netcdf-c/nc_test4/tst_h5_endians.c
Dennis Heimbigner df3636b959 Mitigate S3 test interference + Unlimited Dimensions in NCZarr
This PR started as an attempt to add unlimited dimensions to NCZarr.
It did that, but this exposed significant problems with test interference.
So this PR is mostly about fixing -- well mitigating anyway -- test
interference.

The problem of test interference is now documented in the document docs/internal.md.
The solutions implemented here are also describe in that document.
The solution is somewhat fragile but multiple cleanup mechanisms
are provided. Note that this feature requires that the
AWS command line utility must be installed.

## Unlimited Dimensions.
The existing NCZarr extensions to Zarr are modified to support unlimited dimensions.
NCzarr extends the Zarr meta-data for the ".zgroup" object to include netcdf-4 model extensions. This information is stored in ".zgroup" as dictionary named "_nczarr_group".
Inside "_nczarr_group", there is a key named "dims" that stores information about netcdf-4 named dimensions. The value of "dims" is a dictionary whose keys are the named dimensions. The value associated with each dimension name has one of two forms
Form 1 is a special case of form 2, and is kept for backward compatibility. Whenever a new file is written, it uses format 1 if possible, otherwise format 2.
* Form 1: An integer representing the size of the dimension, which is used for simple named dimensions.
* Form 2: A dictionary with the following keys and values"
   - "size" with an integer value representing the (current) size of the dimension.
   - "unlimited" with a value of either "1" or "0" to indicate if this dimension is an unlimited dimension.

For Unlimited dimensions, the size is initially zero, and as variables extend the length of that dimension, the size value for the dimension increases.
That dimension size is shared by all arrays referencing that dimension, so if one array extends an unlimited dimension, it is implicitly extended for all other arrays that reference that dimension.
This is the standard semantics for unlimited dimensions.

Adding unlimited dimensions required a number of other changes to the NCZarr code-base. These included the following.
* Did a partial refactor of the slice handling code in zwalk.c to clean it up.
* Added a number of tests for unlimited dimensions derived from the same test in nc_test4.
* Added several NCZarr specific unlimited tests; more are needed.
* Add test of endianness.

## Misc. Other Changes
* Modify libdispatch/ncs3sdk_aws.cpp to optionally support use of the
   AWS Transfer Utility mechanism. This is controlled by the
   ```#define TRANSFER```` command in that file. It defaults to being disabled.
* Parameterize both the standard Unidata S3 bucket (S3TESTBUCKET) and the netcdf-c test data prefix (S3TESTSUBTREE).
* Fixed an obscure memory leak in ncdump.
* Removed some obsolete unit testing code and test cases.
* Uncovered a bug in the netcdf-c handling of big-endian floats and doubles. Have not fixed yet. See tst_h5_endians.c.
* Renamed some nczarr_tests testcases to avoid name conflicts with nc_test4.
* Modify the semantics of zmap\#ncsmap_write to only allow total rewrite of objects.
* Modify the semantics of zodom to properly handle stride > 1.
* Add a truncate operation to the libnczarr zmap code.
2023-09-26 16:56:48 -06:00

338 lines
10 KiB
C

/*! Test for NCF-331. Added May 11, 2015.
* Copyright 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002,
* 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011, 2012, 2013, 2014,
* 2015, 2016, 2017, 2018
* University Corporation for Atmospheric Research/Unidata.
*
* See \ref copyright file for more info.
