netcdf-c/ncgen/util.c
Dennis Heimbigner 11fe00ea05 Add filter support to NCZarr
Filter support has three goals:

1. Use the existing HDF5 filter implementations,
2. Allow filter metadata to be stored in the NumCodecs metadata format used by Zarr,
3. Allow filters to be used even when HDF5 is disabled

Detailed usage directions are define in docs/filters.md.

For now, the existing filter API is left in place. So filters
are defined using ''nc_def_var_filter'' using the HDF5 style
where the id and parameters are unsigned integers.

This is a big change since filters affect many parts of the code.

In the following, the terms "compressor" and "filter" and "codec" are generally
used synonomously.

### Filter-Related Changes:
* In order to support dynamic loading of shared filter libraries, a new library was added in the libncpoco directory; it helps to isolate dynamic loading across multiple platforms.
* Provide a json parsing library for use by plugins; this is created by merging libdispatch/ncjson.c with include/ncjson.h.
* Add a new _Codecs attribute to allow clients to see what codecs are being used; let ncdump -s print it out.
* Provide special headers to help support compilation of HDF5 filters when HDF5 is not enabled: netcdf_filter_hdf5_build.h and netcdf_filter_build.h.
* Add a number of new test to test the new nczarr filters.
* Let ncgen parse _Codecs attribute, although it is ignored.

### Plugin directory changes:
* Add support for the Blosc compressor; this is essential because it is the most common compressor used in Zarr datasets. This also necessitated adding a CMake FindBlosc.cmake file
* Add NCZarr support for the big-four filters provided by HDF5: shuffle, fletcher32, deflate (zlib), and szip
* Add a Codec defaulter (see docs/filters.md) for the big four filters.
* Make plugins work with windows by properly adding __declspec declaration.

### Misc. Non-Filter Changes
* Replace most uses of USE_NETCDF4 (deprecated) with USE_HDF5.
* Improve support for caching
* More fixes for path conversion code
* Fix misc. memory leaks
* Add new utility -- ncdump/ncpathcvt -- that does more or less the same thing as cygpath.
* Add a number of new test to test the non-filter fixes.
* Update the parsers
* Convert most instances of '#ifdef _MSC_VER' to '#ifdef _WIN32'
2021-09-02 17:04:26 -06:00

