs>=idtab_g.nalloc) {
|
|
|
|
|
idtab_g.nalloc = MAX(256, 2*idtab_g.nalloc);
|
|
|
|
|
idtab_g.obj = realloc(idtab_g.obj,
|
|
|
|
|
idtab_g.nalloc*sizeof(idtab_g.obj[0]));
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Insert the entry */
|
|
|
|
|
n = idtab_g.nobjs++;
|
|
|
|
|
idtab_g.obj[n].id[0] = sb->objno[0];
|
|
|
|
|
idtab_g.obj[n].id[1] = sb->objno[1];
|
|
|
|
|
idtab_g.obj[n].name = malloc(strlen(name)+1);
|
|
|
|
|
strcpy(idtab_g.obj[n].name, name);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
/*-------------------------------------------------------------------------
|
|
|
|
|
* Function: sym_lookup
|
|
|
|
|
*
|
|
|
|
|
* Purpose: Find another name for the specified object.
|
|
|
|
|
*
|
|
|
|
|
* Return: Success: Ptr to another name.
|
|
|
|
|
*
|
|
|
|
|
* Failure: NULL
|
|
|
|
|
*
|
|
|
|
|
* Programmer: Robb Matzke
|
|
|
|
|
* Thursday, January 21, 1999
|
|
|
|
|
*
|
|
|
|
|
* Modifications:
|
|
|
|
|
*
|
|
|
|
|
*-------------------------------------------------------------------------
|
|
|
|
|
*/
|
|
|
|
|
static char *
|
|
|
|
|
sym_lookup(H5G_stat_t *sb)
|
|
|
|
|
{
|
|
|
|
|
int n;
|
|
|
|
|
|
|
|
|
|
if (sb->nlink<2) return NULL; /*only one name possible*/
|
|
|
|
|
for (n=0; n<idtab_g.nobjs; n++) {
|
|
|
|
|
if (idtab_g.obj[n].id[0]==sb->objno[0] &&
|
|
|
|
|
idtab_g.obj[n].id[1]==sb->objno[1]) {
|
|
|
|
|
return idtab_g.obj[n].name;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
return NULL;
|
|
|
|
|
}
|
|
|
|
|
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
|
|
|
|
|
/*-------------------------------------------------------------------------
|
|
|
|
|
* Function: display_string
|
|
|
|
|
*
|
1999-01-22 02:33:39 +08:00
|
|
|
|
* Purpose: Print a string value by escaping unusual characters. If
|
|
|
|
|
* STREAM is null then we only count how large the output would
|
|
|
|
|
* be.
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
*
|
|
|
|
|
* Return: Number of characters printed.
|
|
|
|
|
*
|
|
|
|
|
* Programmer: Robb Matzke
|
|
|
|
|
* Thursday, November 5, 1998
|
|
|
|
|
*
|
|
|
|
|
* Modifications:
|
|
|
|
|
*
|
|
|
|
|
*-------------------------------------------------------------------------
|
|
|
|
|
*/
|
|
|
|
|
static int
|
1999-01-22 02:33:39 +08:00
|
|
|
|
display_string(FILE *stream, const char *s, hbool_t escape_spaces)
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
{
|
|
|
|
|
int nprint=0;
|
|
|
|
|
|
|
|
|
|
for (/*void*/; s && *s; s++) {
|
|
|
|
|
switch (*s) {
|
|
|
|
|
case '"':
|
1999-01-22 02:33:39 +08:00
|
|
|
|
if (stream) fprintf(stream, "\\\"");
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
nprint += 2;
|
|
|
|
|
break;
|
|
|
|
|
case '\\':
|
1999-01-22 02:33:39 +08:00
|
|
|
|
if (stream) fprintf(stream, "\\\\");
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
nprint += 2;
|
|
|
|
|
break;
|
|
|
|
|
case '\b':
|
1999-01-22 02:33:39 +08:00
|
|
|
|
if (stream) fprintf(stream, "\\b");
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
nprint += 2;
|
|
|
|
|
break;
|
|
|
|
|
case '\f':
|
1999-01-22 02:33:39 +08:00
|
|
|
|
if (stream) fprintf(stream, "\\f");
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
nprint += 2;
|
|
|
|
|
break;
|
|
|
|
|
case '\n':
|
1999-01-22 02:33:39 +08:00
|
|
|
|
if (stream) fprintf(stream, "\\n");
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
nprint += 2;
|
|
|
|
|
break;
|
|
|
|
|
case '\r':
|
1999-01-22 02:33:39 +08:00
|
|
|
|
if (stream) fprintf(stream, "\\r");
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
nprint += 2;
|
|
|
|
|
break;
|
|
|
|
|
case '\t':
|
1999-01-22 02:33:39 +08:00
|
|
|
|
if (stream) fprintf(stream, "\\t");
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
nprint += 2;
|
|
|
|
|
break;
|
1999-01-22 02:33:39 +08:00
|
|
|
|
case ' ':
|
|
|
|
|
if (escape_spaces) {
|
|
|
|
|
if (stream) fprintf(stream, "\\ ");
|
|
|
|
|
nprint += 2;
|
|
|
|
|
} else {
|
|
|
|
|
if (stream) fprintf(stream, " ");
|
|
|
|
|
nprint++;
|
|
|
|
|
}
|
|
|
|
|
break;
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
default:
|
1999-09-28 00:11:15 +08:00
|
|
|
|
if (isprint((int)*s)) {
|
1999-01-22 02:33:39 +08:00
|
|
|
|
if (stream) putc(*s, stream);
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
nprint++;
|
|
|
|
|
} else {
|
1999-01-22 02:33:39 +08:00
|
|
|
|
if (stream) {
|
|
|
|
|
fprintf(stream, "\\%03o", *((const unsigned char*)s));
|
|
|
|
|
}
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
nprint += 4;
|
|
|
|
|
}
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
return nprint;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
/*-------------------------------------------------------------------------
|
|
|
|
|
* Function: display_native_type
|
|
|
|
|
*
|
|
|
|
|
* Purpose: Prints the name of a native C data type.
|
|
|
|
|
*
|
|
|
|
|
* Return: Success: TRUE
|
|
|
|
|
*
|
|
|
|
|
* Failure: FALSE, nothing printed.
|
|
|
|
|
*
|
|
|
|
|
* Programmer: Robb Matzke
|
|
|
|
|
* Thursday, November 5, 1998
|
|
|
|
|
*
|
|
|
|
|
* Modifications:
|
1999-06-14 23:07:58 +08:00
|
|
|
|
* Robb Matzke, 1999-06-11
|
|
|
|
|
* Added the C9x types, but we still prefer to display the types
|
|
|
|
|
* from the C language itself (like `int' vs. `int32_t').
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
*
|
|
|
|
|
*-------------------------------------------------------------------------
|
|
|
|
|
*/
|
|
|
|
|
static hbool_t
|
1999-06-15 22:58:25 +08:00
|
|
|
|
display_native_type(hid_t type, int UNUSED ind)
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
{
|
1998-11-26 01:21:21 +08:00
|
|
|
|
if (H5Tequal(type, H5T_NATIVE_SCHAR)) {
|
|
|
|
|
printf("native signed char");
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_UCHAR)) {
|
|
|
|
|
printf("native unsigned char");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_INT)) {
|
|
|
|
|
printf("native int");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_UINT)) {
|
|
|
|
|
printf("native unsigned int");
|
1998-11-20 01:36:27 +08:00
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_SHORT)) {
|
|
|
|
|
printf("native short");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_USHORT)) {
|
|
|
|
|
printf("native unsigned short");
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_LONG)) {
|
|
|
|
|
printf("native long");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_ULONG)) {
|
|
|
|
|
printf("native unsigned long");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_LLONG)) {
|
|
|
|
|
printf("native long long");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_ULLONG)) {
|
|
|
|
|
printf("native unsigned long long");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_FLOAT)) {
|
|
|
|
|
printf("native float");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_DOUBLE)) {
|
|
|
|
|
printf("native double");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_LDOUBLE)) {
|
|
|
|
|
printf("native long double");
|
1999-06-14 23:07:58 +08:00
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_INT8)) {
|
|
|
|
|
printf("native int8_t");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_UINT8)) {
|
|
|
|
|
printf("native uint8_t");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_INT16)) {
|
|
|
|
|
printf("native int16_t");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_UINT16)) {
|
|
|
|
|
printf("native uint16_t");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_INT32)) {
|
|
|
|
|
printf("native int32_t");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_UINT32)) {
|
|
|
|
|
printf("native uint32_t");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_INT64)) {
|
|
|
|
|
printf("native int64_t");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_UINT64)) {
|
|
|
|
|
printf("native uint64_t");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_INT_LEAST8)) {
|
|
|
|
|
printf("native int_least8_t");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_UINT_LEAST8)) {
|
|
|
|
|
printf("native uint_least8_t");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_INT_LEAST16)) {
|
|
|
|
|
printf("native int_least16_t");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_UINT_LEAST16)) {
|
|
|
|
|
printf("native uint_least16_t");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_INT_LEAST32)) {
|
|
|
|
|
printf("native int_least32_t");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_UINT_LEAST32)) {
|
|
|
|
|
printf("native uint_least32_t");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_INT_LEAST64)) {
|
|
|
|
|
printf("native int_least64_t");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_UINT_LEAST64)) {
|
|
|
|
|
printf("native uint_least64_t");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_INT_FAST8)) {
|
|
|
|
|
printf("native int_fast8_t");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_UINT_FAST8)) {
|
|
|
|
|
printf("native uint_fast8_t");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_INT_FAST16)) {
|
|
|
|
|
printf("native int_fast16_t");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_UINT_FAST16)) {
|
|
|
|
|
printf("native uint_fast16_t");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_INT_FAST32)) {
|
|
|
|
|
printf("native int_fast32_t");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_UINT_FAST32)) {
|
|
|
|
|
printf("native uint_fast32_t");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_INT_FAST64)) {
|
|
|
|
|
printf("native int_fast64_t");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_UINT_FAST64)) {
|
|
|
|
|
printf("native uint_fast64_t");
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_B8)) {
|
|
|
|
|
printf("native 8-bit field");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_B16)) {
|
|
|
|
|
printf("native 16-bit field");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_B32)) {
|
|
|
|
|
printf("native 32-bit field");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_B64)) {
|
|
|
|
|
printf("native 64-bit field");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_HSIZE)) {
|
|
|
|
|
printf("native hsize_t");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_HSSIZE)) {
|
|
|
|
|
printf("native hssize_t");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_HERR)) {
|
|
|
|
|
printf("native herr_t");
|
|
|
|
|
} else if (H5Tequal(type, H5T_NATIVE_HBOOL)) {
|
|
|
|
|
printf("native hbool_t");
|
|
|
|
|
} else {
|
|
|
|
|
return FALSE;
|
|
|
|
|
}
|
|
|
|
|
return TRUE;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
/*-------------------------------------------------------------------------
|
|
|
|
|
* Function: display_ieee_type
|
|
|
|
|
*
|
|
|
|
|
* Purpose: Print the name of an IEEE floating-point data type.
|
|
|
|
|
*
|
|
|
|
|
* Return: Success: TRUE
|
|
|
|
|
*
|
|
|
|
|
* Failure: FALSE, nothing printed
|
|
|
|
|
*
|
|
|
|
|
* Programmer: Robb Matzke
|
|
|
|
|
* Thursday, November 5, 1998
|
|
|
|
|
*
|
|
|
|
|
* Modifications:
|
|
|
|
|
*
|
|
|
|
|
*-------------------------------------------------------------------------
|
|
|
|
|
*/
|
|
|
|
|
static hbool_t
|
1999-06-15 22:58:25 +08:00
|
|
|
|
display_ieee_type(hid_t type, int UNUSED ind)
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
{
|
|
|
|
|
if (H5Tequal(type, H5T_IEEE_F32BE)) {
|
|
|
|
|
printf("IEEE 32-bit big-endian float");
|
|
|
|
|
} else if (H5Tequal(type, H5T_IEEE_F32LE)) {
|
|
|
|
|
printf("IEEE 32-bit little-endian float");
|
|
|
|
|
} else if (H5Tequal(type, H5T_IEEE_F64BE)) {
|
|
|
|
|
printf("IEEE 64-bit big-endian float");
|
|
|
|
|
} else if (H5Tequal(type, H5T_IEEE_F64LE)) {
|
|
|
|
|
printf("IEEE 64-bit little-endian float");
|
|
|
|
|
} else {
|
|
|
|
|
return FALSE;
|
|
|
|
|
}
|
|
|
|
|
return TRUE;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
/*-------------------------------------------------------------------------
|
|
|
|
|
* Function: display_precision
|
|
|
|
|
*
|
|
|
|
|
* Purpose: Prints information on the next line about precision and
|
|
|
|
|
* padding if the precision is less than the total data type
|
|
|
|
|
* size.
|
|
|
|
|
*
|
|
|
|
|
* Return: void
|
|
|
|
|
*
|
|
|
|
|
* Programmer: Robb Matzke
|
|
|
|
|
* Thursday, November 5, 1998
|
|
|
|
|
*
|
|
|
|
|
* Modifications:
|
|
|
|
|
*
|
|
|
|
|
*-------------------------------------------------------------------------
|
|
|
|
|
*/
|
|
|
|
|
static void
|
1999-06-15 22:58:25 +08:00
|
|
|
|
display_precision(hid_t type, int ind)
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
{
|
|
|
|
|
size_t prec; /*precision */
|
|
|
|
|
H5T_pad_t plsb, pmsb; /*lsb and msb padding */
|
|
|
|
|
const char *plsb_s=NULL; /*lsb padding string */
|
|
|
|
|
const char *pmsb_s=NULL; /*msb padding string */
|
|
|
|
|
size_t nbits; /*number of bits */
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* If the precision is less than the total size then show the precision
|
|
|
|
|
* and offset on the following line. Also display the padding
|
|
|
|
|
* information.
|
|
|
|
|
*/
|
|
|
|
|
if (8*H5Tget_size(type)!=(prec=H5Tget_precision(type))) {
|
|
|
|
|
printf("\n%*s(%lu bit%s of precision beginning at bit %lu)",
|
1999-06-15 22:58:25 +08:00
|
|
|
|
ind, "", (unsigned long)prec, 1==prec?"":"s",
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
(unsigned long)H5Tget_offset(type));
|
|
|
|
|
|
|
|
|
|
H5Tget_pad(type, &plsb, &pmsb);
|
|
|
|
|
if (H5Tget_offset(type)>0) {
|
|
|
|
|
switch (plsb) {
|
|
|
|
|
case H5T_PAD_ZERO:
|
|
|
|
|
plsb_s = "zero";
|
|
|
|
|
break;
|
|
|
|
|
case H5T_PAD_ONE:
|
|
|
|
|
plsb_s = "one";
|
|
|
|
|
break;
|
|
|
|
|
case H5T_PAD_BACKGROUND:
|
|
|
|
|
plsb_s = "bkg";
|
|
|
|
|
break;
|
|
|
|
|
case H5T_PAD_ERROR:
|
|
|
|
|
case H5T_NPAD:
|
|
|
|
|
plsb_s = "unknown";
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
if (H5Tget_offset(type)+prec<8*H5Tget_size(type)) {
|
|
|
|
|
switch (pmsb) {
|
|
|
|
|
case H5T_PAD_ZERO:
|
|
|
|
|
pmsb_s = "zero";
|
|
|
|
|
break;
|
|
|
|
|
case H5T_PAD_ONE:
|
|
|
|
|
pmsb_s = "one";
|
|
|
|
|
break;
|
|
|
|
|
case H5T_PAD_BACKGROUND:
|
|
|
|
|
pmsb_s = "bkg";
|
|
|
|
|
break;
|
|
|
|
|
case H5T_PAD_ERROR:
|
|
|
|
|
case H5T_NPAD:
|
|
|
|
|
pmsb_s = "unknown";
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
if (plsb_s || pmsb_s) {
|
1999-06-15 22:58:25 +08:00
|
|
|
|
printf("\n%*s(", ind, "");
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
if (plsb_s) {
|
|
|
|
|
nbits = H5Tget_offset(type);
|
|
|
|
|
printf("%lu %s bit%s at bit 0",
|
|
|
|
|
(unsigned long)nbits, plsb_s, 1==nbits?"":"s");
|
|
|
|
|
}
|
|
|
|
|
if (plsb_s && pmsb_s) printf(", ");
|
|
|
|
|
if (pmsb_s) {
|
|
|
|
|
nbits = 8*H5Tget_size(type)-(H5Tget_offset(type)+prec);
|
|
|
|
|
printf("%lu %s bit%s at bit %lu",
|
|
|
|
|
(unsigned long)nbits, pmsb_s, 1==nbits?"":"s",
|
|
|
|
|
(unsigned long)(8*H5Tget_size(type)-nbits));
|
|
|
|
|
}
|
|
|
|
|
printf(")");
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
/*-------------------------------------------------------------------------
|
|
|
|
|
* Function: display_int_type
|
|
|
|
|
*
|
|
|
|
|
* Purpose: Print the name of an integer data type. Common information
|
|
|
|
|
* like number of bits, byte order, and sign scheme appear on
|
|
|
|
|
* the first line. Additional information might appear in
|
|
|
|
|
* parentheses on the following lines.