* See the following links for more information:
*
* o Issue on GitHub: https://github.com/Unidata/netcdf-c/issues/112
* o Issue in JIRA: https://bugtracking.unidata.ucar.edu/browse/NCF-331
*
* Test contributed by Jeff Whitaker
*/
#include <nc_tests.h>
#include <string.h>
#include "nc_logging.h"
#undef BE_DEBUG
#ifdef TESTNCZARR
#include "test_utils.h"
#define FILE_NAME_NC "file://tmp_h5_endians.file#mode=nczarr,file"
#else
#define FILE_NAME_NC "tst_h5_endians.nc"
#endif
#define NDIM 10
#define NLON 20
#define DIM_NAME "x"
#define DIM_LEN 4
#define GRP_NAME "grp"
#define LE_FLOAT_VARNAME "fl_le"
#define BE_FLOAT_VARNAME "fl_be"
#define LE_INT_VARNAME "int_le"
#define BE_INT_VARNAME "int_be"
#define LE_DBL_VARNAME "dbl_le"
#define BE_DBL_VARNAME "dbl_be"
#if defined BE_DEBUG || defined TESTNCZARR
static float
f32swap(float x)
{
union {
unsigned char bytes[4];
float f;
} u;
unsigned char c;
u.f = x;
c = u.bytes[0]; u.bytes[0] = u.bytes[3]; u.bytes[3] = c;
c = u.bytes[1]; u.bytes[1] = u.bytes[2]; u.bytes[2] = c;
return u.f;
}
static double
f64swap(double x)
{
union {
unsigned char bytes[8];
double d;
} u;
unsigned char c;
u.d = x;
c = u.bytes[0]; u.bytes[0] = u.bytes[7]; u.bytes[7] = c;
c = u.bytes[1]; u.bytes[1] = u.bytes[6]; u.bytes[6] = c;
c = u.bytes[2]; u.bytes[2] = u.bytes[5]; u.bytes[5] = c;
c = u.bytes[3]; u.bytes[3] = u.bytes[4]; u.bytes[4] = c;
return u.d;
}
#endif
int main() {
int ncid, dimid;
int le_float_varid;
int be_float_varid;
int le_int_varid;
int be_int_varid;
int le_dbl_varid;
int be_dbl_varid;
int ed;
int failures = 0;
int retval;
printf("* Checking that endianness is properly read from file.\n");
printf("** Generating test files.\n");
/*
* 1. Create a netcdf file with endianness as desired.
*/
{
printf("*** Creating a file via netcdf API: %s.\n",FILE_NAME_NC);
if ((retval = nc_create(FILE_NAME_NC, NC_NETCDF4 | NC_CLOBBER, &ncid)))
return retval;
if ((retval = nc_def_dim(ncid, DIM_NAME, NDIM, &dimid)))
return retval;
/* Little-Endian Float */
if ((retval = nc_def_var(ncid, LE_FLOAT_VARNAME, NC_FLOAT, 1, &dimid, &le_float_varid)))
return retval;
if ((retval = nc_def_var_endian(ncid, le_float_varid, NC_ENDIAN_LITTLE)))
return retval;
/* Big-Endian Float */
if ((retval = nc_def_var(ncid, BE_FLOAT_VARNAME, NC_FLOAT, 1, &dimid, &be_float_varid)))
return retval;
if ((retval = nc_def_var_endian(ncid, be_float_varid, NC_ENDIAN_BIG)))
return retval;
/* Little-Endian Int */
if ((retval = nc_def_var(ncid, LE_INT_VARNAME, NC_INT, 1, &dimid, &le_int_varid)))
return retval;
if ((retval = nc_def_var_endian(ncid, le_int_varid, NC_ENDIAN_LITTLE)))
return retval;
/* Big-Endian Int */
if ((retval = nc_def_var(ncid, BE_INT_VARNAME, NC_INT, 1, &dimid, &be_int_varid)))
return retval;
if ((retval = nc_def_var_endian(ncid, be_int_varid, NC_ENDIAN_BIG)))
return retval;
/* Little-Endian Double */
if ((retval = nc_def_var(ncid, LE_DBL_VARNAME, NC_DOUBLE, 1, &dimid, &le_dbl_varid)))
return retval;
if ((retval = nc_def_var_endian(ncid, le_dbl_varid, NC_ENDIAN_LITTLE)))
return retval;
/* Big-Endian Double */
if ((retval = nc_def_var(ncid, BE_DBL_VARNAME, NC_DOUBLE, 1, &dimid, &be_dbl_varid)))
return retval;
if ((retval = nc_def_var_endian(ncid, be_dbl_varid, NC_ENDIAN_BIG)))
return retval;
if ((retval = nc_close(ncid)))
return retval;
}
/*
* 2. Reopen netcdf-generated file, check to see if the endianness attribute
* exists.