672 lines
14 KiB
C

/*********************************************************************
* Copyright 2018, UCAR/Unidata
* See netcdf/COPYRIGHT file for copying and redistribution conditions.
* $Header: /upc/share/CVS/netcdf-3/ncgen/util.c,v 1.4 2010/04/14 22:04:59 dmh Exp $
*********************************************************************/
#include "includes.h"
/* Track primitive symbol instances (initialized in ncgen.y) */
Symbol* primsymbols[PRIMNO];
char*
append(const char* s1, const char* s2)
{
int len = (s1?strlen(s1):0)+(s2?strlen(s2):0);
char* result = (char*)ecalloc(len+1);
result[0] = '\0';
if(s1) strcat(result,s1);
if(s2) strcat(result,s2);
return result;
}
unsigned int
chartohex(char c)
{
switch (c) {
case '0': case '1': case '2': case '3': case '4':
case '5': case '6': case '7': case '8': case '9':
return (c - '0');
case 'A': case 'B': case 'C':
case 'D': case 'E': case 'F':
return (c - 'A') + 0x0a;
case 'a': case 'b': case 'c':
case 'd': case 'e': case 'f':
return (c - 'a') + 0x0a;
}
return 0;
}
/*
* For generated Fortran, change 'e' to 'd' in exponent of double precision
* constants.
*/
void
expe2d(
char *cp) /* string containing double constant */
{
char *expchar = strrchr(cp,'e');
if (expchar) {
*expchar = 'd';
}
}
/* Returns non-zero if n is a power of 2, 0 otherwise */
int
pow2(
int n)
{
int m = n;
int p = 1;
while (m > 0) {
m /= 2;
p *= 2;
}
return p == 2*n;
}
/*
* Remove trailing zeros (after decimal point) but not trailing decimal
* point from ss, a string representation of a floating-point number that
* might include an exponent part.
*/
void
tztrim(
char *ss /* returned string representing dd */
)
{
char *cp, *ep;
cp = ss;
if (*cp == '-')
cp++;
while(isdigit((int)*cp) || *cp == '.')
cp++;
if (*--cp == '.')
return;
ep = cp+1;
while (*cp == '0')
cp--;
cp++;
if (cp == ep)
return;
while (*ep)
*cp++ = *ep++;
*cp = '\0';
return;
}
static void
clearSpecialdata(Specialdata* data)
{
if(data == NULL) return;
reclaimdatalist(data->_Fillvalue);
if(data->_ChunkSizes)
efree(data->_ChunkSizes);
if(data->_Filters) {
int i;
for(i=0;i<data->nfilters;i++) {
NC_H5_Filterspec* f = data->_Filters[i];
ncaux_h5filterspec_free(f);
}
efree(data->_Filters);
}
if(data->_Codecs)
efree(data->_Codecs);
}
void
freeSymbol(Symbol* sym)
{
if(sym == NULL) return;
switch (sym->objectclass) {
case NC_VAR:
clearSpecialdata(&sym->var.special);
listfree(sym->var.attributes);
break;
case NC_TYPE:
if(sym->typ.econst)
reclaimconstant(sym->typ.econst);
if(sym->typ._Fillvalue)
reclaimdatalist(sym->typ._Fillvalue);
break;
case NC_GRP:
if(sym->file.filename)
efree(sym->file.filename);
break;
default: break;
}
/* Universal */
if(sym->name) efree(sym->name);
if(sym->fqn) efree(sym->fqn);
listfree(sym->prefix);
if(sym->data)
reclaimdatalist(sym->data);
listfree(sym->subnodes);
efree(sym);
}
char* nctypenames[17] = {
"NC_NAT",
"NC_BYTE", "NC_CHAR", "NC_SHORT", "NC_INT",
"NC_FLOAT", "NC_DOUBLE",
"NC_UBYTE", "NC_USHORT", "NC_UINT",
"NC_INT64", "NC_UINT64",
"NC_STRING",
"NC_VLEN", "NC_OPAQUE", "NC_ENUM", "NC_COMPOUND"
};
char* nctypenamesextend[9] = {
"NC_GRP", "NC_DIM", "NC_VAR", "NC_ATT", "NC_TYPE",
"NC_ECONST","NC_FIELD", "NC_ARRAY","NC_PRIM"
};
char*
nctypename(nc_type nctype)
{
char* s;
if(nctype >= NC_NAT && nctype <= NC_COMPOUND)
return nctypenames[nctype];
if(nctype >= NC_GRP && nctype <= NC_PRIM)
return nctypenamesextend[(nctype - NC_GRP)];
if(nctype == NC_FILLVALUE) return "NC_FILL";
if(nctype == NC_NIL) return "NC_NIL";
s = poolalloc(128);
sprintf(s,"NC_<%d>",nctype);
return s;
}
/* These are the augmented NC_ values (0 based from NC_GRP)*/
char* ncclassnames[9] = {
"NC_GRP", "NC_DIM", "NC_VAR", "NC_ATT",
"NC_TYP", "NC_ECONST", "NC_FIELD", "NC_ARRAY",
"NC_PRIM"
};
char*
ncclassname(nc_class ncc)
{
char* s;
if(ncc >= NC_NAT && ncc <= NC_COMPOUND)
return nctypename((nc_type)ncc);
if(ncc == NC_FILLVALUE) return "NC_FILL";
if(ncc >= NC_GRP && ncc <= NC_PRIM)
return ncclassnames[ncc - NC_GRP];
s = poolalloc(128);
sprintf(s,"NC_<%d>",ncc);
return s;
}
int ncsizes[17] = {
0,
1,1,2,4,
4,8,
1,2,4,
8,8,
sizeof(char*),
sizeof(nc_vlen_t),
0,0,0
};
int
ncsize(nc_type nctype)
{