|
|
|
|
|
*
|
|
|
|
|
* Return: Success: TRUE
|
|
|
|
|
*
|
|
|
|
|
* Failure: FALSE, nothing printed
|
|
|
|
|
*
|
|
|
|
|
* Programmer: Robb Matzke
|
|
|
|
|
* Thursday, November 5, 1998
|
|
|
|
|
*
|
|
|
|
|
* Modifications:
|
|
|
|
|
*
|
|
|
|
|
*-------------------------------------------------------------------------
|
|
|
|
|
*/
|
|
|
|
|
static hbool_t
|
1999-06-15 22:58:25 +08:00
|
|
|
|
display_int_type(hid_t type, int ind)
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
{
|
|
|
|
|
H5T_order_t order; /*byte order value */
|
|
|
|
|
const char *order_s=NULL; /*byte order string */
|
|
|
|
|
H5T_sign_t sign; /*sign scheme value */
|
|
|
|
|
const char *sign_s=NULL; /*sign scheme string */
|
|
|
|
|
|
|
|
|
|
if (H5T_INTEGER!=H5Tget_class(type)) return FALSE;
|
|
|
|
|
|
|
|
|
|
/* Byte order */
|
|
|
|
|
if (H5Tget_size(type)>1) {
|
|
|
|
|
order = H5Tget_order(type);
|
|
|
|
|
if (H5T_ORDER_LE==order) {
|
|
|
|
|
order_s = " little-endian";
|
|
|
|
|
} else if (H5T_ORDER_BE==order) {
|
|
|
|
|
order_s = " big-endian";
|
|
|
|
|
} else if (H5T_ORDER_VAX==order) {
|
|
|
|
|
order_s = " mixed-endian";
|
|
|
|
|
} else {
|
|
|
|
|
order_s = " unknown-byte-order";
|
|
|
|
|
}
|
|
|
|
|
} else {
|
|
|
|
|
order_s = "";
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Sign */
|
1999-01-07 19:42:27 +08:00
|
|
|
|
if ((sign=H5Tget_sign(type))>=0) {
|
|
|
|
|
if (H5T_SGN_NONE==sign) {
|
|
|
|
|
sign_s = " unsigned";
|
|
|
|
|
} else if (H5T_SGN_2==sign) {
|
|
|
|
|
sign_s = "";
|
|
|
|
|
} else {
|
|
|
|
|
sign_s = " unknown-sign";
|
|
|
|
|
}
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
} else {
|
|
|
|
|
sign_s = " unknown-sign";
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* Print size, order, and sign on first line, precision and padding
|
|
|
|
|
* information on the subsequent lines
|
|
|
|
|
*/
|
|
|
|
|
printf("%lu-bit%s%s integer",
|
|
|
|
|
(unsigned long)(8*H5Tget_size(type)), order_s, sign_s);
|
1999-06-15 22:58:25 +08:00
|
|
|
|
display_precision(type, ind);
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
return TRUE;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
/*-------------------------------------------------------------------------
|
|
|
|
|
* Function: display_float_type
|
|
|
|
|
*
|
|
|
|
|
* Purpose: Print info about a floating point data type.
|
|
|
|
|
*
|
|
|
|
|
* Return: Success: TRUE
|
|
|
|
|
*
|
|
|
|
|
* Failure: FALSE, nothing printed
|
|
|
|
|
*
|
|
|
|
|
* Programmer: Robb Matzke
|
|
|
|
|
* Thursday, November 5, 1998
|
|
|
|
|
*
|
|
|
|
|
* Modifications:
|
|
|
|
|
*
|
|
|
|
|
*-------------------------------------------------------------------------
|
|
|
|
|
*/
|
|
|
|
|
static hbool_t
|
1999-06-15 22:58:25 +08:00
|
|
|
|
display_float_type(hid_t type, int ind)
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
{
|
|
|
|
|
H5T_order_t order; /*byte order value */
|
|
|
|
|
const char *order_s=NULL; /*byte order string */
|
|
|
|
|
size_t spos; /*sign bit position */
|
|
|
|
|
size_t esize, epos; /*exponent size and position */
|
|
|
|
|
size_t msize, mpos; /*significand size and position */
|
|
|
|
|
size_t ebias; /*exponent bias */
|
|
|
|
|
H5T_norm_t norm; /*significand normalization */
|
|
|
|
|
const char *norm_s=NULL; /*normalization string */
|
|
|
|
|
H5T_pad_t pad; /*internal padding value */
|
|
|
|
|
const char *pad_s=NULL; /*internal padding string */
|
|
|
|
|
|
|
|
|
|
if (H5T_FLOAT!=H5Tget_class(type)) return FALSE;
|
|
|
|
|
|
|
|
|
|
/* Byte order */
|
|
|
|
|
if (H5Tget_size(type)>1) {
|
|
|
|
|
order = H5Tget_order(type);
|
|
|
|
|
if (H5T_ORDER_LE==order) {
|
|
|
|
|
order_s = " little-endian";
|
|
|
|
|
} else if (H5T_ORDER_BE==order) {
|
|
|
|
|
order_s = " big-endian";
|
|
|
|
|
} else if (H5T_ORDER_VAX==order) {
|
|
|
|
|
order_s = " mixed-endian";
|
|
|
|
|
} else {
|
|
|
|
|
order_s = " unknown-byte-order";
|
|
|
|
|
}
|
|
|
|
|
} else {
|
|
|
|
|
order_s = "";
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* Print size and byte order on first line, precision and padding on
|
|
|
|
|
* subsequent lines.
|
|
|
|
|
*/
|
|
|
|
|
printf("%lu-bit%s floating-point",
|
|
|
|
|
(unsigned long)(8*H5Tget_size(type)), order_s);
|
1999-06-15 22:58:25 +08:00
|
|
|
|
display_precision(type, ind);
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
|
|
|
|
|
/* Print sizes, locations, and other information about each field */
|
|
|
|
|
H5Tget_fields (type, &spos, &epos, &esize, &mpos, &msize);
|
|
|
|
|
ebias = H5Tget_ebias(type);
|
|
|
|
|
norm = H5Tget_norm(type);
|
|
|
|
|
switch (norm) {
|
|
|
|
|
case H5T_NORM_IMPLIED:
|
|
|
|
|
norm_s = ", msb implied";
|
|
|
|
|
break;
|
|
|
|
|
case H5T_NORM_MSBSET:
|
|
|
|
|
norm_s = ", msb always set";
|
|
|
|
|
break;
|
|
|
|
|
case H5T_NORM_NONE:
|
|
|
|
|
norm_s = ", no normalization";
|
|
|
|
|
break;
|
|
|
|
|
case H5T_NORM_ERROR:
|
|
|
|
|
norm_s = ", unknown normalization";
|
|
|
|
|
break;
|
|
|
|
|
}
|
1999-06-15 22:58:25 +08:00
|
|
|
|
printf("\n%*s(significant for %lu bit%s at bit %lu%s)", ind, "",
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
(unsigned long)msize, 1==msize?"":"s", (unsigned long)mpos,
|
|
|
|
|
norm_s);
|
|
|
|
|
printf("\n%*s(exponent for %lu bit%s at bit %lu, bias is 0x%lx)",
|
1999-06-15 22:58:25 +08:00
|
|
|
|
ind, "", (unsigned long)esize, 1==esize?"":"s",
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
(unsigned long)epos, (unsigned long)ebias);
|
1999-06-15 22:58:25 +08:00
|
|
|
|
printf("\n%*s(sign bit at %lu)", ind, "", (unsigned long)spos);
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
|
|
|
|
|
/* Display internal padding */
|
|
|
|
|
if (1+esize+msize<H5Tget_precision(type)) {
|
|
|
|
|
pad = H5Tget_inpad(type);
|
|
|
|
|
switch (pad) {
|
|
|
|
|
case H5T_PAD_ZERO:
|
|
|
|
|
pad_s = "zero";
|
|
|
|
|
break;
|
|
|
|
|
case H5T_PAD_ONE:
|
|
|
|
|
pad_s = "one";
|
|
|
|
|
break;
|
|
|
|
|
case H5T_PAD_BACKGROUND:
|
|
|
|
|
pad_s = "bkg";
|
|
|
|
|
break;
|
|
|
|
|
case H5T_PAD_ERROR:
|
|
|
|
|
case H5T_NPAD:
|
|
|
|
|
pad_s = "unknown";
|
|
|
|
|
break;
|
|
|
|
|
}
|
1999-06-15 22:58:25 +08:00
|
|
|
|
printf("\n%*s(internal padding bits are %s)", ind, "", pad_s);
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
}
|
|
|
|
|
return TRUE;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
/*-------------------------------------------------------------------------
|
|
|
|
|
* Function: display_cmpd_type
|
|
|
|
|
*
|
|
|
|
|
* Purpose: Print info about a compound data type.
|
|
|
|
|
*
|
|
|
|
|
* Return: Success: TRUE
|
|
|
|
|
*
|
|
|
|
|
* Failure: FALSE, nothing printed
|
|
|
|
|
*
|
|
|
|
|
* Programmer: Robb Matzke
|
|
|
|
|
* Thursday, November 5, 1998
|
|
|
|
|
*
|
|
|
|
|
* Modifications:
|
|
|
|
|
*
|
|
|
|
|
*-------------------------------------------------------------------------
|
|
|
|
|
*/
|
|
|
|
|
static hbool_t
|
1999-06-15 22:58:25 +08:00
|
|
|
|
display_cmpd_type(hid_t type, int ind)
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
{
|
|
|
|
|
char *name=NULL; /*member name */
|
|
|
|
|
int ndims; /*dimensionality */
|
|
|
|
|
size_t dims[8]; /*dimensions */
|
1999-03-30 19:38:34 +08:00
|
|
|
|
size_t size; /*total size of type in bytes */
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
int perm[8]; /*index permutation */
|
|
|
|
|
hid_t subtype; /*member data type */
|
|
|
|
|
int i, j, n; /*miscellaneous counters */
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
if (H5T_COMPOUND!=H5Tget_class(type)) return FALSE;
|
|
|
|
|
printf("struct {");
|
|
|
|
|
for (i=0; i<H5Tget_nmembers(type); i++) {
|
|
|
|
|
|
|
|
|
|
/* Name and offset */
|
|
|
|
|
name = H5Tget_member_name(type, i);
|
1999-06-15 22:58:25 +08:00
|
|
|
|
printf("\n%*s\"", ind+4, "");
|
1999-01-22 02:33:39 +08:00
|
|
|
|
n = display_string(stdout, name, FALSE);
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
printf("\"%*s +%-4lu ", MAX(0, 16-n), "",
|
|
|
|
|
(unsigned long)H5Tget_member_offset(type, i));
|
|
|
|
|
free(name);
|
|
|
|
|
|
|
|
|
|
/* Dimensions and permutation */
|
|
|
|
|
ndims = H5Tget_member_dims(type, i, dims, perm);
|
|
|
|
|
if (ndims>0) {
|
|
|
|
|
printf("[");
|
|
|
|
|
for (j=0; j<ndims; j++) {
|
|
|
|
|
printf("%s%lu", j?",":"", (unsigned long)(dims[j]));
|
|
|
|
|
}
|
|
|
|
|
printf("]");
|
|
|
|
|
for (j=0; j<ndims; j++) {
|
|
|
|
|
if (perm[j]!=j) break;
|
|
|
|
|
}
|
|
|
|
|
if (j<ndims) {
|
|
|
|
|
printf("x[");
|
|
|
|
|
for (j=0; j<ndims; j++) {
|
|
|
|
|
printf("%s%d", j?",":"", perm[j]);
|
|
|
|
|
}
|
|
|
|
|
printf("]");
|
|
|
|
|
}
|
|
|
|
|
printf(" ");
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Data type */
|
|
|
|
|
subtype = H5Tget_member_type(type, i);
|
1999-06-15 22:58:25 +08:00
|
|
|
|
display_type(subtype, ind+4);
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
H5Tclose(subtype);
|
|
|
|
|
}
|
1999-03-30 19:38:34 +08:00
|
|
|
|
size = H5Tget_size(type);
|
|
|
|
|
printf("\n%*s} %lu byte%s",
|
1999-06-15 22:58:25 +08:00
|
|
|
|
ind, "", (unsigned long)size, 1==size?"":"s");
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
return TRUE;
|
|
|
|
|
}
|
|
|
|
|
|
1999-01-07 19:42:27 +08:00
|
|
|
|
|
|
|
|
|
/*-------------------------------------------------------------------------
|
|
|
|
|
* Function: display_enum_type
|
|
|
|
|
*
|
|
|
|
|
* Purpose: Print info about an enumeration data type.
|
|
|
|
|
*
|
|
|
|
|
* Return: Success: TRUE
|
|
|
|
|
*
|
|
|
|
|
* Failure: FALSE, nothing printed
|
|
|
|
|
*
|
|
|
|
|
* Programmer: Robb Matzke
|
|
|
|
|
* Wednesday, December 23, 1998
|
|
|
|
|
*
|
|
|
|
|
* Modifications:
|
|
|
|
|
*
|
|
|
|
|
*-------------------------------------------------------------------------
|
|
|
|
|
*/
|
|
|
|
|
static hbool_t
|
1999-06-15 22:58:25 +08:00
|
|
|
|
display_enum_type(hid_t type, int ind)
|
1999-01-07 19:42:27 +08:00
|
|
|
|
{
|
|
|
|
|
char **name=NULL; /*member names */
|
|
|
|
|
unsigned char *value=NULL; /*value array */
|
|
|
|
|
int nmembs; /*number of members */
|
|
|
|
|
int nchars; /*number of output characters */
|
|
|
|
|
hid_t super; /*enum base integer type */
|
|
|
|
|
hid_t native=-1; /*native integer data type */
|
|
|
|
|
size_t dst_size; /*destination value type size */
|
|
|
|
|
int i; /*miscellaneous counters */
|
|
|
|
|
size_t j;
|
|
|
|
|
|
|
|
|
|
if (H5T_ENUM!=H5Tget_class(type)) return FALSE;
|
|
|
|
|
nmembs = H5Tget_nmembers(type);
|
|
|
|
|
super = H5Tget_super(type);
|
|
|
|
|
printf("enum ");
|
1999-06-15 22:58:25 +08:00
|
|
|
|
display_type(super, ind+4);
|
1999-01-07 19:42:27 +08:00
|
|
|
|
printf(" {");
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* Determine what data type to use for the native values. To simplify
|
|
|
|
|
* things we entertain three possibilities:
|
|
|
|
|
* 1. long_long -- the largest native signed integer
|
|
|
|
|
* 2. unsigned long_long -- the largest native unsigned integer
|
|
|
|
|
* 3. raw format
|
|
|
|
|
*/
|
|
|
|
|
if (H5Tget_size(type)<=sizeof(long_long)) {
|
|
|
|
|
dst_size = sizeof(long_long);
|
|
|
|
|
if (H5T_SGN_NONE==H5Tget_sign(type)) {
|
|
|
|
|
native = H5T_NATIVE_ULLONG;
|
|
|
|
|
} else {
|
|
|
|
|
native = H5T_NATIVE_LLONG;
|
|
|
|
|
}
|
|
|
|
|
} else {
|
|
|
|
|
dst_size = H5Tget_size(type);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Get the names and raw values of all members */
|
|
|
|
|
name = calloc(nmembs, sizeof(char*));
|
|
|
|
|
value = calloc(nmembs, MAX(H5Tget_size(type), dst_size));
|
|
|
|
|
for (i=0; i<nmembs; i++) {
|
|
|
|
|
name[i] = H5Tget_member_name(type, i);
|
|
|
|
|
H5Tget_member_value(type, i, value+i*H5Tget_size(type));
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Convert values to native data type */
|
1999-07-03 18:34:45 +08:00
|
|
|
|
if (native>0) H5Tconvert(super, native, nmembs, value, NULL, H5P_DEFAULT);
|
1999-01-07 19:42:27 +08:00
|
|
|
|
|
|
|
|
|
/* Sort members by increasing value */
|
|
|
|
|
/*not implemented yet*/
|
|
|
|
|
|
|
|
|
|
/* Print members */
|
|
|
|
|
for (i=0; i<nmembs; i++) {
|
1999-06-15 22:58:25 +08:00
|
|
|
|
printf("\n%*s", ind+4, "");
|
1999-01-22 02:33:39 +08:00
|
|
|
|
nchars = display_string(stdout, name[i], TRUE);
|
1999-01-07 19:42:27 +08:00
|
|
|
|
printf("%*s = ", MAX(0, 16-nchars), "");
|
|
|
|
|
|
|
|
|
|
if (native<0) {
|
|
|
|
|
printf("0x");
|
|
|
|
|
for (j=0; j<dst_size; j++) {
|
|
|
|
|
printf("%02x", value[i*dst_size+j]);
|
|
|
|
|
}
|
|
|
|
|
} else if (H5T_SGN_NONE==H5Tget_sign(native)) {
|
|
|
|
|
printf("%"PRINTF_LL_WIDTH"u",
|
1999-09-30 00:54:00 +08:00
|
|
|
|
*((unsigned long_long*)((void*)(value+i*dst_size))));
|
1999-01-07 19:42:27 +08:00
|
|
|
|
} else {
|
|
|
|
|
printf("%"PRINTF_LL_WIDTH"d",
|
1999-09-30 00:54:00 +08:00
|
|
|
|
*((long_long*)((void*)(value+i*dst_size))));
|
1999-01-07 19:42:27 +08:00
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Release resources */
|
|
|
|
|
for (i=0; i<nmembs; i++) free(name[i]);
|
|
|
|
|
free(name);
|
|
|
|
|
free(value);
|
|
|
|
|
H5Tclose(super);
|
|
|
|
|
|
1999-06-15 22:58:25 +08:00
|
|
|
|
if (0==nmembs) printf("\n%*s <empty>", ind+4, "");
|
|
|
|
|
printf("\n%*s}", ind, "");
|
1999-01-07 19:42:27 +08:00
|
|
|
|
return TRUE;
|
|
|
|
|
}
|
|
|
|
|
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
|
|
|
|
|
/*-------------------------------------------------------------------------
|
1999-01-22 02:33:39 +08:00
|
|
|
|
* Function: display_string_type
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
*
|
|
|
|
|
* Purpose: Print information about a string data type.