*/
printf("** Checking test files.\n");
{
ncid = 0;
le_float_varid = 0;
be_float_varid = 0;
le_int_varid = 0;
be_int_varid = 0;
le_dbl_varid = 0;
be_dbl_varid = 0;
printf("*** %s\n",FILE_NAME_NC);
if ((retval = nc_open(FILE_NAME_NC, NC_NETCDF4 | NC_NOWRITE, &ncid)))
return retval;
if ((retval = nc_inq_varid(ncid,LE_FLOAT_VARNAME,&le_float_varid)))
return retval;
if ((retval = nc_inq_varid(ncid,BE_FLOAT_VARNAME,&be_float_varid)))
return retval;
if ((retval = nc_inq_varid(ncid,LE_INT_VARNAME,&le_int_varid)))
return retval;
if ((retval = nc_inq_varid(ncid,BE_INT_VARNAME,&be_int_varid)))
return retval;
if ((retval = nc_inq_varid(ncid,LE_DBL_VARNAME,&le_dbl_varid)))
return retval;
if ((retval = nc_inq_varid(ncid,BE_DBL_VARNAME,&be_dbl_varid)))
return retval;
printf("\tLittle-Endian Float...\t");
if ((retval = nc_inq_var_endian(ncid,le_float_varid,&ed)))
return retval;
if(ed == NC_ENDIAN_LITTLE) printf("passed\n"); else {printf("failed\n"); failures++;}
printf("\tBig-Endian Float...\t");
if ((retval = nc_inq_var_endian(ncid,be_float_varid,&ed)))
return retval;
if(ed == NC_ENDIAN_BIG) printf("passed\n"); else {printf("failed\n"); failures++;}
printf("\tLittle-Endian Int...\t");
if ((retval = nc_inq_var_endian(ncid,le_int_varid,&ed)))
return retval;
if(ed == NC_ENDIAN_LITTLE) printf("passed\n"); else {printf("failed\n"); failures++;}
printf("\tBig-Endian Int...\t");
if ((retval = nc_inq_var_endian(ncid,be_int_varid,&ed)))
return retval;
if(ed == NC_ENDIAN_BIG) printf("passed\n"); else {printf("failed\n"); failures++;}
printf("\tLittle-Endian Double...\t");
if ((retval = nc_inq_var_endian(ncid,le_dbl_varid,&ed)))
return retval;
if(ed == NC_ENDIAN_LITTLE) printf("passed\n"); else {printf("failed\n"); failures++;}
printf("\tBig-Endian Double...\t");
if ((retval = nc_inq_var_endian(ncid,be_dbl_varid,&ed)))
return retval;
if(ed == NC_ENDIAN_BIG) printf("passed\n"); else {printf("failed\n"); failures++;}
if ((retval = nc_close(ncid)))
return retval;
}
/*
* 3. Reopen netcdf-generated file, write data and reread.
*/
printf("** Read/Write test files.\n");
{
ncid = 0;
le_float_varid = 0;
be_float_varid = 0;
le_int_varid = 0;
be_int_varid = 0;
le_dbl_varid = 0;
be_dbl_varid = 0;
printf("*** %s\n",FILE_NAME_NC);
if ((retval = nc_open(FILE_NAME_NC, NC_NETCDF4 | NC_WRITE, &ncid)))
return retval;
if ((retval = nc_inq_varid(ncid,LE_FLOAT_VARNAME,&le_float_varid)))
return retval;
if ((retval = nc_inq_varid(ncid,BE_FLOAT_VARNAME,&be_float_varid)))
return retval;
if ((retval = nc_inq_varid(ncid,LE_INT_VARNAME,&le_int_varid)))
return retval;
if ((retval = nc_inq_varid(ncid,BE_INT_VARNAME,&be_int_varid)))
return retval;
if ((retval = nc_inq_varid(ncid,LE_DBL_VARNAME,&le_dbl_varid)))
return retval;
if ((retval = nc_inq_varid(ncid,BE_DBL_VARNAME,&be_dbl_varid)))
return retval;
{
int i, failed;
int idata_in[NDIM];
float fdata_in[NDIM];
double ddata_in[NDIM];
int idata_le_out[NDIM];
float fdata_le_out[NDIM];
double ddata_le_out[NDIM];