if(nctype >= NC_NAT && nctype <= NC_COMPOUND)
return ncsizes[nctype];
return 0;
}
int
hasunlimited(Dimset* dimset)
{
int i;
for(i=0;i<dimset->ndims;i++) {
Symbol* dim = dimset->dimsyms[i];
if(dim->dim.declsize == NC_UNLIMITED) return 1;
}
return 0;
}
/* return 1 if first dimension is unlimited*/
int
isunlimited0(Dimset* dimset)
{
return (dimset->ndims > 0 && dimset->dimsyms[0]->dim.declsize == NC_UNLIMITED);
}
/* True only if dim[0] is unlimited all rest are bounded*/
/* or all are bounded*/
int
classicunlimited(Dimset* dimset)
{
int i;
int last = -1;
for(i=0;i<dimset->ndims;i++) {
Symbol* dim = dimset->dimsyms[i];
if(dim->dim.declsize == NC_UNLIMITED) last = i;
}
return (last < 1);
}
/* True only iff no dimension is unlimited*/
int
isbounded(Dimset* dimset)
{
int i;
for(i=0;i<dimset->ndims;i++) {
Symbol* dim = dimset->dimsyms[i];
if(dim->dim.declsize == NC_UNLIMITED) return 0;
}
return 1;
}
int
signedtype(nc_type nctype)
{
switch (nctype) {
case NC_BYTE:
case NC_SHORT:
case NC_INT:
case NC_INT64:
return nctype;
case NC_UBYTE: return NC_BYTE;
case NC_USHORT: return NC_SHORT;
case NC_UINT: return NC_INT;
case NC_UINT64: return NC_INT64;
default: break;
}
return nctype;
}
int
unsignedtype(nc_type nctype)
{
switch (nctype) {
case NC_UBYTE:
case NC_USHORT:
case NC_UINT:
case NC_UINT64:
return nctype;
case NC_BYTE: return NC_UBYTE;
case NC_SHORT: return NC_USHORT;
case NC_INT: return NC_UINT;
case NC_INT64: return NC_UINT64;
default: break;
}
return nctype;
}
int
isinttype(nc_type nctype)
{
return (nctype != NC_CHAR)
&& ((nctype >= NC_BYTE && nctype <= NC_INT)
|| (nctype >= NC_UBYTE && nctype <= NC_UINT64));
}
int
isuinttype(nc_type t)
{
return isinttype(t)
&& t >= NC_UBYTE
&& t <= NC_UINT64
&& t != NC_INT64;
}
int
isfloattype(nc_type nctype)
{
return (nctype == NC_FLOAT || nctype <= NC_DOUBLE);
}
int
isclassicprim(nc_type nctype)
{
return (nctype >= NC_BYTE && nctype <= NC_DOUBLE)
;
}
int
isclassicprimplus(nc_type nctype)
{
return (nctype >= NC_BYTE && nctype <= NC_DOUBLE)
|| (nctype == NC_STRING)
;
}
int
isprim(nc_type nctype)
{
return (nctype >= NC_BYTE && nctype <= NC_STRING)
;
}
int
isprimplus(nc_type nctype)
{
return (nctype >= NC_BYTE && nctype <= NC_STRING)
|| (nctype == NC_ECONST)
|| (nctype == NC_OPAQUE)
;
}
void
collectpath(Symbol* grp, List* grpstack)
{
while(grp != NULL) {
listpush(grpstack,(void*)grp);
grp = grp->container;
}
}
#ifdef USE_NETCDF4
/* Result is pool'd*/
char*
prefixtostring(List* prefix, char* separator)
{
int slen=0;
int plen;
int i;
char* result;
if(prefix == NULL) return pooldup("");
plen = prefixlen(prefix);
if(plen == 0) { /* root prefix*/
slen=0;
/* slen += strlen(separator);*/
slen++; /* for null terminator*/
result = poolalloc(slen);
result[0] = '\0';
/*strcat(result,separator);*/
} else {
for(i=0;i<plen;i++) {
Symbol* sym = (Symbol*)listget(prefix,i);
slen += (strlen(separator)+strlen(sym->name));
}
slen++; /* for null terminator*/
result = poolalloc(slen);
result[0] = '\0';
for(i=0;i<plen;i++) {
Symbol* sym = (Symbol*)listget(prefix,i);
strcat(result,separator);
strcat(result,sym->name); /* append "/<prefix[i]>"*/
}
}
return result;
}
#endif
/* Result is pool'd*/
char*
fullname(Symbol* sym)
{
#ifdef USE_NETCDF4
char* s1;
char* result;
char* prefix;
prefix = prefixtostring(sym->prefix,PATHSEPARATOR);
s1 = poolcat(prefix,PATHSEPARATOR);
result = poolcat(s1,sym->name);
return result;
#else
return nulldup(sym->name);
#endif
}
int
prefixeq(List* x1, List* x2)
{
Symbol** l1;
Symbol** l2;
int len,i;
if((len=listlength(x1)) != listlength(x2)) return 0;
l1=(Symbol**)listcontents(x1);
l2=(Symbol**)listcontents(x2);
for(i=0;i<len;i++) {
if(strcmp(l1[i]->name,l2[i]->name) != 0) return 0;
}
return 1;
}
List*
prefixdup(List* prefix)
{
List* dupseq;
int i;
if(prefix == NULL) return listnew();
dupseq = listnew();
listsetalloc(dupseq,listlength(prefix));
for(i=0;i<listlength(prefix);i++) listpush(dupseq,listget(prefix,i));
return dupseq;
}
/*
Many of the generate routines need to construct
heap strings for short periods. Remembering to
free such space is error prone, so provide a
pseudo-GC to handle these short term requests.
The idea is to have a fixed size pool