|
|
|
|
|
*
|
|
|
|
|
* Return: Success: TRUE
|
|
|
|
|
*
|
|
|
|
|
* Failure: FALSE, nothing printed
|
|
|
|
|
*
|
|
|
|
|
* Programmer: Robb Matzke
|
|
|
|
|
* Thursday, November 5, 1998
|
|
|
|
|
*
|
|
|
|
|
* Modifications:
|
|
|
|
|
*
|
|
|
|
|
*-------------------------------------------------------------------------
|
|
|
|
|
*/
|
|
|
|
|
static hbool_t
|
1999-06-15 22:58:25 +08:00
|
|
|
|
display_string_type(hid_t type, int UNUSED ind)
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
{
|
|
|
|
|
H5T_str_t pad;
|
|
|
|
|
const char *pad_s=NULL;
|
|
|
|
|
H5T_cset_t cset;
|
|
|
|
|
const char *cset_s=NULL;
|
|
|
|
|
|
|
|
|
|
if (H5T_STRING!=H5Tget_class(type)) return FALSE;
|
|
|
|
|
|
|
|
|
|
/* Padding */
|
|
|
|
|
pad = H5Tget_strpad(type);
|
|
|
|
|
switch (pad) {
|
|
|
|
|
case H5T_STR_NULLTERM:
|
|
|
|
|
pad_s = "null-terminated";
|
|
|
|
|
break;
|
|
|
|
|
case H5T_STR_NULLPAD:
|
|
|
|
|
pad_s = "null-padded";
|
|
|
|
|
break;
|
|
|
|
|
case H5T_STR_SPACEPAD:
|
|
|
|
|
pad_s = "space-padded";
|
|
|
|
|
break;
|
|
|
|
|
case H5T_STR_RESERVED_3:
|
|
|
|
|
case H5T_STR_RESERVED_4:
|
|
|
|
|
case H5T_STR_RESERVED_5:
|
|
|
|
|
case H5T_STR_RESERVED_6:
|
|
|
|
|
case H5T_STR_RESERVED_7:
|
|
|
|
|
case H5T_STR_RESERVED_8:
|
|
|
|
|
case H5T_STR_RESERVED_9:
|
|
|
|
|
case H5T_STR_RESERVED_10:
|
|
|
|
|
case H5T_STR_RESERVED_11:
|
|
|
|
|
case H5T_STR_RESERVED_12:
|
|
|
|
|
case H5T_STR_RESERVED_13:
|
|
|
|
|
case H5T_STR_RESERVED_14:
|
|
|
|
|
case H5T_STR_RESERVED_15:
|
|
|
|
|
case H5T_STR_ERROR:
|
|
|
|
|
pad_s = "unknown-format";
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Character set */
|
|
|
|
|
cset = H5Tget_cset(type);
|
|
|
|
|
switch (cset) {
|
|
|
|
|
case H5T_CSET_ASCII:
|
|
|
|
|
cset_s = "ASCII";
|
|
|
|
|
break;
|
|
|
|
|
case H5T_CSET_RESERVED_1:
|
|
|
|
|
case H5T_CSET_RESERVED_2:
|
|
|
|
|
case H5T_CSET_RESERVED_3:
|
|
|
|
|
case H5T_CSET_RESERVED_4:
|
|
|
|
|
case H5T_CSET_RESERVED_5:
|
|
|
|
|
case H5T_CSET_RESERVED_6:
|
|
|
|
|
case H5T_CSET_RESERVED_7:
|
|
|
|
|
case H5T_CSET_RESERVED_8:
|
|
|
|
|
case H5T_CSET_RESERVED_9:
|
|
|
|
|
case H5T_CSET_RESERVED_10:
|
|
|
|
|
case H5T_CSET_RESERVED_11:
|
|
|
|
|
case H5T_CSET_RESERVED_12:
|
|
|
|
|
case H5T_CSET_RESERVED_13:
|
|
|
|
|
case H5T_CSET_RESERVED_14:
|
|
|
|
|
case H5T_CSET_RESERVED_15:
|
|
|
|
|
case H5T_CSET_ERROR:
|
|
|
|
|
cset_s = "unknown-character-set";
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
printf("%lu-byte %s %s string",
|
|
|
|
|
(unsigned long)H5Tget_size(type), pad_s, cset_s);
|
|
|
|
|
return TRUE;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
/*-------------------------------------------------------------------------
|
|
|
|
|
* Function: display_reference_type
|
|
|
|
|
*
|
|
|
|
|
* Purpose: Prints information about a reference data type.
|
|
|
|
|
*
|
|
|
|
|
* Return: Success: TRUE
|
|
|
|
|
*
|
|
|
|
|
* Failure: FALSE, nothing printed
|
|
|
|
|
*
|
|
|
|
|
* Programmer: Robb Matzke
|
|
|
|
|
* Thursday, November 5, 1998
|
|
|
|
|
*
|
|
|
|
|
* Modifications:
|
1999-06-07 23:05:02 +08:00
|
|
|
|
* Robb Matzke, 1999-06-04
|
|
|
|
|
* Knows about object and dataset region references.
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
*
|
|
|
|
|
*-------------------------------------------------------------------------
|
|
|
|
|
*/
|
|
|
|
|
static hbool_t
|
1999-06-15 22:58:25 +08:00
|
|
|
|
display_reference_type(hid_t type, int UNUSED ind)
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
{
|
|
|
|
|
if (H5T_REFERENCE!=H5Tget_class(type)) return FALSE;
|
|
|
|
|
|
1999-06-07 23:05:02 +08:00
|
|
|
|
if (H5Tequal(type, H5T_STD_REF_OBJ)) {
|
|
|
|
|
printf("object reference");
|
|
|
|
|
} else if (H5Tequal(type, H5T_STD_REF_DSETREG)) {
|
|
|
|
|
printf("dataset region reference");
|
|
|
|
|
} else {
|
|
|
|
|
printf("%lu-byte unknown reference",
|
|
|
|
|
(unsigned long)H5Tget_size(type));
|
|
|
|
|
}
|
|
|
|
|
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
return TRUE;
|
|
|
|
|
}
|
|
|
|
|
|
1999-06-08 04:20:32 +08:00
|
|
|
|
|
|
|
|
|
/*-------------------------------------------------------------------------
|
|
|
|
|
* Function: display_opaque_type
|
|
|
|
|
*
|
|
|
|
|
* Purpose: Prints information about an opaque data type.
|
|
|
|
|
*
|
|
|
|
|
* Return: Success: TRUE
|
|
|
|
|
*
|
|
|
|
|
* Failure: FALSE, nothing printed
|
|
|
|
|
*
|
|
|
|
|
* Programmer: Robb Matzke
|
|
|
|
|
* Monday, June 7, 1999
|
|
|
|
|
*
|
|
|
|
|
* Modifications:
|
|
|
|
|
*
|
|
|
|
|
*-------------------------------------------------------------------------
|
|
|
|
|
*/
|
|
|
|
|
static hbool_t
|
1999-06-15 22:58:25 +08:00
|
|
|
|
display_opaque_type(hid_t type, int ind)
|
1999-06-08 04:20:32 +08:00
|
|
|
|
{
|
|
|
|
|
char *tag;
|
|
|
|
|
size_t size;
|
|
|
|
|
|
|
|
|
|
if (H5T_OPAQUE!=H5Tget_class(type)) return FALSE;
|
|
|
|
|
|
|
|
|
|
size = H5Tget_size(type);
|
|
|
|
|
printf("%lu-byte opaque type", (unsigned long)size);
|
|
|
|
|
if ((tag=H5Tget_tag(type))) {
|
1999-06-15 22:58:25 +08:00
|
|
|
|
printf("\n%*s(tag = \"", ind, "");
|
1999-06-08 04:20:32 +08:00
|
|
|
|
display_string(stdout, tag, FALSE);
|
|
|
|
|
printf("\")");
|
|
|
|
|
free(tag);
|
|
|
|
|
}
|
|
|
|
|
return TRUE;
|
|
|
|
|
}
|
|
|
|
|
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
|
|
|
|
|
/*-------------------------------------------------------------------------
|
|
|
|
|
* Function: display_type
|
|
|
|
|
*
|
|
|
|
|
* Purpose: Prints a data type definition. The definition is printed
|
|
|
|
|
* without any leading space or trailing line-feed (although
|
|
|
|
|
* there might be line-feeds inside the type definition). The
|
1999-06-15 22:58:25 +08:00
|
|
|
|
* first line is assumed to have IND characters before it on
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
* the same line (printed by the caller).
|
|
|
|
|
*
|
|
|
|
|
* Return: void
|
|
|
|
|
*
|
|
|
|
|
* Programmer: Robb Matzke
|
|
|
|
|
* Thursday, November 5, 1998
|
|
|
|
|
*
|
|
|
|
|
* Modifications:
|
1999-06-14 23:07:58 +08:00
|
|
|
|
* Robb Matzke, 1999-06-11
|
|
|
|
|
* Prints the OID of shared data types.
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
*
|
|
|
|
|
*-------------------------------------------------------------------------
|
|
|
|
|
*/
|
|
|
|
|
static void
|
1999-06-15 22:58:25 +08:00
|
|
|
|
display_type(hid_t type, int ind)
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
{
|
1999-02-16 01:38:04 +08:00
|
|
|
|
H5T_class_t data_class = H5Tget_class(type);
|
1999-06-14 23:07:58 +08:00
|
|
|
|
H5G_stat_t sb;
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
|
|
|
|
|
/* Bad data type */
|
|
|
|
|
if (type<0) {
|
|
|
|
|
printf("<ERROR>");
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
|
1999-06-14 23:07:58 +08:00
|
|
|
|
/* Shared? If so then print the type's OID */
|
|
|
|
|
if (H5Tcommitted(type)) {
|
|
|
|
|
if (H5Gget_objinfo(type, ".", FALSE, &sb)>=0) {
|
|
|
|
|
printf("shared-%lu:%lu:%lu:%lu ",
|
|
|
|
|
sb.fileno[1], sb.fileno[0],
|
|
|
|
|
sb.objno[1], sb.objno[0]);
|
|
|
|
|
} else {
|
|
|
|
|
printf("shared ");
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
/* Print the type */
|
1999-06-15 22:58:25 +08:00
|
|
|
|
if (display_native_type(type, ind) ||
|
|
|
|
|
display_ieee_type(type, ind) ||
|
|
|
|
|
display_int_type(type, ind) ||
|
|
|
|
|
display_float_type(type, ind) ||
|
|
|
|
|
display_cmpd_type(type, ind) ||
|
|
|
|
|
display_enum_type(type, ind) ||
|
|
|
|
|
display_string_type(type, ind) ||
|
|
|
|
|
display_reference_type(type, ind) ||
|
|
|
|
|
display_opaque_type(type, ind)) {
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Unknown type */
|
|
|
|
|
printf("%lu-byte class-%u unknown",
|
|
|
|
|
(unsigned long)H5Tget_size(type),
|
|
|
|
|
(unsigned)data_class);
|
|
|
|
|
}
|
|
|
|
|
|
[svn-r526] Changes since 19980721
----------------------
./tools/h5ls.c
If the dataset is of type H5T_NATIVE_CHAR then we print the
value as a string. This is temporary -- I plan to add better
control of this later but needed something now for debugging.
./src/H5Fistore.c
Squashed a bug in the chunk caching code that caused the wrong
chunk to be returned.
./bin/trace
./src/H5.c
Added support for printing values of array arguments when the
size of the array is supplied by some previous argument. You
must declare the argument as an array in order for the
automatic tracing stuff to work. For instance, the third
argument of H5Pset_chunk() is an array whose size is
determined by the second argument `ndims'. Here's how you
should declare it:
herr_t
H5Pset_chunk(hid_t plist_id, intn rank, hsize_t dims[/*rank*/])
The comment inside the `[]' is the name of some previous
integer argument (int, unsigned, size_t, ssize_t, hsize_t,
hssize_t). The trace output will look something like:
H5Pset_chunk(plist=1234567, rank=2, dims=0x112233 {11, 22})
Changed tracing so that data types are printed out
symbolically when possible.
Changed tracing so data type initializations are not printed.
This used to be confusing because lots of H5Tcopy() and
H5Tregister...() calls showed up before the applications first
explicit API call.
./src/H5Ipublic.h
Changed the file atom group from zero to one so printing of
atoms during tracing is more consistent -- they're all big
numbers now.
./src/H5A.c
./src/H5E.c
./src/H5F.c
./src/H5G.c
./src/H5Sselect.c
./src/H5T.c
./src/H5TB.c
./src/H5Z.c
Accidently modified these when working on the tracing, but
nothing should have changed.
./src/H5P.c
Changed the definition of H5Pset_chunk() for tracing.
./src/H5S.c
./src/H5Spublic.h
Changed the definitions of H5Sset_extent_simple() and
H5Screate_simple() for tracing. Changed the FUNC_ENTER() name
for H5Screate_simple() so tracing shows the correct name.
1998-07-22 21:51:51 +08:00
|
|
|
|
|
|
|
|
|
/*-------------------------------------------------------------------------
|
|
|
|
|
* Function: dump_dataset_values
|
|
|
|
|
*
|
|
|
|
|
* Purpose: Prints all values of a dataset.
|
|
|
|
|
*
|
|
|
|
|
* Return: void
|
|
|
|
|
*
|
|
|
|
|
* Programmer: Robb Matzke
|
|
|
|
|
* Tuesday, July 21, 1998
|
|
|
|
|
*
|
|
|
|
|
* Modifications:
|
1999-10-01 00:14:14 +08:00
|
|
|
|
* Robb Matzke, 1999-09-27
|
|
|
|
|
* Understands the simple_output_g switch which causes data to
|
|
|
|
|
* be displayed in a more machine-readable format.
|
[svn-r526] Changes since 19980721
----------------------
./tools/h5ls.c
If the dataset is of type H5T_NATIVE_CHAR then we print the
value as a string. This is temporary -- I plan to add better
control of this later but needed something now for debugging.
./src/H5Fistore.c
Squashed a bug in the chunk caching code that caused the wrong
chunk to be returned.
./bin/trace
./src/H5.c
Added support for printing values of array arguments when the
size of the array is supplied by some previous argument. You
must declare the argument as an array in order for the
automatic tracing stuff to work. For instance, the third
argument of H5Pset_chunk() is an array whose size is
determined by the second argument `ndims'. Here's how you
should declare it:
herr_t
H5Pset_chunk(hid_t plist_id, intn rank, hsize_t dims[/*rank*/])
The comment inside the `[]' is the name of some previous
integer argument (int, unsigned, size_t, ssize_t, hsize_t,
hssize_t). The trace output will look something like:
H5Pset_chunk(plist=1234567, rank=2, dims=0x112233 {11, 22})
Changed tracing so that data types are printed out
symbolically when possible.
Changed tracing so data type initializations are not printed.
This used to be confusing because lots of H5Tcopy() and
H5Tregister...() calls showed up before the applications first
explicit API call.
./src/H5Ipublic.h
Changed the file atom group from zero to one so printing of
atoms during tracing is more consistent -- they're all big
numbers now.
./src/H5A.c
./src/H5E.c
./src/H5F.c
./src/H5G.c
./src/H5Sselect.c
./src/H5T.c
./src/H5TB.c
./src/H5Z.c
Accidently modified these when working on the tracing, but
nothing should have changed.
./src/H5P.c
Changed the definition of H5Pset_chunk() for tracing.
./src/H5S.c
./src/H5Spublic.h
Changed the definitions of H5Sset_extent_simple() and
H5Screate_simple() for tracing. Changed the FUNC_ENTER() name
for H5Screate_simple() so tracing shows the correct name.
1998-07-22 21:51:51 +08:00
|
|
|
|
*-------------------------------------------------------------------------
|
|
|
|
|
*/
|
|
|
|
|
static void
|
|
|
|
|
dump_dataset_values(hid_t dset)
|
|
|
|
|
{
|
1998-07-24 22:28:32 +08:00
|
|
|
|
hid_t f_type = H5Dget_type(dset);
|
|
|
|
|
size_t size = H5Tget_size(f_type);
|
[svn-r537] Changes since 19980722
----------------------
./src/H5A.c
./src/H5Apublic.h
./test/tattr.c
Switched the order of the second and third argument of
H5Aget_name() to make it consistent with other functions that
take buffers and buffer sizes.
./src/H5G.c
./src/H5Gpublic.h
./src/H5Gprivate.h
The H5Gget_comment() function returns the size of the comment
including the null terminator. If the object has no comment
then zero is returned. If an error occurs then a negative
value is returned.
./MANIFEST
./tools/Makefile.in
./tools/h5tools.h [NEW]
./tools/h5dump.c [NEW]
Created a library for printing values of datasets in a way
that looks nice. It's not done yet, but I needed it for
debugging the contents of files from Jim Reus.
./tools/h5ls.c
Added the `-d' and `--dump' options which cause the contents
of a dataset to be printed. Added `-w N' and `--width=N'
options to control how wide the raw data output should be. If
you want single-column output then say `-w1'.
Printing dataset values can now handle datasets of any integer
or floating point atomic type. As a special case, integers
which are one byte wide are treated a character strings for
now.
Sample output:
$ h5ls --dump --width=60 banana.hdf
ARCHIVE 0:0:0:744 Dataset {52/Inf}
Data:
(0) "U struct complex { double R; double I; };\012V"
(43) " double;\012"
U 0:0:0:2500 Dataset {256/512}
Data: printing of compound data types is not implemented yet
V 0:0:0:3928 Dataset {256/512}
Data:
(0) 0, 0.015625, 0.03125, 0.046875, 0.0625,
(5) 0.078125, 0.09375, 0.109375, 0.125, 0.140625,
(10) 0.15625, 0.171875, 0.1875, 0.203125, 0.21875,
(15) 0.234375, 0.25, 0.265625, 0.28125, 0.296875,
...
1998-07-24 05:19:17 +08:00
|
|
|
|
h5dump_t info;
|
1999-10-01 00:14:14 +08:00
|
|
|
|
char string_prefix[64];
|
|
|
|
|
|
[svn-r537] Changes since 19980722
----------------------
./src/H5A.c
./src/H5Apublic.h
./test/tattr.c
Switched the order of the second and third argument of
H5Aget_name() to make it consistent with other functions that
take buffers and buffer sizes.
./src/H5G.c
./src/H5Gpublic.h
./src/H5Gprivate.h
The H5Gget_comment() function returns the size of the comment
including the null terminator. If the object has no comment
then zero is returned. If an error occurs then a negative
value is returned.
./MANIFEST
./tools/Makefile.in
./tools/h5tools.h [NEW]
./tools/h5dump.c [NEW]
Created a library for printing values of datasets in a way
that looks nice. It's not done yet, but I needed it for
debugging the contents of files from Jim Reus.
./tools/h5ls.c
Added the `-d' and `--dump' options which cause the contents
of a dataset to be printed. Added `-w N' and `--width=N'
options to control how wide the raw data output should be. If
you want single-column output then say `-w1'.
Printing dataset values can now handle datasets of any integer
or floating point atomic type. As a special case, integers
which are one byte wide are treated a character strings for
now.
Sample output:
$ h5ls --dump --width=60 banana.hdf
ARCHIVE 0:0:0:744 Dataset {52/Inf}
Data:
(0) "U struct complex { double R; double I; };\012V"
(43) " double;\012"
U 0:0:0:2500 Dataset {256/512}
Data: printing of compound data types is not implemented yet
V 0:0:0:3928 Dataset {256/512}
Data:
(0) 0, 0.015625, 0.03125, 0.046875, 0.0625,
(5) 0.078125, 0.09375, 0.109375, 0.125, 0.140625,
(10) 0.15625, 0.171875, 0.1875, 0.203125, 0.21875,
(15) 0.234375, 0.25, 0.265625, 0.28125, 0.296875,
...