int idata_be_out[NDIM];
#if defined BE_DEBUG || defined TESTNCZARR
float fdata_be_out[NDIM];
double ddata_be_out[NDIM];
#endif
/* Setup data in/out */
for(i=0;i<NDIM;i++) {
idata_in[i] = (int)i;
fdata_in[i] = (float)i;
ddata_in[i] = (double)i;
idata_le_out[i] = 0;
fdata_le_out[i] = 0.0f;
ddata_le_out[i] = 0.0;
idata_be_out[i] = 0;
#if defined BE_DEBUG || defined TESTNCZARR
fdata_be_out[i] = 0.0f;
ddata_be_out[i] = 0.0;
#endif
}
printf("\tLittle-Endian Float...\t");
if ((retval = nc_put_var(ncid,le_float_varid,fdata_in)))
return retval;
if ((retval = nc_get_var(ncid,le_float_varid,fdata_le_out)))
return retval;
for(failed=0,i=0;i<NDIM;i++) {if(fdata_in[i] != fdata_le_out[i]) {printf("failed\n"); failures++; failed++; break;}}
if(!failed) printf("passed\n");
#if defined BE_DEBUG || defined TESTNCZARR
/* There appears to be a bug in the handling of big-endian floats and doubles */
printf("\tBig-Endian Float...\t");
if ((retval = nc_put_var(ncid,be_float_varid,fdata_in)))
return retval;
if ((retval = nc_get_var(ncid,be_float_varid,fdata_be_out)))
return retval;
for(i=0;i<NDIM;i++) {
float f = f32swap(fdata_be_out[i]);
fprintf(stderr,"[%d] %f\n",i,f);
}
for(failed=0,i=0;i<NDIM;i++) {
if(fdata_in[i] != fdata_be_out[i]) {
printf("failed\n"); failures++; failed++; break;
}
}
if(!failed) printf("passed\n");
#endif
printf("\tLittle-Endian Int...\t");
if ((retval = nc_put_var(ncid,le_int_varid,idata_in)))
return retval;
if ((retval = nc_get_var(ncid,le_int_varid,idata_le_out)))
return retval;
for(failed=0,i=0;i<NDIM;i++) {if(idata_in[i] != idata_le_out[i]) {printf("failed\n"); failures++; failed++; break;}}
if(!failed) printf("passed\n");
printf("\tBig-Endian Int...\t");
if ((retval = nc_put_var(ncid,be_int_varid,idata_in)))
return retval;
if ((retval = nc_get_var(ncid,be_int_varid,idata_be_out)))
for(failed=0,i=0;i<NDIM;i++) {if(idata_in[i] != idata_be_out[i]) {printf("failed\n"); failures++; failed++; break;}}
if(!failed) printf("passed\n");
printf("\tLittle-Endian Double...\t");
if ((retval = nc_put_var(ncid,le_dbl_varid,ddata_in)))
return retval;
if ((retval = nc_get_var(ncid,le_dbl_varid,ddata_le_out)))
return retval;
for(failed=0,i=0;i<NDIM;i++) {if(ddata_in[i] != ddata_le_out[i]) {printf("failed\n"); failures++; failed++; break;}}
if(!failed) printf("passed\n");
#if defined BE_DEBUG || defined TESTNCZARR
/* There appears to be a bug in the handling of big-endian floats and doubles */
printf("\tBig-Endian Double...\t");
if ((retval = nc_put_var(ncid,be_dbl_varid,ddata_in)))
return retval;
if ((retval = nc_get_var(ncid,be_dbl_varid,ddata_be_out)))
return retval;
for(i=0;i<NDIM;i++) {
double d = f64swap(ddata_be_out[i]);
fprintf(stderr,"[%d] %lf\n",i,d);
}
for(failed=0,i=0;i<NDIM;i++) {if(ddata_in[i] != ddata_be_out[i]) {printf("failed\n"); failures++; failed++; break;}}
if(!failed) printf("passed\n");
#endif
if ((retval = nc_close(ncid)))
return retval;
}
}
printf("** Failures Returned: [%d]\n",failures);
return failures;
}