tracking malloc requests and automatically
releasing when the pool gets full.
*/
/* Max number of allocated pool items*/
#define POOLMAX 100
static char* pool[POOLMAX];
static int poolindex = -1;
#define POOL_DEFAULT 256
char*
poolalloc(size_t length)
{
if(poolindex == -1) { /* initialize*/
memset((void*)pool,0,sizeof(pool));
poolindex = 0;
}
if(poolindex == POOLMAX) poolindex=0;
if(length == 0) length = POOL_DEFAULT;
if(pool[poolindex] != NULL) efree(pool[poolindex]);
pool[poolindex] = (char*)ecalloc(length);
return pool[poolindex++];
}
char*
pooldup(const char* s)
{
char* sdup = poolalloc(strlen(s)+1);
strncpy(sdup,s,(strlen(s)+1));
return sdup;
}
char*
poolcat(const char* s1, const char* s2)
{
int len1, len2;
char* cat;
if(s1 == NULL && s2 == NULL) return NULL;
len1 = (s1?strlen(s1):0);
len2 = (s2?strlen(s2):0);
cat = poolalloc(len1+len2+1);
cat[0] = '\0';
if(s1 != NULL) strcat(cat,s1);
if(s2 != NULL) strcat(cat,s2);
return cat;
}
/* Result is malloc'd*/
unsigned char*
makebytestring(char* s, size_t* lenp)
{
unsigned char* bytes;
unsigned char* b;
size_t slen = strlen(s); /* # nibbles */
size_t blen = slen/2; /* # bytes */
int i;
ASSERT((slen%2) == 0);
ASSERT(blen > 0);
bytes = (unsigned char*)ecalloc(blen);
b = bytes;
for(i=0;i<slen;i+=2) {
unsigned int digit1 = chartohex(*s++);
unsigned int digit2 = chartohex(*s++);
unsigned int byte = (digit1 << 4) | digit2;
*b++ = byte;
}
if(lenp) *lenp = blen;
return bytes;
}
int
getpadding(int offset, int alignment)
{
int rem = (alignment==0?0:(offset % alignment));
int pad = (rem==0?0:(alignment - rem));
return pad;
}
static void
reclaimSymbols(void)
{
int i;
for(i=0;i<listlength(symlist);i++) {
Symbol* sym = listget(symlist,i);
freeSymbol(sym);
}
}
void
cleanup()
{
reclaimSymbols();
listfree(symlist);
listfree(grpdefs);
listfree(dimdefs);
listfree(attdefs);
listfree(gattdefs);
listfree(xattdefs);
listfree(typdefs);
listfree(vardefs);
filldatalist->readonly = 0;
freedatalist(filldatalist);
}
/* compute the total n-dimensional size as 1 long array;
if stop == 0, then stop = dimset->ndims.
*/
size_t
crossproduct(Dimset* dimset, int start, int stop)
{
size_t totalsize = 1;
int i;
for(i=start;i<stop;i++) {
totalsize = totalsize * dimset->dimsyms[i]->dim.declsize;
}
return totalsize;
}
/* Do the "complement" of crossproduct;
compute the total n-dimensional size of an array
starting at 0 thru the 'last' array index.
stop if we encounter an unlimited dimension
*/
size_t
prefixarraylength(Dimset* dimset, int last)
{
return crossproduct(dimset,0,last+1);
}
#ifdef USE_HDF5
extern int H5Eprint1(FILE * stream);
#endif
void
check_err(const int stat, const int line, const char* file, const char* func)
{
check_err2(stat,-1,line,file,func);
}
void check_err2(const int stat, const int cdlline, const int line, const char* file, const char* func)
{
if (stat != NC_NOERR) {
if(cdlline >= 0)
fprintf(stderr, "ncgen: cdl line %d; %s\n", cdlline, nc_strerror(stat));
else
fprintf(stderr, "ncgen: %s\n", nc_strerror(stat));
fprintf(stderr, "\t(%s:%s:%d)\n", file,func,line);
#ifdef USE_HDF5
H5Eprint1(stderr);
#endif
fflush(stderr);
finalize_netcdf(1);
}
}
/**
Find the index of the first unlimited
dimension at or after 'start'.
If no unlimited exists, return |dimset|
*/
int
findunlimited(Dimset* dimset, int start)
{
for(;start<dimset->ndims;start++) {
if(dimset->dimsyms[start]->dim.isunlimited)
return start;
}
return dimset->ndims;
}
/**
Find the index of the last unlimited
dimension.
If no unlimited exists, return |dimset|
*/
int
findlastunlimited(Dimset* dimset)
{
int i;
for(i=dimset->ndims-1;i>=0;i--) {
if(dimset->dimsyms[i]->dim.isunlimited)
return i;
}
return dimset->ndims;
}
/**
Count the number of unlimited dimensions.
*/
int
countunlimited(Dimset* dimset)
{
int i, count;
for(count=0,i=dimset->ndims-1;i>=0;i--) {
if(dimset->dimsyms[i]->dim.isunlimited)
count++;
}
return count;
}
/* Return standard format string */
const char *
kind_string(int kind)
{
switch (kind) {
case 1: return "classic";
case 2: return "64-bit offset";
case 3: return "netCDF-4";
case 4: return "netCDF-4 classic model";
default:
derror("Unknown format index: %d\n",kind);
}
return NULL;
}