1998-07-24 05:19:17 +08:00
|
|
|
|
/* Set to all default values and then override */
|
|
|
|
|
memset(&info, 0, sizeof info);
|
[svn-r526] Changes since 19980721
----------------------
./tools/h5ls.c
If the dataset is of type H5T_NATIVE_CHAR then we print the
value as a string. This is temporary -- I plan to add better
control of this later but needed something now for debugging.
./src/H5Fistore.c
Squashed a bug in the chunk caching code that caused the wrong
chunk to be returned.
./bin/trace
./src/H5.c
Added support for printing values of array arguments when the
size of the array is supplied by some previous argument. You
must declare the argument as an array in order for the
automatic tracing stuff to work. For instance, the third
argument of H5Pset_chunk() is an array whose size is
determined by the second argument `ndims'. Here's how you
should declare it:
herr_t
H5Pset_chunk(hid_t plist_id, intn rank, hsize_t dims[/*rank*/])
The comment inside the `[]' is the name of some previous
integer argument (int, unsigned, size_t, ssize_t, hsize_t,
hssize_t). The trace output will look something like:
H5Pset_chunk(plist=1234567, rank=2, dims=0x112233 {11, 22})
Changed tracing so that data types are printed out
symbolically when possible.
Changed tracing so data type initializations are not printed.
This used to be confusing because lots of H5Tcopy() and
H5Tregister...() calls showed up before the applications first
explicit API call.
./src/H5Ipublic.h
Changed the file atom group from zero to one so printing of
atoms during tracing is more consistent -- they're all big
numbers now.
./src/H5A.c
./src/H5E.c
./src/H5F.c
./src/H5G.c
./src/H5Sselect.c
./src/H5T.c
./src/H5TB.c
./src/H5Z.c
Accidently modified these when working on the tracing, but
nothing should have changed.
./src/H5P.c
Changed the definition of H5Pset_chunk() for tracing.
./src/H5S.c
./src/H5Spublic.h
Changed the definitions of H5Sset_extent_simple() and
H5Screate_simple() for tracing. Changed the FUNC_ENTER() name
for H5Screate_simple() so tracing shows the correct name.
1998-07-22 21:51:51 +08:00
|
|
|
|
|
1999-10-01 00:14:14 +08:00
|
|
|
|
if (simple_output_g) {
|
|
|
|
|
info.idx_fmt = "";
|
|
|
|
|
info.line_ncols = 65535; /*something big*/
|
|
|
|
|
info.line_per_line = 1;
|
|
|
|
|
info.line_multi_new = 0;
|
|
|
|
|
info.line_pre = " ";
|
|
|
|
|
info.line_cont = " ";
|
|
|
|
|
|
|
|
|
|
info.arr_pre = "";
|
|
|
|
|
info.arr_suf = "";
|
|
|
|
|
info.arr_sep = " ";
|
|
|
|
|
|
|
|
|
|
info.cmpd_pre = "";
|
|
|
|
|
info.cmpd_suf = "";
|
|
|
|
|
info.cmpd_sep = " ";
|
2000-02-18 02:35:30 +08:00
|
|
|
|
|
1999-10-01 00:14:14 +08:00
|
|
|
|
if (label_g) info.cmpd_name = "%s=";
|
|
|
|
|
|
|
|
|
|
info.elmt_suf1 = " ";
|
|
|
|
|
info.str_locale = ESCAPE_HTML;
|
|
|
|
|
|
|
|
|
|
} else {
|
|
|
|
|
info.idx_fmt = "(%s)";
|
|
|
|
|
info.line_ncols = width_g;
|
|
|
|
|
info.line_multi_new = 1;
|
|
|
|
|
if (label_g) info.cmpd_name = "%s=";
|
|
|
|
|
info.line_pre = " %s ";
|
|
|
|
|
info.line_cont = " %s ";
|
|
|
|
|
info.str_repeat = 8;
|
|
|
|
|
}
|
2000-02-24 01:58:43 +08:00
|
|
|
|
|
|
|
|
|
info.dset_format = "DSET-%lu:%lu:%lu:%lu-";
|
|
|
|
|
info.dset_hidefileno = 0;
|
|
|
|
|
|
|
|
|
|
info.obj_format = "-%lu:%lu:%lu:%lu";
|
|
|
|
|
info.obj_hidefileno = 0;
|
|
|
|
|
|
|
|
|
|
info.dset_blockformat_pre = "%sBlk%lu: ";
|
|
|
|
|
info.dset_ptformat_pre = "%sPt%lu: ";
|
|
|
|
|
|
|
|
|
|
info.line_indent = "";
|
1999-10-01 00:14:14 +08:00
|
|
|
|
|
1999-06-07 23:58:05 +08:00
|
|
|
|
if (hexdump_g) {
|
|
|
|
|
/*
|
|
|
|
|
* Print all data in hexadecimal format if the `-x' or `--hexdump'
|
|
|
|
|
* command line switch was given.
|
|
|
|
|
*/
|
|
|
|
|
info.raw = TRUE;
|
|
|
|
|
} else if (string_g && 1==size && H5T_INTEGER==H5Tget_class(f_type)) {
|
|
|
|
|
/*
|
|
|
|
|
* Print 1-byte integer data as an ASCI character string instead of
|
|
|
|
|
* integers if the `-s' or `--string' command-line option was given.
|
|
|
|
|
*/
|
[svn-r876] Changes since 19981102
----------------------
./bin/snapshot
Made same fix as for the release script yesterday.
./src/H5D.c
./src/H5Dprivate.h
./src/H5G.c
./src/H5Gprivate.h
./src/H5Gpublic.h
./src/H5O.c
./src/H5Oprivate.h
./src/H5RA.c
./src/H5RAprivate.h
./src/H5T.c
./src/H5Tprivate.h
Improved object type checking. Instead of determining the
object type by trying to open each of the possible types, we
keep a table of associations between object type number (like
H5G_GROUP, H5G_DATASET, H5D_TYPE, and H5D_RAGGED) and an `isa'
function that returns true if the object header has the right
messages to make the object a particular type. This mechanism
also allows specialization of object types by permitting an
object to satisfy more than one `isa' function.
Added `isa' functions for groups, datasets, ragged arrays, and
committed data types.
./src/H5config.h.in
Added HAVE_STAT_ST_BLOCKS. I thought this had already been
added, but apparently not.
./tools/h5ls.c
Removed system include files since they're already included by
H5private.h and since I wasn't including them portably anyway.
By default, 1-byte integer types are printed as integer values
instead of ASCII characters. However, the `-s' or `--string'
command-line switch causes the data to be interpretted as
ASCII. String data types are always printed as character
data.
Ragged arrays are now identified as ragged arrays and h5ls
doesn't descend into the group automatically. This uses the
new object type specialization stuff.
./tools/h5tools.c
./tools/h5tools.h
Added the ability to print 1-byte integer types as either
ASCII or numeric data instead of always ASCII. The default is
to print as numeric data.
1998-11-06 04:28:34 +08:00
|
|
|
|
info.ascii = TRUE;
|
1998-07-24 22:28:32 +08:00
|
|
|
|
info.elmt_suf1 = "";
|
|
|
|
|
info.elmt_suf2 = "";
|
1999-10-01 00:14:14 +08:00
|
|
|
|
strcpy(string_prefix, info.line_pre);
|
|
|
|
|
strcat(string_prefix, "\"");
|
|
|
|
|
info.line_pre = string_prefix;
|
1998-07-24 22:28:32 +08:00
|
|
|
|
info.line_suf = "\"";
|
[svn-r537] Changes since 19980722
----------------------
./src/H5A.c
./src/H5Apublic.h
./test/tattr.c
Switched the order of the second and third argument of
H5Aget_name() to make it consistent with other functions that
take buffers and buffer sizes.
./src/H5G.c
./src/H5Gpublic.h
./src/H5Gprivate.h
The H5Gget_comment() function returns the size of the comment
including the null terminator. If the object has no comment
then zero is returned. If an error occurs then a negative
value is returned.
./MANIFEST
./tools/Makefile.in
./tools/h5tools.h [NEW]
./tools/h5dump.c [NEW]
Created a library for printing values of datasets in a way
that looks nice. It's not done yet, but I needed it for
debugging the contents of files from Jim Reus.
./tools/h5ls.c
Added the `-d' and `--dump' options which cause the contents
of a dataset to be printed. Added `-w N' and `--width=N'
options to control how wide the raw data output should be. If
you want single-column output then say `-w1'.
Printing dataset values can now handle datasets of any integer
or floating point atomic type. As a special case, integers
which are one byte wide are treated a character strings for
now.
Sample output:
$ h5ls --dump --width=60 banana.hdf
ARCHIVE 0:0:0:744 Dataset {52/Inf}
Data:
(0) "U struct complex { double R; double I; };\012V"
(43) " double;\012"
U 0:0:0:2500 Dataset {256/512}
Data: printing of compound data types is not implemented yet
V 0:0:0:3928 Dataset {256/512}
Data:
(0) 0, 0.015625, 0.03125, 0.046875, 0.0625,
(5) 0.078125, 0.09375, 0.109375, 0.125, 0.140625,
(10) 0.15625, 0.171875, 0.1875, 0.203125, 0.21875,
(15) 0.234375, 0.25, 0.265625, 0.28125, 0.296875,
...
1998-07-24 05:19:17 +08:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* Print all the values.
|
|
|
|
|
*/
|
|
|
|
|
printf(" Data:\n");
|
2000-02-15 23:46:26 +08:00
|
|
|
|
if (h5dump_dset(stdout, &info, dset, -1, -1)<0) {
|
1998-07-24 22:28:32 +08:00
|
|
|
|
printf(" Unable to print data.\n");
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
H5Tclose(f_type);
|
[svn-r526] Changes since 19980721
----------------------
./tools/h5ls.c
If the dataset is of type H5T_NATIVE_CHAR then we print the
value as a string. This is temporary -- I plan to add better
control of this later but needed something now for debugging.
./src/H5Fistore.c
Squashed a bug in the chunk caching code that caused the wrong
chunk to be returned.
./bin/trace
./src/H5.c
Added support for printing values of array arguments when the
size of the array is supplied by some previous argument. You
must declare the argument as an array in order for the
automatic tracing stuff to work. For instance, the third
argument of H5Pset_chunk() is an array whose size is
determined by the second argument `ndims'. Here's how you
should declare it:
herr_t
H5Pset_chunk(hid_t plist_id, intn rank, hsize_t dims[/*rank*/])
The comment inside the `[]' is the name of some previous
integer argument (int, unsigned, size_t, ssize_t, hsize_t,
hssize_t). The trace output will look something like:
H5Pset_chunk(plist=1234567, rank=2, dims=0x112233 {11, 22})
Changed tracing so that data types are printed out
symbolically when possible.
Changed tracing so data type initializations are not printed.
This used to be confusing because lots of H5Tcopy() and
H5Tregister...() calls showed up before the applications first
explicit API call.
./src/H5Ipublic.h
Changed the file atom group from zero to one so printing of
atoms during tracing is more consistent -- they're all big
numbers now.
./src/H5A.c
./src/H5E.c
./src/H5F.c
./src/H5G.c
./src/H5Sselect.c
./src/H5T.c
./src/H5TB.c
./src/H5Z.c
Accidently modified these when working on the tracing, but
nothing should have changed.
./src/H5P.c
Changed the definition of H5Pset_chunk() for tracing.
./src/H5S.c
./src/H5Spublic.h
Changed the definitions of H5Sset_extent_simple() and
H5Screate_simple() for tracing. Changed the FUNC_ENTER() name
for H5Screate_simple() so tracing shows the correct name.
1998-07-22 21:51:51 +08:00
|
|
|
|
}
|
|
|
|
|
|
1998-06-06 05:03:49 +08:00
|
|
|
|
|
|
|
|
|
/*-------------------------------------------------------------------------
|
|
|
|
|
* Function: list_attr
|
|
|
|
|
*
|
|
|
|
|
* Purpose: Prints information about attributes.
|
|
|
|
|
*
|
|
|
|
|
* Return: Success: 0
|
|
|
|
|
*
|
|
|
|
|
* Failure: -1
|
|
|
|
|
*
|
|
|
|
|
* Programmer: Robb Matzke
|
|
|
|
|
* Friday, June 5, 1998
|
|
|
|
|
*
|
|
|
|
|
* Modifications:
|
|
|
|
|
*
|
|
|
|
|
*-------------------------------------------------------------------------
|
|
|
|
|
*/
|
|
|
|
|
static herr_t
|
1999-04-16 03:57:50 +08:00
|
|
|
|
list_attr (hid_t obj, const char *attr_name, void UNUSED *op_data)
|
1998-06-06 05:03:49 +08:00
|
|
|
|
{
|
1999-01-22 02:33:39 +08:00
|
|
|
|
hid_t attr, space, type, p_type;
|
|
|
|
|
hsize_t size[64], nelmts=1;
|
|
|
|
|
int ndims, i, n;
|
|
|
|
|
size_t need;
|
|
|
|
|
void *buf;
|
|
|
|
|
h5dump_t info;
|
|
|
|
|
|
|
|
|
|
printf(" Attribute: ");
|
|
|
|
|
n = display_string(stdout, attr_name, TRUE);
|
1999-03-30 19:38:34 +08:00
|
|
|
|
printf("%*s", MAX(0, 9-n), "");
|
1999-01-22 02:33:39 +08:00
|
|
|
|
if ((attr = H5Aopen_name(obj, attr_name))) {
|
|
|
|
|
space = H5Aget_space(attr);
|
|
|
|
|
type = H5Aget_type(attr);
|
|
|
|
|
|
|
|
|
|
/* Data space */
|
|
|
|
|
ndims = H5Sget_simple_extent_dims(space, size, NULL);
|
1999-03-30 19:38:34 +08:00
|
|
|
|
if (0==ndims) {
|
|
|
|
|
puts(" scalar");
|
|
|
|
|
} else {
|
|
|
|
|
printf(" {");
|
|
|
|
|
for (i=0; i<ndims; i++) {
|
|
|
|
|
HDfprintf(stdout, "%s%Hu", i?", ":"", size[i]);
|
|
|
|
|
nelmts *= size[i];
|
|
|
|
|
}
|
|
|
|
|
puts("}");
|
1999-01-22 02:33:39 +08:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Data type */
|
1999-03-30 19:38:34 +08:00
|
|
|
|
printf(" Type: ");
|
1999-01-22 02:33:39 +08:00
|
|
|
|
display_type(type, 15);
|
|
|
|
|
putchar('\n');
|
|
|
|
|
|
|
|
|
|
/* Data */
|
|
|
|
|
memset(&info, 0, sizeof info);
|
1999-04-27 22:47:54 +08:00
|
|
|
|
info.line_multi_new = 1;
|
1999-03-30 19:38:34 +08:00
|
|
|
|
if (nelmts<5) {
|
1999-04-27 22:47:54 +08:00
|
|
|
|
info.idx_fmt = "";
|
|
|
|
|
info.line_1st = " Data: ";
|
|
|
|
|
info.line_pre = " ";
|
|
|
|
|
info.line_cont = " ";
|
1999-10-01 00:14:14 +08:00
|
|
|
|
info.str_repeat = 8;
|
1999-04-27 22:47:54 +08:00
|
|
|
|
|
1999-03-30 19:38:34 +08:00
|
|
|
|
} else {
|
1999-08-26 22:21:45 +08:00
|
|
|
|
printf(" Data:\n");
|
1999-04-27 22:47:54 +08:00
|
|
|
|
info.idx_fmt = "(%s)";
|
|
|
|
|
info.line_pre = " %s ";
|
|
|
|
|
info.line_cont = " %s ";
|
1999-10-01 00:14:14 +08:00
|
|
|
|
info.str_repeat = 8;
|
1999-03-30 19:38:34 +08:00
|
|
|
|
}
|
1999-01-22 02:33:39 +08:00
|
|
|
|
info.line_ncols = width_g;
|
|
|
|
|
if (label_g) info.cmpd_name = "%s=";
|
|
|
|
|
if (string_g && 1==H5Tget_size(type) &&
|
|
|
|
|
H5T_INTEGER==H5Tget_class(type)) {
|
|
|
|
|
info.ascii = TRUE;
|
|
|
|
|
info.elmt_suf1 = "";
|
|
|
|
|
info.elmt_suf2 = "";
|
1999-04-27 22:47:54 +08:00
|
|
|
|
info.idx_fmt = "(%s)";
|
|
|
|
|
info.line_pre = " %s \"";
|
1999-01-22 02:33:39 +08:00
|
|
|
|
info.line_suf = "\"";
|
|
|
|
|
}
|
1999-06-07 23:58:05 +08:00
|
|
|
|
if (hexdump_g) {
|
|
|
|
|
p_type = H5Tcopy(type);
|
|
|
|
|
} else {
|
|
|
|
|
p_type = h5dump_fixtype(type);
|
|
|
|
|
}
|
|
|
|
|
if (p_type>=0) {
|
1999-01-22 02:33:39 +08:00
|
|
|
|
need = nelmts * MAX(H5Tget_size(type), H5Tget_size(p_type));
|
|
|
|
|
buf = malloc(need);
|
|
|
|
|
assert(buf);
|
|
|
|
|
if (H5Aread(attr, p_type, buf)>=0) {
|
2000-02-15 23:46:26 +08:00
|
|
|
|
h5dump_mem(stdout, &info, p_type, space, buf,-1);
|
1999-01-22 02:33:39 +08:00
|
|
|
|
}
|
|
|
|
|
free(buf);
|
|
|
|
|
H5Tclose(p_type);
|
1998-06-06 05:03:49 +08:00
|
|
|
|
}
|
1999-01-22 02:33:39 +08:00
|
|
|
|
|
|
|
|
|
H5Sclose(space);
|
|
|
|
|
H5Tclose(type);
|
|
|
|
|
H5Aclose(attr);
|
|
|
|
|
} else {
|
|
|
|
|
putchar('\n');
|
1998-06-06 05:03:49 +08:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
1998-05-14 01:58:24 +08:00
|
|
|
|
|
1998-08-28 23:24:12 +08:00
|
|
|
|
|
|
|
|
|
/*-------------------------------------------------------------------------
|
|
|
|
|
* Function: dataset_list1
|
|
|
|
|
*
|
|
|
|
|
* Purpose: List information about a dataset which should appear on the
|
|
|
|
|
* same line as the dataset name. This information will precede
|
|
|
|
|
* information which is applicable to all objects which will be
|
|
|
|
|
* printed by the caller.
|
|
|
|
|
*
|
|
|
|
|
* Return: Success: 0
|
|
|
|
|
*
|
|
|
|
|
* Failure: -1
|
|
|
|
|
*
|
|
|
|
|
* Programmer: Robb Matzke
|
|
|
|
|
* Thursday, August 27, 1998
|
|
|
|
|
*
|
|
|
|
|
* Modifications:
|
|
|
|
|
*
|
|
|
|
|
*-------------------------------------------------------------------------
|
|
|
|
|
*/
|
|
|
|
|
static herr_t
|
|
|
|
|
dataset_list1(hid_t dset)
|
|
|
|
|
{
|
|
|
|
|
hsize_t cur_size[64]; /*current dataset dimensions */
|
|
|
|
|
hsize_t max_size[64]; /*maximum dataset dimensions */
|
|
|
|
|
hid_t space; /*data space */
|
|
|
|
|
int ndims; /*dimensionality */
|
|
|
|
|
int i;
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* Information that goes on the same row as the name. The name has
|
|
|
|
|
* already been printed.
|
|
|
|
|
*/
|
|
|
|
|
space = H5Dget_space(dset);
|
1998-09-01 11:35:23 +08:00
|
|
|
|
ndims = H5Sget_simple_extent_dims(space, cur_size, max_size);
|
1998-08-28 23:24:12 +08:00
|
|
|
|
printf (" {");
|
|
|
|
|
for (i=0; i<ndims; i++) {
|
|
|
|
|
HDfprintf (stdout, "%s%Hu", i?", ":"", cur_size[i]);
|
|
|
|
|
if (max_size[i]==H5S_UNLIMITED) {
|
|
|
|
|
HDfprintf (stdout, "/%s", "Inf");
|
|
|
|
|
} else if (max_size[i]!=cur_size[i] || verbose_g>0) {
|
|
|
|
|
HDfprintf(stdout, "/%Hu", max_size[i]);
|
|
|
|
|
}
|
|
|
|
|
}
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
if (0==ndims) printf("SCALAR");
|
1998-08-28 23:24:12 +08:00
|
|
|
|
putchar('}');
|
|
|
|
|
H5Sclose (space);
|
|
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
/*-------------------------------------------------------------------------
|
|
|
|
|
* Function: dataset_list2
|
|
|
|
|
*
|
|
|
|
|
* Purpose: List information about a dataset which should appear after
|
|
|
|
|
* information which is general to all objects.
|
|
|
|
|
*
|
|
|
|
|
* Return: Success: 0
|
|
|
|
|
*
|
|
|
|
|
* Failure: -1
|
|
|
|
|
*
|
|
|
|
|
* Programmer: Robb Matzke
|
|
|
|
|
* Thursday, August 27, 1998
|
|
|
|
|
*
|
|
|
|
|
* Modifications:
|
|
|
|
|
*
|
|
|
|
|
*-------------------------------------------------------------------------
|
|
|
|
|
*/
|
|
|
|
|
static herr_t
|
1999-04-16 03:57:50 +08:00
|
|
|
|
dataset_list2(hid_t dset, const char UNUSED *name)
|
1998-08-28 23:24:12 +08:00
|
|
|
|
{
|
|
|
|
|
hid_t dcpl; /*dataset creation property list*/
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
hid_t type; /*data type of dataset */
|
1998-11-12 00:33:33 +08:00
|
|
|
|
hid_t space; /*data space of dataset */
|
1998-08-28 23:24:12 +08:00
|
|
|
|
int nf; /*number of filters */
|
|
|
|
|
unsigned filt_flags; /*filter flags */
|
|
|
|
|
H5Z_filter_t filt_id; /*filter identification number */
|
|
|
|
|
unsigned cd_values[20]; /*filter client data values */
|
|
|
|
|
size_t cd_nelmts; /*filter client number of values*/
|
|
|
|
|
size_t cd_num; /*filter client data counter */
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
char f_name[256]; /*filter/file name */
|
1998-08-28 23:24:12 +08:00
|
|
|
|
char s[64]; /*temporary string buffer */
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
off_t f_offset; /*offset in external file */
|
|
|
|
|
hsize_t f_size; /*bytes used in external file */
|
1999-04-23 20:31:21 +08:00
|
|
|
|
hsize_t total, used; /*total size or offset */
|
1998-11-12 00:33:33 +08:00
|
|
|
|
hsize_t chsize[64]; /*chunk size in elements */
|
|
|
|
|
int ndims; /*dimensionality */
|
1999-01-22 02:33:39 +08:00
|
|
|
|
int n, max_len; /*max extern file name length */
|
1999-04-27 22:47:54 +08:00
|
|
|
|
double utilization; /*percent utilization of storage*/
|
1998-08-28 23:24:12 +08:00
|
|
|
|
int i;
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
|
1998-08-28 23:24:12 +08:00
|
|
|
|
if (verbose_g>0) {
|
|
|
|
|
dcpl = H5Dget_create_plist(dset);
|
1998-11-12 00:33:33 +08:00
|
|
|
|
space = H5Dget_space(dset);
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
type = H5Dget_type(dset);
|
1998-11-12 00:33:33 +08:00
|
|
|
|
|
|
|
|
|
/* Print information about chunked storage */
|
|
|
|
|
if (H5D_CHUNKED==H5Pget_layout(dcpl)) {
|
|
|
|
|
ndims = H5Pget_chunk(dcpl, NELMTS(chsize), chsize/*out*/);
|
|
|
|
|
printf(" %-10s {", "Chunks:");
|
|
|
|
|
total = H5Tget_size(type);
|
|
|
|
|
for (i=0; i<ndims; i++) {
|
|
|
|
|
printf("%s%lu", i?", ":"", (unsigned long)(chsize[i]));
|
|
|
|
|
total *= chsize[i];
|
|
|
|
|
}
|
|
|
|
|
printf("} %lu bytes\n", (unsigned long)total);
|
|
|
|
|
}
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
|
1999-04-23 20:31:21 +08:00
|
|
|
|
/* Print total raw storage size */
|
|
|
|
|
used = H5Sget_simple_extent_npoints(space) * H5Tget_size(type);
|
|
|
|
|
total = H5Dget_storage_size(dset);
|
|
|
|
|
printf(" %-10s ", "Storage:");
|
|
|
|
|
printf("%lu logical byte%s, %lu allocated byte%s",
|
|
|
|
|
(unsigned long)used, 1==used?"":"s",
|
|
|
|
|
(unsigned long)total, 1==total?"":"s");
|
|
|
|
|
if (total>0) {
|
1999-04-27 22:47:54 +08:00
|
|
|
|
#ifdef WIN32
|
1999-05-04 21:12:42 +08:00
|
|
|
|
hsize_t mask = (hsize_t)1 << (8*sizeof(hsize_t)-1);
|
1999-04-27 22:47:54 +08:00
|
|
|
|
if ((used & mask) || (total & mask)) {
|
|
|
|
|
total = 0; /*prevent utilization printing*/
|
|
|
|
|
} else {
|
|
|
|
|
utilization = (hssize_t)used*100.0 /(hssize_t)total;
|
|
|
|
|
}
|
|
|
|
|
#else
|
|
|
|
|
utilization = (used*100.0)/total;
|
|
|
|
|
#endif
|
1999-05-13 23:18:12 +08:00
|
|
|
|
printf(", %1.2f%% utilization", utilization/*(used*100.0)/total*/);
|
1999-04-23 20:31:21 +08:00
|
|
|
|
}
|
|
|
|
|
putchar('\n');
|
|
|
|
|
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
/* Print information about external strorage */
|
|
|
|
|
if ((nf = H5Pget_external_count(dcpl))>0) {
|
1999-01-22 02:33:39 +08:00
|
|
|
|
for (i=0, max_len=0; i<nf; i++) {
|
|
|
|
|
H5Pget_external(dcpl, i, sizeof(f_name), f_name, NULL, NULL);
|
|
|
|
|
n = display_string(NULL, f_name, TRUE);
|
|
|
|
|
max_len = MAX(max_len, n);
|
|
|
|
|
}
|
|
|
|
|
printf(" %-10s %d external file%s\n",
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
"Extern:", nf, 1==nf?"":"s");
|
1999-01-22 02:33:39 +08:00
|
|
|
|
printf(" %4s %10s %10s %10s %s\n",
|
|
|
|
|
"ID", "DSet-Addr", "File-Addr", "Bytes", "File");
|
|
|
|
|
printf(" %4s %10s %10s %10s ",
|
|
|
|
|
"----", "----------", "----------", "----------");
|
|
|
|
|
for (i=0; i<max_len; i++) putchar('-');
|
|
|
|
|
putchar('\n');
|
1998-11-12 00:33:33 +08:00
|
|
|
|
for (i=0, total=0; i<nf; i++) {
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
if (H5Pget_external(dcpl, i, sizeof(f_name), f_name, &f_offset,
|
|
|
|
|
&f_size)<0) {
|
|
|
|
|
HDfprintf(stdout,
|
|
|
|
|
" #%03d %10Hu %10s %10s ***ERROR*** %s\n",
|
1998-11-12 00:33:33 +08:00
|
|
|
|
i, total, "", "",
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
i+1<nf?"Following addresses are incorrect":"");
|
|
|
|
|
} else if (H5S_UNLIMITED==f_size) {
|
1999-01-22 02:33:39 +08:00
|
|
|
|
HDfprintf(stdout, " #%03d %10Hu %10Hu %10s ",
|
1998-11-12 00:33:33 +08:00
|
|
|
|
i, total, (hsize_t)f_offset, "INF");
|
1999-01-22 02:33:39 +08:00
|
|
|
|
display_string(stdout, f_name, TRUE);
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
} else {
|
1999-01-22 02:33:39 +08:00
|
|
|
|
HDfprintf(stdout, " #%03d %10Hu %10Hu %10Hu ",
|
1998-11-12 00:33:33 +08:00
|
|
|
|
i, total, (hsize_t)f_offset, f_size);
|
1999-01-22 02:33:39 +08:00
|
|
|
|
display_string(stdout, f_name, TRUE);
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
}
|
1999-01-22 02:33:39 +08:00
|
|
|
|
putchar('\n');
|
1998-11-12 00:33:33 +08:00
|
|
|
|
total += f_size;
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
}
|
1999-01-22 02:33:39 +08:00
|
|
|
|
printf(" %4s %10s %10s %10s ",
|
|
|
|
|
"----", "----------", "----------", "----------");
|
|
|
|
|
for (i=0; i<max_len; i++) putchar('-');
|
|
|
|
|
putchar('\n');
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
}
|
|
|
|
|
|
1998-08-28 23:24:12 +08:00
|
|
|
|
/* Print information about raw data filters */
|
|
|
|
|
if ((nf = H5Pget_nfilters(dcpl))>0) {
|
|
|
|
|
for (i=0; i<nf; i++) {
|
|
|
|
|
cd_nelmts = NELMTS(cd_values);
|
|
|
|
|
filt_id = H5Pget_filter(dcpl, i, &filt_flags, &cd_nelmts,
|
|
|
|
|
cd_values, sizeof(f_name), f_name);
|
|
|
|
|
f_name[sizeof(f_name)-1] = '\0';
|
|
|
|
|
sprintf(s, "Filter-%d:", i);
|
|
|
|
|
printf(" %-10s %s-%u %s {", s,
|
|
|
|
|
f_name[0]?f_name:"method",
|
|
|
|
|
(unsigned)filt_id,
|
|
|
|
|
filt_flags & H5Z_FLAG_OPTIONAL?"OPT":"");
|
|
|
|
|
for (cd_num=0; cd_num<cd_nelmts; cd_num++) {
|
|
|
|
|
printf("%s%u", cd_num?", ":"", cd_values[cd_num]);
|
|
|
|
|
}
|
|
|
|
|
printf("}\n");
|
|
|
|
|
}
|
|
|
|
|
}
|
1998-11-12 00:33:33 +08:00
|
|
|
|
|
|
|
|
|
/* Print data type */
|
|
|
|
|
printf(" %-10s ", "Type:");
|
|
|
|
|
display_type(type, 15);
|
|
|
|
|
printf("\n");
|
|
|
|
|
|
1999-04-30 23:54:52 +08:00
|
|
|
|
/* Print address information */
|
|
|
|
|
if (address_g) H5Ddebug(dset, 0);
|
|
|
|
|
|
1998-11-12 00:33:33 +08:00
|
|
|
|
/* Close stuff */
|
|
|
|
|
H5Tclose(type);
|
|
|
|
|
H5Sclose(space);
|
1998-08-28 23:24:12 +08:00
|
|
|
|
H5Pclose(dcpl);
|
|
|
|
|
}
|
|
|
|
|
|
1999-04-30 23:54:52 +08:00
|
|
|
|
if (data_g) dump_dataset_values(dset);
|
1998-08-28 23:24:12 +08:00
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
1999-01-22 02:33:39 +08:00
|
|
|
|
|
|
|
|
|
/*-------------------------------------------------------------------------
|
|
|
|
|
* Function: group_list2
|
|
|
|
|
*
|
|
|
|
|
* Purpose: List information about a group which should appear after
|
|
|
|
|
* information which is general to all objects.
|
|
|
|
|
*
|
|
|
|
|
* Return: Success: 0
|
|
|
|
|
*
|
|
|
|
|
* Failure: -1
|
|
|
|
|
*
|
|
|
|
|
* Programmer: Robb Matzke
|
|
|
|
|
* Thursday, January 21, 1999
|
|
|
|
|
*
|
|
|
|
|
* Modifications:
|
|
|
|
|
*
|
|
|
|
|
*-------------------------------------------------------------------------
|
|
|
|
|
*/
|
|
|
|
|
static herr_t
|
|
|
|
|
group_list2(hid_t grp, const char *name)
|
|
|
|
|
{
|
|
|
|
|
iter_t iter;
|
|
|
|
|
|
|
|
|
|
if (recursive_g) {
|
|
|
|
|
iter.container = name;
|
|
|
|
|
H5Giterate(grp, ".", NULL, list, &iter);
|
|
|
|
|
}
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
|
|
|
|
|
/*-------------------------------------------------------------------------
|
|
|
|
|
* Function: datatype_list2
|
|
|
|
|
*
|
|
|
|
|
* Purpose: List information about a data type which should appear after
|
|
|
|
|
* information which is general to all objects.
|
|
|
|
|
*
|
|
|
|
|
* Return: Success: 0
|
|
|
|
|
*
|
|
|
|
|
* Failure: -1
|
|
|
|
|
*
|
|
|
|
|
* Programmer: Robb Matzke
|
|
|
|
|
* Thursday, November 5, 1998
|
|
|
|
|
*
|
|
|
|
|
* Modifications:
|
|
|
|
|
*
|
|
|
|
|
*-------------------------------------------------------------------------
|
|
|
|
|
*/
|
|
|
|
|
static herr_t
|
1999-04-16 03:57:50 +08:00
|
|
|
|
datatype_list2(hid_t type, const char UNUSED *name)
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
{
|
1999-03-30 19:38:34 +08:00
|
|
|
|
if (verbose_g>0) {
|
|
|
|
|
printf(" %-10s ", "Type:");
|
|
|
|
|
display_type(type, 15);
|
|
|
|
|
printf("\n");
|
|
|
|
|
}
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
[svn-r876] Changes since 19981102
----------------------
./bin/snapshot
Made same fix as for the release script yesterday.
./src/H5D.c
./src/H5Dprivate.h
./src/H5G.c
./src/H5Gprivate.h
./src/H5Gpublic.h
./src/H5O.c
./src/H5Oprivate.h
./src/H5RA.c
./src/H5RAprivate.h
./src/H5T.c
./src/H5Tprivate.h
Improved object type checking. Instead of determining the
object type by trying to open each of the possible types, we
keep a table of associations between object type number (like
H5G_GROUP, H5G_DATASET, H5D_TYPE, and H5D_RAGGED) and an `isa'
function that returns true if the object header has the right
messages to make the object a particular type. This mechanism
also allows specialization of object types by permitting an
object to satisfy more than one `isa' function.
Added `isa' functions for groups, datasets, ragged arrays, and
committed data types.
./src/H5config.h.in
Added HAVE_STAT_ST_BLOCKS. I thought this had already been
added, but apparently not.
./tools/h5ls.c
Removed system include files since they're already included by
H5private.h and since I wasn't including them portably anyway.
By default, 1-byte integer types are printed as integer values
instead of ASCII characters. However, the `-s' or `--string'
command-line switch causes the data to be interpretted as
ASCII. String data types are always printed as character
data.
Ragged arrays are now identified as ragged arrays and h5ls
doesn't descend into the group automatically. This uses the
new object type specialization stuff.
./tools/h5tools.c
./tools/h5tools.h
Added the ability to print 1-byte integer types as either
ASCII or numeric data instead of always ASCII. The default is
to print as numeric data.
1998-11-06 04:28:34 +08:00
|
|
|
|
|
|
|
|
|
/*-------------------------------------------------------------------------
|
|
|
|
|
* Function: ragged_list2
|
|
|
|
|
*
|
|
|
|
|
* Purpose: List information about a ragged array which should appear
|
|
|
|
|
* after information which is general to all objects.
|
|
|
|
|
*
|
|
|
|
|
* Return: Success: 0
|
|
|
|
|
*
|
|
|
|
|
* Failure: -1
|
|
|
|
|
*
|
|
|
|
|
* Programmer: Robb Matzke
|
|
|
|
|
* Thursday, November 5, 1998
|
|
|
|
|
*
|
|
|
|
|
* Modifications:
|
|
|
|
|
*
|
|
|
|
|
*-------------------------------------------------------------------------
|
|
|
|
|
*/
|
|
|
|
|
static herr_t
|
1999-04-16 03:57:50 +08:00
|
|
|
|
ragged_list2(hid_t UNUSED ra, const char UNUSED *name)
|
[svn-r876] Changes since 19981102
----------------------
./bin/snapshot
Made same fix as for the release script yesterday.
./src/H5D.c
./src/H5Dprivate.h
./src/H5G.c
./src/H5Gprivate.h
./src/H5Gpublic.h
./src/H5O.c
./src/H5Oprivate.h
./src/H5RA.c
./src/H5RAprivate.h
./src/H5T.c
./src/H5Tprivate.h
Improved object type checking. Instead of determining the
object type by trying to open each of the possible types, we
keep a table of associations between object type number (like
H5G_GROUP, H5G_DATASET, H5D_TYPE, and H5D_RAGGED) and an `isa'
function that returns true if the object header has the right
messages to make the object a particular type. This mechanism
also allows specialization of object types by permitting an
object to satisfy more than one `isa' function.
Added `isa' functions for groups, datasets, ragged arrays, and
committed data types.
./src/H5config.h.in
Added HAVE_STAT_ST_BLOCKS. I thought this had already been
added, but apparently not.
./tools/h5ls.c
Removed system include files since they're already included by
H5private.h and since I wasn't including them portably anyway.
By default, 1-byte integer types are printed as integer values
instead of ASCII characters. However, the `-s' or `--string'
command-line switch causes the data to be interpretted as
ASCII. String data types are always printed as character
data.
Ragged arrays are now identified as ragged arrays and h5ls
doesn't descend into the group automatically. This uses the
new object type specialization stuff.
./tools/h5tools.c
./tools/h5tools.h
Added the ability to print 1-byte integer types as either
ASCII or numeric data instead of always ASCII. The default is
to print as numeric data.
1998-11-06 04:28:34 +08:00
|
|
|
|
{
|
1999-04-30 23:54:52 +08:00
|
|
|
|
if (data_g) {
|
[svn-r876] Changes since 19981102
----------------------
./bin/snapshot
Made same fix as for the release script yesterday.
./src/H5D.c
./src/H5Dprivate.h
./src/H5G.c
./src/H5Gprivate.h
./src/H5Gpublic.h
./src/H5O.c
./src/H5Oprivate.h
./src/H5RA.c
./src/H5RAprivate.h
./src/H5T.c
./src/H5Tprivate.h
Improved object type checking. Instead of determining the
object type by trying to open each of the possible types, we
keep a table of associations between object type number (like
H5G_GROUP, H5G_DATASET, H5D_TYPE, and H5D_RAGGED) and an `isa'
function that returns true if the object header has the right
messages to make the object a particular type. This mechanism
also allows specialization of object types by permitting an
object to satisfy more than one `isa' function.
Added `isa' functions for groups, datasets, ragged arrays, and
committed data types.
./src/H5config.h.in
Added HAVE_STAT_ST_BLOCKS. I thought this had already been
added, but apparently not.
./tools/h5ls.c
Removed system include files since they're already included by
H5private.h and since I wasn't including them portably anyway.
By default, 1-byte integer types are printed as integer values
instead of ASCII characters. However, the `-s' or `--string'
command-line switch causes the data to be interpretted as
ASCII. String data types are always printed as character
data.
Ragged arrays are now identified as ragged arrays and h5ls
doesn't descend into the group automatically. This uses the
new object type specialization stuff.
./tools/h5tools.c
./tools/h5tools.h
Added the ability to print 1-byte integer types as either
ASCII or numeric data instead of always ASCII. The default is
to print as numeric data.
1998-11-06 04:28:34 +08:00
|
|
|
|
puts(" Data: Not implemented yet (see values of member");
|
|
|
|
|
puts(" datasets `raw', `over', and `meta')");
|
|
|
|
|
}
|
|
|
|
|
return -1;
|
|
|
|
|
}
|
|
|
|
|
|
1998-08-28 23:24:12 +08:00
|
|
|
|
|
|
|
|
|
/*-------------------------------------------------------------------------
|
|
|
|
|
* Function: link_open
|
|
|
|
|
*
|
|
|
|
|
* Purpose: This gets called to open a symbolic link. Since symbolic
|
|
|
|
|
* links don't correspond to actual objects we simply print the
|
|
|
|
|
* link information and return failure.
|
|
|
|
|
*
|
1999-06-14 23:07:58 +08:00
|
|
|
|
* Return: Success: 0 - an invalid object but successful return
|
|
|
|
|
* of this function.
|
1998-08-28 23:24:12 +08:00
|
|
|
|
*
|
|
|
|
|
* Failure: -1
|
|
|
|
|
*
|
|
|
|
|
* Programmer: Robb Matzke
|
|
|
|
|
* Thursday, August 27, 1998
|
|
|
|
|
*
|
|
|
|
|
* Modifications:
|
|
|
|
|
*
|
|
|
|
|
*-------------------------------------------------------------------------
|
|
|
|
|
*/
|
|
|
|
|
static hid_t
|
|
|
|
|
link_open(hid_t location, const char *name)
|
|
|
|
|
{
|
|
|
|
|
char buf[64];
|
|
|
|
|
|
|
|
|
|
if (H5Gget_linkval (location, name, sizeof(buf), buf)<0) return -1;
|
|
|
|
|
if (NULL==HDmemchr(buf, 0, sizeof(buf))) {
|
|
|
|
|
strcpy(buf+sizeof(buf)-4, "...");
|
|
|
|
|
}
|
1999-06-14 23:07:58 +08:00
|
|
|
|
fputs(buf, stdout);
|
1998-08-28 23:24:12 +08:00
|
|
|
|
|
1999-06-14 23:07:58 +08:00
|
|
|
|
return 0;
|
1998-08-28 23:24:12 +08:00
|
|
|
|
}
|
|
|
|
|
|
1998-03-25 07:18:34 +08:00
|
|
|
|
|
|
|
|
|
/*-------------------------------------------------------------------------
|
|
|
|
|
* Function: list
|
|
|
|
|
*
|
|
|
|
|
* Purpose: Prints the group member name.
|
|
|
|
|
*
|
|
|
|
|
* Return: Success: 0
|
|
|
|
|
*
|
|
|
|
|
* Failure: -1
|
|
|
|
|
*
|
|
|
|
|
* Programmer: Robb Matzke
|
|
|
|
|
* Monday, March 23, 1998
|
|
|
|
|
*
|
|
|
|
|
* Modifications:
|
|
|
|
|
*
|
|
|
|
|
*-------------------------------------------------------------------------
|
|
|
|
|
*/
|
|
|
|
|
static herr_t
|
1999-01-22 02:33:39 +08:00
|
|
|
|
list (hid_t group, const char *name, void *_iter)
|
1998-03-25 07:18:34 +08:00
|
|
|
|
{
|
1999-01-22 02:33:39 +08:00
|
|
|
|
hid_t obj=-1;
|
|
|
|
|
char buf[512], comment[50], *fullname=NULL, *s=NULL;
|
1998-08-28 23:24:12 +08:00
|
|
|
|
H5G_stat_t sb;
|
1998-07-30 00:43:59 +08:00
|
|
|
|
struct tm *tm;
|
1998-08-28 23:24:12 +08:00
|
|
|
|
herr_t status;
|
1999-01-22 02:33:39 +08:00
|
|
|
|
iter_t *iter = (iter_t*)_iter;
|
|
|
|
|
int n;
|
1998-04-15 00:44:46 +08:00
|
|
|
|
|
1999-01-22 02:33:39 +08:00
|
|
|
|
/* Print the object name, either full name or base name */
|
|
|
|
|
fullname = fix_name(iter->container, name);
|
|
|
|
|
if (fullname_g) {
|
|
|
|
|
n = display_string(stdout, fullname, TRUE);
|
1999-04-26 22:43:53 +08:00
|
|
|
|
printf("%*s ", MAX(0, 24-n), "");
|
1999-01-22 02:33:39 +08:00
|
|
|
|
} else {
|
|
|
|
|
n = display_string(stdout, name, TRUE);
|
1999-04-26 22:43:53 +08:00
|
|
|
|
printf("%*s ", MAX(0, 24-n), "");
|
1999-01-22 02:33:39 +08:00
|
|
|
|
}
|
|
|
|
|
|
1998-08-28 23:24:12 +08:00
|
|
|
|
/* Get object information */
|
|
|
|
|
H5E_BEGIN_TRY {
|
1998-09-03 05:16:24 +08:00
|
|
|
|
status = H5Gget_objinfo(group, name, FALSE, &sb);
|
1998-08-28 23:24:12 +08:00
|
|
|
|
} H5E_END_TRY;
|
|
|
|
|
if (status<0) {
|
|
|
|
|
puts("**NOT FOUND**");
|
|
|
|
|
return 0;
|
|
|
|
|
} else if (sb.type<0 || sb.type>=H5G_NTYPES) {
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
printf("Unknown type(%d)", sb.type);
|
|
|
|
|
sb.type = -1;
|
|
|
|
|
}
|
|
|
|
|
if (sb.type>=0 && dispatch_g[sb.type].name) {
|
|
|
|
|
fputs(dispatch_g[sb.type].name, stdout);
|
1998-03-25 07:18:34 +08:00
|
|
|
|
}
|
1998-08-28 23:24:12 +08:00
|
|
|
|
|
1999-01-22 02:33:39 +08:00
|
|
|
|
/*
|
|
|
|
|
* If the object has already been printed then just show the object ID
|
|
|
|
|
* and return.
|
|
|
|
|
*/
|
|
|
|
|
if ((s=sym_lookup(&sb))) {
|
|
|
|
|
printf(", same as ");
|
|
|
|
|
display_string(stdout, s, TRUE);
|
|
|
|
|
printf("\n");
|
|
|
|
|
goto done;
|
|
|
|
|
} else {
|
|
|
|
|
sym_insert(&sb, fullname);
|
|
|
|
|
}
|
|
|
|
|
|
1998-08-28 23:24:12 +08:00
|
|
|
|
/*
|
|
|
|
|
* Open the object. Not all objects can be opened. If this is the case
|
|
|
|
|
* then return right away.
|
|
|
|
|
*/
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
if (sb.type>=0 &&
|
|
|
|
|
(NULL==dispatch_g[sb.type].open ||
|
1999-01-22 02:33:39 +08:00
|
|
|
|
(obj=(dispatch_g[sb.type].open)(group, name))<0)) {
|
|
|
|
|
printf(" *ERROR*\n");
|
|
|
|
|
goto done;
|
|
|
|
|
}
|
|
|
|
|
|
1998-08-28 23:24:12 +08:00
|
|
|
|
/*
|
|
|
|
|
* List the first line of information for the object.
|
|
|
|
|
*/
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
if (sb.type>=0 && dispatch_g[sb.type].list1) {
|
|
|
|
|
(dispatch_g[sb.type].list1)(obj);
|
|
|
|
|
}
|
1998-08-28 23:24:12 +08:00
|
|
|
|
putchar('\n');
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* Show detailed information about the object, beginning with information
|
|
|
|
|
* which is common to all objects.
|
|
|
|
|
*/
|
1999-06-14 23:07:58 +08:00
|
|
|
|
if (verbose_g>0 && H5G_LINK!=sb.type) {
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
if (sb.type>=0) H5Aiterate(obj, NULL, list_attr, NULL);
|
1998-08-28 23:24:12 +08:00
|
|
|
|
printf(" %-10s %lu:%lu:%lu:%lu\n", "Location:",
|
|
|
|
|
sb.fileno[1], sb.fileno[0], sb.objno[1], sb.objno[0]);
|
|
|
|
|
printf(" %-10s %u\n", "Links:", sb.nlink);
|
|
|
|
|
if (sb.mtime>0 && NULL!=(tm=localtime(&(sb.mtime)))) {
|
|
|
|
|
strftime(buf, sizeof(buf), "%Y-%m-%d %H:%M:%S %Z", tm);
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
printf(" %-10s %s\n", "Modified:", buf);
|
1998-07-30 00:43:59 +08:00
|
|
|
|
}
|
|
|
|
|
comment[0] = '\0';
|
|
|
|
|
H5Gget_comment(group, name, sizeof(comment), comment);
|
|
|
|
|
strcpy(comment+sizeof(comment)-4, "...");
|
1999-01-22 02:33:39 +08:00
|
|
|
|
if (comment[0]) {
|
|
|
|
|
printf(" %-10s \"", "Comment:");
|
|
|
|
|
display_string(stdout, comment, FALSE);
|
|
|
|
|
puts("\"");
|
|
|
|
|
}
|
1998-07-30 00:43:59 +08:00
|
|
|
|
}
|
1999-06-14 23:07:58 +08:00
|
|
|
|
if (sb.type>=0 && dispatch_g[sb.type].list2) {
|
1999-01-22 02:33:39 +08:00
|
|
|
|
(dispatch_g[sb.type].list2)(obj, fullname);
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
}
|
1998-07-30 00:43:59 +08:00
|
|
|
|
|
1998-08-28 23:24:12 +08:00
|
|
|
|
/*
|
|
|
|
|
* Close the object.
|
|
|
|
|
*/
|
1999-01-22 02:33:39 +08:00
|
|
|
|
done:
|
1999-06-14 23:07:58 +08:00
|
|
|
|
if (sb.type>=0 && obj>=0 && dispatch_g[sb.type].close) {
|
|
|
|
|
(dispatch_g[sb.type].close)(obj);
|
|
|
|
|
}
|
1999-01-22 02:33:39 +08:00
|
|
|
|
if (fullname) free(fullname);
|
1998-03-25 07:18:34 +08:00
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
1999-01-22 02:33:39 +08:00
|
|
|
|
|
|
|
|
|
/*-------------------------------------------------------------------------
|
|
|
|
|
* Function: fix_name
|
|
|
|
|
*
|
|
|
|
|
* Purpose: Returns a malloc'd buffer that contains the PATH and BASE
|
|
|
|
|
* names separated by a single slash. It also removes duplicate
|
1999-06-14 23:07:58 +08:00
|
|
|
|
* and trailing slashes.
|
1999-01-22 02:33:39 +08:00
|
|
|
|
*
|
|
|
|
|
* Return: Success: Ptr to fixed name from malloc()
|
|
|
|
|
*
|
|
|
|
|
* Failure: NULL
|
|
|
|
|
*
|
|
|
|
|
* Programmer: Robb Matzke
|
|
|
|
|
* Thursday, January 21, 1999
|
|
|
|
|
*
|
|
|
|
|
* Modifications:
|
|
|
|
|
*
|
|
|
|
|
*-------------------------------------------------------------------------
|
|
|
|
|
*/
|
|
|
|
|
static char *
|
|
|
|
|
fix_name(const char *path, const char *base)
|
|
|
|
|
{
|
|
|
|
|
size_t n = (path?strlen(path):0) + (base?strlen(base):0) + 3;
|
|
|
|
|
char *s = malloc(n), prev='\0';
|
|
|
|
|
int len=0;
|
|
|
|
|
|
|
|
|
|
if (path) {
|
1999-06-14 23:07:58 +08:00
|
|
|
|
/* Path, followed by slash */
|
|
|
|
|
#ifdef H5LS_PREPEND_FILENAME
|
|
|
|
|
if ('/'!=*path) s[len++] = '/';
|
|
|
|
|
#endif
|
1999-01-22 02:33:39 +08:00
|
|
|
|
for (/*void*/; *path; path++) {
|
|
|
|
|
if ('/'!=*path || '/'!=prev) prev = s[len++] = *path;
|
|
|
|
|
}
|
|
|
|
|
if ('/'!=prev) prev = s[len++] = '/';
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (base) {
|
|
|
|
|
/* Base name w/o trailing slashes */
|
|
|
|
|
const char *end = base + strlen(base);
|
|
|
|
|
while (end>base && '/'==end[-1]) --end;
|
|
|
|
|
for (/*void*/; base<end; base++) {
|
|
|
|
|
if ('/'!=*base || '/'!=prev) prev = s[len++] = *base;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
s[len] = '\0';
|
|
|
|
|
return s;
|
|
|
|
|
}
|
|
|
|
|
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
|
|
|
|
|
/*-------------------------------------------------------------------------
|
|
|
|
|
* Function: get_width
|
|
|
|
|
*
|
|
|
|
|
* Purpose: Figure out how wide the screen is. This is highly
|
|
|
|
|
* unportable, but the user can always override the width we
|
|
|
|
|
* detect by giving a command-line option. These code snippets
|
|
|
|
|
* were borrowed from the GNU less(1).
|
|
|
|
|
*
|
|
|
|
|
* Return: Success: Number of columns.
|
|
|
|
|
*
|
|
|
|
|
* Failure: Some default number of columms.
|
|
|
|
|
*
|
|
|
|
|
* Programmer: Robb Matzke
|
|
|
|
|
* Friday, November 6, 1998
|
|
|
|
|
*
|
|
|
|
|
* Modifications:
|
|
|
|
|
*
|
|
|
|
|
*-------------------------------------------------------------------------
|
|
|
|
|
*/
|
|
|
|
|
static int
|
|
|
|
|
get_width(void)
|
|
|
|
|
{
|
|
|
|
|
int width = 80; /*the default */
|
|
|
|
|
char *s;
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* Try to get it from the COLUMNS environment variable first since it's
|
|
|
|
|
* value is sometimes wrong.
|
|
|
|
|
*/
|
1999-09-28 00:11:15 +08:00
|
|
|
|
if ((s=getenv("COLUMNS")) && *s && isdigit((int)*s)) {
|
1999-02-25 23:40:27 +08:00
|
|
|
|
width = (int)strtol(s, NULL, 0);
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
#if defined(HAVE_STRUCT_VIDEOCONFIG) && defined(HAVE__GETVIDEOCONFIG)
|
|
|
|
|
{
|
|
|
|
|
/* Microsoft C */
|
|
|
|
|
struct videoconfig w;
|
|
|
|
|
_getvideoconfig(&w);
|
|
|
|
|
width = w.numtextcols;
|
|
|
|
|
}
|
|
|
|
|
#elif defined(HAVE_STRUCT_TEXT_INFO) && defined(HAVE_GETTEXTINFO)
|
|
|
|
|
{
|
|
|
|
|
/* Borland C or DJGPPC */
|
|
|
|
|
struct text_info w;
|
|
|
|
|
gettextinfo(&w);
|
|
|
|
|
width = w.screenwidth;
|
|
|
|
|
}
|
|
|
|
|
#elif defined(HAVE_GETCONSOLESCREENBUFFERINFO)
|
|
|
|
|
{
|
|
|
|
|
/* Win32 C */
|
|
|
|
|
CONSOLE_SCREEN_BUFFER_INFO scr;
|
|
|
|
|
GetConsoleScreenBufferInfo(con_out, &scr);
|
|
|
|
|
width = scr.srWindow.Right - scr.srWindow.Left + 1;
|
|
|
|
|
}
|
|
|
|
|
#elif defined(HAVE__SCRSIZE)
|
|
|
|
|
{
|
|
|
|
|
/* OS/2 */
|
|
|
|
|
int w[2];
|
|
|
|
|
_scrsize(w);
|
|
|
|
|
width = w[0];
|
|
|
|
|
}
|
|
|
|
|
#elif defined(HAVE_TIOCGWINSZ) && defined(HAVE_IOCTL)
|
|
|
|
|
{
|
|
|
|
|
/* Unix with ioctl(TIOCGWINSZ) */
|
|
|
|
|
struct winsize w;
|
|
|
|
|
if (ioctl(2, TIOCGWINSZ, &w)>=0 && w.ws_col>0) {
|
|
|
|
|
width = w.ws_col;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
#elif defined(HAVE_TIOCGETD) && defined(HAVE_IOCTL)
|
|
|
|
|
{
|
|
|
|
|
/* Unix with ioctl(TIOCGETD) */
|
|
|
|
|
struct uwdata w;
|
|
|
|
|
if (ioctl(2, WIOCGETD, &w)>=0 && w.uw_width>0) {
|
|
|
|
|
width = w.uw_width / w.uw_hs;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
/* Set to at least 1 */
|
|
|
|
|
if (width<1) width = 1;
|
|
|
|
|
return width;
|
|
|
|
|
}
|
|
|
|
|
|
1998-03-25 07:18:34 +08:00
|
|
|
|
|
|
|
|
|
/*-------------------------------------------------------------------------
|
|
|
|
|
* Function: main
|
|
|
|
|
*
|
|
|
|
|
* Purpose: Opens a file and lists the specified group
|
|
|
|
|
*
|
|
|
|
|
* Return: Success: 0
|
|
|
|
|
*
|
|
|
|
|
* Failure: 1
|
|
|
|
|
*
|
|
|
|
|
* Programmer: Robb Matzke
|
|
|
|
|
* Monday, March 23, 1998
|
|
|
|
|
*
|
|
|
|
|
* Modifications:
|
|
|
|
|
*
|
|
|
|
|
*-------------------------------------------------------------------------
|
|
|
|
|
*/
|
|
|
|
|
int
|
|
|
|
|
main (int argc, char *argv[])
|
|
|
|
|
{
|
2000-06-07 02:07:20 +08:00
|
|
|
|
hid_t file=-1, root=-1;
|
1999-06-14 23:07:58 +08:00
|
|
|
|
char *fname=NULL, *oname=NULL, *x;
|
1998-05-14 01:58:24 +08:00
|
|
|
|
const char *progname;
|
1999-03-05 04:22:42 +08:00
|
|
|
|
const char *s = NULL;
|
1999-01-22 02:33:39 +08:00
|
|
|
|
char *rest, *container=NULL;
|
2000-06-07 02:07:20 +08:00
|
|
|
|
int argno;
|
[svn-r876] Changes since 19981102
----------------------
./bin/snapshot
Made same fix as for the release script yesterday.
./src/H5D.c
./src/H5Dprivate.h
./src/H5G.c
./src/H5Gprivate.h
./src/H5Gpublic.h
./src/H5O.c
./src/H5Oprivate.h
./src/H5RA.c
./src/H5RAprivate.h
./src/H5T.c
./src/H5Tprivate.h
Improved object type checking. Instead of determining the
object type by trying to open each of the possible types, we
keep a table of associations between object type number (like
H5G_GROUP, H5G_DATASET, H5D_TYPE, and H5D_RAGGED) and an `isa'
function that returns true if the object header has the right
messages to make the object a particular type. This mechanism
also allows specialization of object types by permitting an
object to satisfy more than one `isa' function.
Added `isa' functions for groups, datasets, ragged arrays, and
committed data types.
./src/H5config.h.in
Added HAVE_STAT_ST_BLOCKS. I thought this had already been
added, but apparently not.
./tools/h5ls.c
Removed system include files since they're already included by
H5private.h and since I wasn't including them portably anyway.
By default, 1-byte integer types are printed as integer values
instead of ASCII characters. However, the `-s' or `--string'
command-line switch causes the data to be interpretted as
ASCII. String data types are always printed as character
data.
Ragged arrays are now identified as ragged arrays and h5ls
doesn't descend into the group automatically. This uses the
new object type specialization stuff.
./tools/h5tools.c
./tools/h5tools.h
Added the ability to print 1-byte integer types as either
ASCII or numeric data instead of always ASCII. The default is
to print as numeric data.
1998-11-06 04:28:34 +08:00
|
|
|
|
H5G_stat_t sb;
|
1999-01-22 02:33:39 +08:00
|
|
|
|
iter_t iter;
|
1999-06-14 23:07:58 +08:00
|
|
|
|
static char root_name[] = "/";
|
2000-06-24 01:52:21 +08:00
|
|
|
|
char drivername[50];
|
2000-05-03 23:48:20 +08:00
|
|
|
|
|
1999-06-14 23:07:58 +08:00
|
|
|
|
|
|
|
|
|
/* Build display table */
|
1998-08-28 23:24:12 +08:00
|
|
|
|
DISPATCH(H5G_DATASET, "Dataset", H5Dopen, H5Dclose,
|
|
|
|
|
dataset_list1, dataset_list2);
|
|
|
|
|
DISPATCH(H5G_GROUP, "Group", H5Gopen, H5Gclose,
|
1999-01-22 02:33:39 +08:00
|
|
|
|
NULL, group_list2);
|
1998-08-28 23:24:12 +08:00
|
|
|
|
DISPATCH(H5G_TYPE, "Type", H5Topen, H5Tclose,
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
NULL, datatype_list2);
|
1998-08-28 23:24:12 +08:00
|
|
|
|
DISPATCH(H5G_LINK, "-> ", link_open, NULL,
|
|
|
|
|
NULL, NULL);
|
[svn-r876] Changes since 19981102
----------------------
./bin/snapshot
Made same fix as for the release script yesterday.
./src/H5D.c
./src/H5Dprivate.h
./src/H5G.c
./src/H5Gprivate.h
./src/H5Gpublic.h
./src/H5O.c
./src/H5Oprivate.h
./src/H5RA.c
./src/H5RAprivate.h
./src/H5T.c
./src/H5Tprivate.h
Improved object type checking. Instead of determining the
object type by trying to open each of the possible types, we
keep a table of associations between object type number (like
H5G_GROUP, H5G_DATASET, H5D_TYPE, and H5D_RAGGED) and an `isa'
function that returns true if the object header has the right
messages to make the object a particular type. This mechanism
also allows specialization of object types by permitting an
object to satisfy more than one `isa' function.
Added `isa' functions for groups, datasets, ragged arrays, and
committed data types.
./src/H5config.h.in
Added HAVE_STAT_ST_BLOCKS. I thought this had already been
added, but apparently not.
./tools/h5ls.c
Removed system include files since they're already included by
H5private.h and since I wasn't including them portably anyway.
By default, 1-byte integer types are printed as integer values
instead of ASCII characters. However, the `-s' or `--string'
command-line switch causes the data to be interpretted as
ASCII. String data types are always printed as character
data.
Ragged arrays are now identified as ragged arrays and h5ls
doesn't descend into the group automatically. This uses the
new object type specialization stuff.
./tools/h5tools.c
./tools/h5tools.h
Added the ability to print 1-byte integer types as either
ASCII or numeric data instead of always ASCII. The default is
to print as numeric data.
1998-11-06 04:28:34 +08:00
|
|
|
|
DISPATCH(H5G_RAGGED, "Ragged Array", H5Gopen, H5Gclose,
|
|
|
|
|
NULL, ragged_list2);
|
1998-08-28 23:24:12 +08:00
|
|
|
|
|
1998-07-18 03:03:43 +08:00
|
|
|
|
/* Name of this program without the path */
|
1999-01-22 02:33:39 +08:00
|
|
|
|
if ((progname=strrchr(argv[0], '/'))) progname++;
|
1998-05-14 01:58:24 +08:00
|
|
|
|
else progname = argv[0];
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
|
|
|
|
|
/* Default output width */
|
|
|
|
|
width_g = get_width();
|
|
|
|
|
|
1998-07-18 03:03:43 +08:00
|
|
|
|
/* Switches come before non-switch arguments */
|
|
|
|
|
for (argno=1; argno<argc && '-'==argv[argno][0]; argno++) {
|
|
|
|
|
if (!strcmp(argv[argno], "--")) {
|
|
|
|
|
/* Last switch */
|
|
|
|
|
argno++;
|
|
|
|
|
break;
|
|
|
|
|
} else if (!strcmp(argv[argno], "--help")) {
|
|
|
|
|
usage(progname);
|
|
|
|
|
exit(0);
|
1999-04-30 23:54:52 +08:00
|
|
|
|
} else if (!strcmp(argv[argno], "--address")) {
|
|
|
|
|
address_g = TRUE;
|
|
|
|
|
} else if (!strcmp(argv[argno], "--data")) {
|
|
|
|
|
data_g = TRUE;
|
1999-08-24 20:52:10 +08:00
|
|
|
|
} else if (!strcmp(argv[argno], "--errors")) {
|
|
|
|
|
show_errors_g = TRUE;
|
1999-01-22 02:33:39 +08:00
|
|
|
|
} else if (!strcmp(argv[argno], "--full")) {
|
|
|
|
|
fullname_g = TRUE;
|
1999-08-24 20:52:10 +08:00
|
|
|
|
} else if (!strcmp(argv[argno], "--group")) {
|
2000-06-07 02:07:20 +08:00
|
|
|
|
grp_literal_g = TRUE;
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
} else if (!strcmp(argv[argno], "--label")) {
|
|
|
|
|
label_g = TRUE;
|
1999-01-22 02:33:39 +08:00
|
|
|
|
} else if (!strcmp(argv[argno], "--recursive")) {
|
|
|
|
|
recursive_g = TRUE;
|
|
|
|
|
fullname_g = TRUE;
|
1999-10-01 00:14:14 +08:00
|
|
|
|
} else if (!strcmp(argv[argno], "--simple")) {
|
|
|
|
|
simple_output_g = TRUE;
|
[svn-r876] Changes since 19981102
----------------------
./bin/snapshot
Made same fix as for the release script yesterday.
./src/H5D.c
./src/H5Dprivate.h
./src/H5G.c
./src/H5Gprivate.h
./src/H5Gpublic.h
./src/H5O.c
./src/H5Oprivate.h
./src/H5RA.c
./src/H5RAprivate.h
./src/H5T.c
./src/H5Tprivate.h
Improved object type checking. Instead of determining the
object type by trying to open each of the possible types, we
keep a table of associations between object type number (like
H5G_GROUP, H5G_DATASET, H5D_TYPE, and H5D_RAGGED) and an `isa'
function that returns true if the object header has the right
messages to make the object a particular type. This mechanism
also allows specialization of object types by permitting an
object to satisfy more than one `isa' function.
Added `isa' functions for groups, datasets, ragged arrays, and
committed data types.
./src/H5config.h.in
Added HAVE_STAT_ST_BLOCKS. I thought this had already been
added, but apparently not.
./tools/h5ls.c
Removed system include files since they're already included by
H5private.h and since I wasn't including them portably anyway.
By default, 1-byte integer types are printed as integer values
instead of ASCII characters. However, the `-s' or `--string'
command-line switch causes the data to be interpretted as
ASCII. String data types are always printed as character
data.
Ragged arrays are now identified as ragged arrays and h5ls
doesn't descend into the group automatically. This uses the
new object type specialization stuff.
./tools/h5tools.c
./tools/h5tools.h
Added the ability to print 1-byte integer types as either
ASCII or numeric data instead of always ASCII. The default is
to print as numeric data.
1998-11-06 04:28:34 +08:00
|
|
|
|
} else if (!strcmp(argv[argno], "--string")) {
|
|
|
|
|
string_g = TRUE;
|
[svn-r537] Changes since 19980722
----------------------
./src/H5A.c
./src/H5Apublic.h
./test/tattr.c
Switched the order of the second and third argument of
H5Aget_name() to make it consistent with other functions that
take buffers and buffer sizes.
./src/H5G.c
./src/H5Gpublic.h
./src/H5Gprivate.h
The H5Gget_comment() function returns the size of the comment
including the null terminator. If the object has no comment
then zero is returned. If an error occurs then a negative
value is returned.
./MANIFEST
./tools/Makefile.in
./tools/h5tools.h [NEW]
./tools/h5dump.c [NEW]
Created a library for printing values of datasets in a way
that looks nice. It's not done yet, but I needed it for
debugging the contents of files from Jim Reus.
./tools/h5ls.c
Added the `-d' and `--dump' options which cause the contents
of a dataset to be printed. Added `-w N' and `--width=N'
options to control how wide the raw data output should be. If
you want single-column output then say `-w1'.
Printing dataset values can now handle datasets of any integer
or floating point atomic type. As a special case, integers
which are one byte wide are treated a character strings for
now.
Sample output:
$ h5ls --dump --width=60 banana.hdf
ARCHIVE 0:0:0:744 Dataset {52/Inf}
Data:
(0) "U struct complex { double R; double I; };\012V"
(43) " double;\012"
U 0:0:0:2500 Dataset {256/512}
Data: printing of compound data types is not implemented yet
V 0:0:0:3928 Dataset {256/512}
Data:
(0) 0, 0.015625, 0.03125, 0.046875, 0.0625,
(5) 0.078125, 0.09375, 0.109375, 0.125, 0.140625,
(10) 0.15625, 0.171875, 0.1875, 0.203125, 0.21875,
(15) 0.234375, 0.25, 0.265625, 0.28125, 0.296875,
...
1998-07-24 05:19:17 +08:00
|
|
|
|
} else if (!strncmp(argv[argno], "--width=", 8)) {
|
1999-02-25 23:40:27 +08:00
|
|
|
|
width_g = (int)strtol(argv[argno]+8, &rest, 0);
|
[svn-r537] Changes since 19980722
----------------------
./src/H5A.c
./src/H5Apublic.h
./test/tattr.c
Switched the order of the second and third argument of
H5Aget_name() to make it consistent with other functions that
take buffers and buffer sizes.
./src/H5G.c
./src/H5Gpublic.h
./src/H5Gprivate.h
The H5Gget_comment() function returns the size of the comment
including the null terminator. If the object has no comment
then zero is returned. If an error occurs then a negative
value is returned.
./MANIFEST
./tools/Makefile.in
./tools/h5tools.h [NEW]
./tools/h5dump.c [NEW]
Created a library for printing values of datasets in a way
that looks nice. It's not done yet, but I needed it for
debugging the contents of files from Jim Reus.
./tools/h5ls.c
Added the `-d' and `--dump' options which cause the contents
of a dataset to be printed. Added `-w N' and `--width=N'
options to control how wide the raw data output should be. If
you want single-column output then say `-w1'.
Printing dataset values can now handle datasets of any integer
or floating point atomic type. As a special case, integers
which are one byte wide are treated a character strings for
now.
Sample output:
$ h5ls --dump --width=60 banana.hdf
ARCHIVE 0:0:0:744 Dataset {52/Inf}
Data:
(0) "U struct complex { double R; double I; };\012V"
(43) " double;\012"
U 0:0:0:2500 Dataset {256/512}
Data: printing of compound data types is not implemented yet
V 0:0:0:3928 Dataset {256/512}
Data:
(0) 0, 0.015625, 0.03125, 0.046875, 0.0625,
(5) 0.078125, 0.09375, 0.109375, 0.125, 0.140625,
(10) 0.15625, 0.171875, 0.1875, 0.203125, 0.21875,
(15) 0.234375, 0.25, 0.265625, 0.28125, 0.296875,
...
1998-07-24 05:19:17 +08:00
|
|
|
|
if (width_g<=0 || *rest) {
|
|
|
|
|
usage(progname);
|
|
|
|
|
exit(1);
|
|
|
|
|
}
|
1999-06-14 23:07:58 +08:00
|
|
|
|
} else if (!strcmp(argv[argno], "--width")) {
|
|
|
|
|
if (argno+1>=argc) {
|
|
|
|
|
usage(progname);
|
|
|
|
|
exit(1);
|
|
|
|
|
} else {
|
|
|
|
|
s = argv[++argno];
|
|
|
|
|
}
|
|
|
|
|
width_g = (int)strtol(s, &rest, 0);
|
|
|
|
|
if (width_g<=0 || *rest) {
|
|
|
|
|
usage(progname);
|
|
|
|
|
exit(1);
|
|
|
|
|
}
|
1998-07-18 03:03:43 +08:00
|
|
|
|
} else if (!strcmp(argv[argno], "--verbose")) {
|
|
|
|
|
verbose_g++;
|
|
|
|
|
} else if (!strcmp(argv[argno], "--version")) {
|
2000-04-22 03:28:09 +08:00
|
|
|
|
print_version(progname);
|
1998-07-18 03:03:43 +08:00
|
|
|
|
exit(0);
|
1999-06-07 23:05:02 +08:00
|
|
|
|
} else if (!strcmp(argv[argno], "--hexdump")) {
|
|
|
|
|
hexdump_g = TRUE;
|
[svn-r537] Changes since 19980722
----------------------
./src/H5A.c
./src/H5Apublic.h
./test/tattr.c
Switched the order of the second and third argument of
H5Aget_name() to make it consistent with other functions that
take buffers and buffer sizes.
./src/H5G.c
./src/H5Gpublic.h
./src/H5Gprivate.h
The H5Gget_comment() function returns the size of the comment
including the null terminator. If the object has no comment
then zero is returned. If an error occurs then a negative
value is returned.
./MANIFEST
./tools/Makefile.in
./tools/h5tools.h [NEW]
./tools/h5dump.c [NEW]
Created a library for printing values of datasets in a way
that looks nice. It's not done yet, but I needed it for
debugging the contents of files from Jim Reus.
./tools/h5ls.c
Added the `-d' and `--dump' options which cause the contents
of a dataset to be printed. Added `-w N' and `--width=N'
options to control how wide the raw data output should be. If
you want single-column output then say `-w1'.
Printing dataset values can now handle datasets of any integer
or floating point atomic type. As a special case, integers
which are one byte wide are treated a character strings for
now.
Sample output:
$ h5ls --dump --width=60 banana.hdf
ARCHIVE 0:0:0:744 Dataset {52/Inf}
Data:
(0) "U struct complex { double R; double I; };\012V"
(43) " double;\012"
U 0:0:0:2500 Dataset {256/512}
Data: printing of compound data types is not implemented yet
V 0:0:0:3928 Dataset {256/512}
Data:
(0) 0, 0.015625, 0.03125, 0.046875, 0.0625,
(5) 0.078125, 0.09375, 0.109375, 0.125, 0.140625,
(10) 0.15625, 0.171875, 0.1875, 0.203125, 0.21875,
(15) 0.234375, 0.25, 0.265625, 0.28125, 0.296875,
...
1998-07-24 05:19:17 +08:00
|
|
|
|
} else if (!strncmp(argv[argno], "-w", 2)) {
|
|
|
|
|
if (argv[argno][2]) {
|
|
|
|
|
s = argv[argno]+2;
|
|
|
|
|
} else if (argno+1>=argc) {
|
|
|
|
|
usage(progname);
|
|
|
|
|
exit(1);
|
|
|
|
|
} else {
|
|
|
|
|
s = argv[++argno];
|
|
|
|
|
}
|
1999-02-25 23:40:27 +08:00
|
|
|
|
width_g = (int)strtol(s, &rest, 0);
|
[svn-r537] Changes since 19980722
----------------------
./src/H5A.c
./src/H5Apublic.h
./test/tattr.c
Switched the order of the second and third argument of
H5Aget_name() to make it consistent with other functions that
take buffers and buffer sizes.
./src/H5G.c
./src/H5Gpublic.h
./src/H5Gprivate.h
The H5Gget_comment() function returns the size of the comment
including the null terminator. If the object has no comment
then zero is returned. If an error occurs then a negative
value is returned.
./MANIFEST
./tools/Makefile.in
./tools/h5tools.h [NEW]
./tools/h5dump.c [NEW]
Created a library for printing values of datasets in a way
that looks nice. It's not done yet, but I needed it for
debugging the contents of files from Jim Reus.
./tools/h5ls.c
Added the `-d' and `--dump' options which cause the contents
of a dataset to be printed. Added `-w N' and `--width=N'
options to control how wide the raw data output should be. If
you want single-column output then say `-w1'.
Printing dataset values can now handle datasets of any integer
or floating point atomic type. As a special case, integers
which are one byte wide are treated a character strings for
now.
Sample output:
$ h5ls --dump --width=60 banana.hdf
ARCHIVE 0:0:0:744 Dataset {52/Inf}
Data:
(0) "U struct complex { double R; double I; };\012V"
(43) " double;\012"
U 0:0:0:2500 Dataset {256/512}
Data: printing of compound data types is not implemented yet
V 0:0:0:3928 Dataset {256/512}
Data:
(0) 0, 0.015625, 0.03125, 0.046875, 0.0625,
(5) 0.078125, 0.09375, 0.109375, 0.125, 0.140625,
(10) 0.15625, 0.171875, 0.1875, 0.203125, 0.21875,
(15) 0.234375, 0.25, 0.265625, 0.28125, 0.296875,
...
1998-07-24 05:19:17 +08:00
|
|
|
|
if (width_g<=0 || *rest) {
|
|
|
|
|
usage(progname);
|
|
|
|
|
exit(1);
|
|
|
|
|
}
|
1998-07-18 03:03:43 +08:00
|
|
|
|
} else if ('-'!=argv[argno][1]) {
|
|
|
|
|
/* Single-letter switches */
|
|
|
|
|
for (s=argv[argno]+1; *s; s++) {
|
|
|
|
|
switch (*s) {
|
|
|
|
|
case '?':
|
|
|
|
|
case 'h': /* --help */
|
|
|
|
|
usage(progname);
|
|
|
|
|
exit(0);
|
1999-04-30 23:54:52 +08:00
|
|
|
|
case 'a': /* --address */
|
|
|
|
|
address_g = TRUE;
|
|
|
|
|
break;
|
|
|
|
|
case 'd': /* --data */
|
|
|
|
|
data_g = TRUE;
|
[svn-r537] Changes since 19980722
----------------------
./src/H5A.c
./src/H5Apublic.h
./test/tattr.c
Switched the order of the second and third argument of
H5Aget_name() to make it consistent with other functions that
take buffers and buffer sizes.
./src/H5G.c
./src/H5Gpublic.h
./src/H5Gprivate.h
The H5Gget_comment() function returns the size of the comment
including the null terminator. If the object has no comment
then zero is returned. If an error occurs then a negative
value is returned.
./MANIFEST
./tools/Makefile.in
./tools/h5tools.h [NEW]
./tools/h5dump.c [NEW]
Created a library for printing values of datasets in a way
that looks nice. It's not done yet, but I needed it for
debugging the contents of files from Jim Reus.
./tools/h5ls.c
Added the `-d' and `--dump' options which cause the contents
of a dataset to be printed. Added `-w N' and `--width=N'
options to control how wide the raw data output should be. If
you want single-column output then say `-w1'.
Printing dataset values can now handle datasets of any integer
or floating point atomic type. As a special case, integers
which are one byte wide are treated a character strings for
now.
Sample output:
$ h5ls --dump --width=60 banana.hdf
ARCHIVE 0:0:0:744 Dataset {52/Inf}
Data:
(0) "U struct complex { double R; double I; };\012V"
(43) " double;\012"
U 0:0:0:2500 Dataset {256/512}
Data: printing of compound data types is not implemented yet
V 0:0:0:3928 Dataset {256/512}
Data:
(0) 0, 0.015625, 0.03125, 0.046875, 0.0625,
(5) 0.078125, 0.09375, 0.109375, 0.125, 0.140625,
(10) 0.15625, 0.171875, 0.1875, 0.203125, 0.21875,
(15) 0.234375, 0.25, 0.265625, 0.28125, 0.296875,
...
1998-07-24 05:19:17 +08:00
|
|
|
|
break;
|
1999-08-24 20:52:10 +08:00
|
|
|
|
case 'e': /* --errors */
|
|
|
|
|
show_errors_g = TRUE;
|
|
|
|
|
break;
|
1999-01-22 02:33:39 +08:00
|
|
|
|
case 'f': /* --full */
|
|
|
|
|
fullname_g = TRUE;
|
|
|
|
|
break;
|
1999-04-16 03:57:50 +08:00
|
|
|
|
case 'g': /* --group */
|
|
|
|
|
grp_literal_g = TRUE;
|
|
|
|
|
break;
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
case 'l': /* --label */
|
|
|
|
|
label_g = TRUE;
|
|
|
|
|
break;
|
1999-01-22 02:33:39 +08:00
|
|
|
|
case 'r': /* --recursive */
|
|
|
|
|
recursive_g = TRUE;
|
|
|
|
|
fullname_g = TRUE;
|
|
|
|
|
break;
|
1999-10-01 00:14:14 +08:00
|
|
|
|
case 'S': /* --simple */
|
|
|
|
|
simple_output_g = TRUE;
|
|
|
|
|
break;
|
[svn-r879] Changes since 19981105
----------------------
./configure.in
./acconfig.h
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Added checks for functions, structs, constants, and header
files used in h5ls.c to determine the output width with
various Windows compilers and Unix variants.
Added check for <sys/stat.h> and defined
HAVE_SYS_STAT_H. Included <sys/stat.h> in H5private.h outside
the Posix.1 #include's section.
./src/H5RA.c
Less aggressive about failing -- rather returns false, which
allows type detection to continue.
./src/h5ls.c
Data types are displayed for datasets and named data types when
`-v' or `--verbose' is specified on the command-line. The
algorithm is a little different than the dumper because we're
trying to be human-friendly, not necessarily machine-friendly.
* Any data type which matches a native C type gets printed
something like `native double'.
* A floating point type that matches one of the IEEE standard
types but not one of the native types gets printed like `IEEE
64-bit big-endian float'.
* Other floating point values have information about sign bit
location; exponent size, location, and bias; and significand
size, location, and normalization.
* Padding and offsets are displayed for types that have
padding (precision != size), including internal padding for
some floating point data types.
* Non-native integer types are displayed like `32-bit
little-endian unsigned integer'.
* Compound data types have each member displayed including the
member name, byte offset within the struct, dimensions, index
permutation, and data type.
* String types are displayed like `256-byte null-terminated
ASCII string'.
* References are displayed like `8-byte unknown reference'
until the reference interface stabilizes a little.
* All other types including types not yet defined will be
printed like `4-byte class-9 unknown'.
The dimensionality of scalar datasets is printed like `{SCALAR}'
instead of just `{}'.
If external raw files are used to store a dataset then the offsets,
sizes, and file names of each are printed if `-v' or `--verbose'
was given on the command-line.
If an object is found and h5ls can't determine the object type then
it still tries to print the number of hard links, the OID, and any
comment that might be present if `-v' or `--verbose' was specified.
If the `-d' or `--dump' switch is turned on then ragged arrays will
report that the data can only be dumped by dumping the component
datasets explicitly. I'm not planning to implement this since
we're going to eventually change the whole way ragged arrays are
stored.
Compound data values do not have the component names displayed by
default when `-v' or `--verbose' is turned on. Instead, the names
can be displayed with `-l' or `--label'.
The output width is determined by the first rule that applies:
* If the `-wN', `-w N' or `--width=N' switch appeared on
the command line then use N for the output width.
* Query the OS for the tty width in a highly unportable way
borrowed from GNU `less' depending on what functions and
data structures were found during configuration (if any):
_getvideoconfig(), gettextinfo(), _srcsize(), ioctl(),
GetConsoleScreenBufferInfo(), struct videoconfig, struct
text_info, the TIOCGWINSZ ioctl, the TIOCGETD ioctl.
* If the `COLUMNS' environment variable is set then use
its value.
* Use the value 80.
Just for kicks, run Mark and Jim's test_vbt and then say `h5ls -dlsv
test.vbt'. You can also try it on the various *.h5 files in the
test/example directories.
./config/linux
Removed turning on parallel by default on Robb's macine.
1998-11-07 02:00:22 +08:00
|
|
|
|
case 's': /* --string */
|
[svn-r876] Changes since 19981102
----------------------
./bin/snapshot
Made same fix as for the release script yesterday.
./src/H5D.c
./src/H5Dprivate.h
./src/H5G.c
./src/H5Gprivate.h
./src/H5Gpublic.h
./src/H5O.c
./src/H5Oprivate.h
./src/H5RA.c
./src/H5RAprivate.h
./src/H5T.c
./src/H5Tprivate.h
Improved object type checking. Instead of determining the
object type by trying to open each of the possible types, we
keep a table of associations between object type number (like
H5G_GROUP, H5G_DATASET, H5D_TYPE, and H5D_RAGGED) and an `isa'
function that returns true if the object header has the right
messages to make the object a particular type. This mechanism
also allows specialization of object types by permitting an
object to satisfy more than one `isa' function.
Added `isa' functions for groups, datasets, ragged arrays, and
committed data types.
./src/H5config.h.in
Added HAVE_STAT_ST_BLOCKS. I thought this had already been
added, but apparently not.
./tools/h5ls.c
Removed system include files since they're already included by
H5private.h and since I wasn't including them portably anyway.
By default, 1-byte integer types are printed as integer values
instead of ASCII characters. However, the `-s' or `--string'
command-line switch causes the data to be interpretted as
ASCII. String data types are always printed as character
data.
Ragged arrays are now identified as ragged arrays and h5ls
doesn't descend into the group automatically. This uses the
new object type specialization stuff.
./tools/h5tools.c
./tools/h5tools.h
Added the ability to print 1-byte integer types as either
ASCII or numeric data instead of always ASCII. The default is
to print as numeric data.
1998-11-06 04:28:34 +08:00
|
|
|
|
string_g = TRUE;
|
|
|
|
|
break;
|
1998-07-18 03:03:43 +08:00
|
|
|
|
case 'v': /* --verbose */
|
|
|
|
|
verbose_g++;
|
|
|
|
|
break;
|
|
|
|
|
case 'V': /* --version */
|
2000-04-22 03:28:09 +08:00
|
|
|
|
print_version(progname);
|
1998-07-18 03:03:43 +08:00
|
|
|
|
exit(0);
|
1999-06-07 23:05:02 +08:00
|
|
|
|
case 'x': /* --hexdump */
|
|
|
|
|
hexdump_g = TRUE;
|
1999-06-07 23:58:05 +08:00
|
|
|
|
break;
|
1998-07-18 03:03:43 +08:00
|
|
|
|
default:
|
|
|
|
|
usage(progname);
|
|
|
|
|
exit(1);
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
} else {
|
|
|
|
|
usage(progname);
|
|
|
|
|
exit(1);
|
|
|
|
|
}
|
|
|
|
|
}
|
1998-03-25 07:18:34 +08:00
|
|
|
|
|
1998-08-28 23:24:12 +08:00
|
|
|
|
/*
|
1999-06-14 23:07:58 +08:00
|
|
|
|
* If no arguments remain then print a usage message (instead of doing
|
|
|
|
|
* absolutely nothing ;-)
|
1998-08-28 23:24:12 +08:00
|
|
|
|
*/
|
1999-06-14 23:07:58 +08:00
|
|
|
|
if (argno>=argc) {
|
1998-07-18 03:03:43 +08:00
|
|
|
|
usage(progname);
|
|
|
|
|
exit(1);
|
|
|
|
|
}
|
1999-08-24 20:52:10 +08:00
|
|
|
|
|
|
|
|
|
/* Turn off HDF5's automatic error printing unless you're debugging h5ls */
|
|
|
|
|
if (!show_errors_g) H5Eset_auto(NULL, NULL);
|
|
|
|
|
|
|
|
|
|
|
1998-08-28 23:24:12 +08:00
|
|
|
|
/*
|
1999-06-14 23:07:58 +08:00
|
|
|
|
* Each remaining argument is an hdf5 file followed by an optional slash
|
|
|
|
|
* and object name.
|
|
|
|
|
*
|
|
|
|
|
* Example: ../dir1/foo/bar/baz
|
|
|
|
|
* \_________/\______/
|
|
|
|
|
* file obj
|
|
|
|
|
*
|
|
|
|
|
* The dichotomy is determined by calling H5Fopen() repeatedly until it
|
|
|
|
|
* succeeds. The first call uses the entire name and each subsequent call
|
|
|
|
|
* chops off the last component. If we reach the beginning of the name
|
|
|
|
|
* then there must have been something wrong with the file (perhaps it
|
|
|
|
|
* doesn't exist).
|
1998-08-28 23:24:12 +08:00
|
|
|
|
*/
|
1999-06-14 23:07:58 +08:00
|
|
|
|
while (argno<argc) {
|
|
|
|
|
fname = argv[argno++];
|
|
|
|
|
oname = NULL;
|
|
|
|
|
file = -1;
|
|
|
|
|
|
|
|
|
|
while (fname && *fname) {
|
2000-06-24 01:52:21 +08:00
|
|
|
|
file = h5dump_fopen(fname, drivername, sizeof drivername);
|
1999-08-24 20:52:10 +08:00
|
|
|
|
if (file>=0) {
|
|
|
|
|
if (verbose_g) {
|
|
|
|
|
printf("Opened \"%s\" with %s driver.\n",
|
2000-06-07 02:07:20 +08:00
|
|
|
|
fname, drivername);
|
1999-08-24 20:52:10 +08:00
|
|
|
|
}
|
|
|
|
|
break; /*success*/
|
|
|
|
|
}
|
|
|
|
|
|
1999-06-14 23:07:58 +08:00
|
|
|
|
/* Shorten the file name; lengthen the object name */
|
|
|
|
|
x = oname;
|
|
|
|
|
oname = strrchr(fname, '/');
|
|
|
|
|
if (x) *x = '/';
|
|
|
|
|
if (!oname) break;
|
|
|
|
|
*oname = '\0';
|
1999-04-16 03:57:50 +08:00
|
|
|
|
}
|
1999-06-14 23:07:58 +08:00
|
|
|
|
if (file<0) {
|
1999-10-08 00:51:49 +08:00
|
|
|
|
fprintf(stderr, "%s: unable to open file\n", argv[argno-1]);
|
1999-06-14 23:07:58 +08:00
|
|
|
|
}
|
|
|
|
|
if (oname) oname++;
|
|
|
|
|
if (!oname || !*oname) oname = root_name;
|
|
|
|
|
|
|
|
|
|
/* Open the object and display it's information */
|
|
|
|
|
if (H5Gget_objinfo(file, oname, TRUE, &sb)>=0 &&
|
|
|
|
|
H5G_GROUP==sb.type && !grp_literal_g) {
|
|
|
|
|
/*
|
|
|
|
|
* Specified name is a group. List the complete contents of the
|
|
|
|
|
* group.
|
|
|
|
|
*/
|
|
|
|
|
sym_insert(&sb, oname);
|
|
|
|
|
#ifdef H5LS_PREPEND_FILENAME
|
|
|
|
|
iter.container = container = fix_name(fname, oname);
|
|
|
|
|
#else
|
|
|
|
|
iter.container = container = fix_name("", oname);
|
|
|
|
|
#endif
|
|
|
|
|
H5Giterate(file, oname, NULL, list, &iter);
|
|
|
|
|
free(container);
|
|
|
|
|
|
|
|
|
|
} else if ((root=H5Gopen(file, "/"))<0) {
|
|
|
|
|
exit(1); /*major problem!*/
|
1999-01-22 02:33:39 +08:00
|
|
|
|
|
1999-06-14 23:07:58 +08:00
|
|
|
|
} else {
|
|
|
|
|
/*
|
|
|
|
|
* Specified name is a non-group object -- list that object. The
|
|
|
|
|
* container for the object is everything up to the base name.
|
|
|
|
|
*/
|
|
|
|
|
#ifdef H5LS_PREPEND_FILENAME
|
|
|
|
|
iter.container = fname;
|
|
|
|
|
#else
|
|
|
|
|
iter.container = "/";
|
|
|
|
|
#endif
|
|
|
|
|
list(root, oname, &iter);
|
|
|
|
|
if (H5Gclose(root)<0) exit(1);
|
1998-08-28 23:24:12 +08:00
|
|
|
|
}
|
1999-06-14 23:07:58 +08:00
|
|
|
|
H5Fclose(file);
|
1998-08-28 23:24:12 +08:00
|
|
|
|
}
|
1999-06-14 23:07:58 +08:00
|
|
|
|
|
1998-03-25 07:18:34 +08:00
|
|
|
|
return 0;
|
|
|
|
|
}
|