ceph/src/zstd/examples/dictionary_compression.c

   1 /*
   2  * Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
   3  * All rights reserved.
   4  *
   5  * This source code is licensed under both the BSD-style license (found in the
   6  * LICENSE file in the root directory of this source tree) and the GPLv2 (found
   7  * in the COPYING file in the root directory of this source tree).
   8  * You may select, at your option, one of the above-listed licenses.
   9  */
  10 #include <stdio.h>     // printf
  11 #include <stdlib.h>    // free
  12 #include <string.h>    // memset, strcat
  13 #include <zstd.h>      // presumes zstd library is installed
  14 #include "common.h"    // Helper functions, CHECK(), and CHECK_ZSTD()
  15
  16 /* createDict() :
  17    `dictFileName` is supposed to have been created using `zstd --train` */
  18 static ZSTD_CDict* createCDict_orDie(const char* dictFileName, int cLevel)
  19 {
  20     size_t dictSize;
  21     printf("loading dictionary %s \n", dictFileName);
  22     void* const dictBuffer = mallocAndLoadFile_orDie(dictFileName, &dictSize);
  23     ZSTD_CDict* const cdict = ZSTD_createCDict(dictBuffer, dictSize, cLevel);
  24     CHECK(cdict != NULL, "ZSTD_createCDict() failed!");
  25     free(dictBuffer);
  26     return cdict;
  27 }
  28
  29
  30 static void compress(const char* fname, const char* oname, const ZSTD_CDict* cdict)
  31 {
  32     size_t fSize;
  33     void* const fBuff = mallocAndLoadFile_orDie(fname, &fSize);
  34     size_t const cBuffSize = ZSTD_compressBound(fSize);
  35     void* const cBuff = malloc_orDie(cBuffSize);
  36
  37     /* Compress using the dictionary.
  38      * This function writes the dictionary id, and content size into the header.
  39      * But, it doesn't use a checksum. You can control these options using the
  40      * advanced API: ZSTD_CCtx_setParameter(), ZSTD_CCtx_refCDict(),
  41      * and ZSTD_compress2().
  42      */
  43     ZSTD_CCtx* const cctx = ZSTD_createCCtx();
  44     CHECK(cctx != NULL, "ZSTD_createCCtx() failed!");
  45     size_t const cSize = ZSTD_compress_usingCDict(cctx, cBuff, cBuffSize, fBuff, fSize, cdict);
  46     CHECK_ZSTD(cSize);
  47
  48     saveFile_orDie(oname, cBuff, cSize);
  49
  50     /* success */
  51     printf("%25s : %6u -> %7u - %s \n", fname, (unsigned)fSize, (unsigned)cSize, oname);
  52
  53     ZSTD_freeCCtx(cctx);   /* never fails */
  54     free(fBuff);
  55     free(cBuff);
  56 }
  57
  58
  59 static char* createOutFilename_orDie(const char* filename)
  60 {
  61     size_t const inL = strlen(filename);
  62     size_t const outL = inL + 5;
  63     void* outSpace = malloc_orDie(outL);
  64     memset(outSpace, 0, outL);
  65     strcat(outSpace, filename);
  66     strcat(outSpace, ".zst");
  67     return (char*)outSpace;
  68 }
  69
  70 int main(int argc, const char** argv)
  71 {
  72     const char* const exeName = argv[0];
  73     int const cLevel = 3;
  74
  75     if (argc<3) {
  76         fprintf(stderr, "wrong arguments\n");
  77         fprintf(stderr, "usage:\n");
  78         fprintf(stderr, "%s [FILES] dictionary\n", exeName);
  79         return 1;
  80     }
  81
  82     /* load dictionary only once */
  83     const char* const dictName = argv[argc-1];
  84     ZSTD_CDict* const dictPtr = createCDict_orDie(dictName, cLevel);
  85
  86     int u;
  87     for (u=1; u<argc-1; u++) {
  88         const char* inFilename = argv[u];
  89         char* const outFilename = createOutFilename_orDie(inFilename);
  90         compress(inFilename, outFilename, dictPtr);
  91         free(outFilename);
  92     }
  93
  94     ZSTD_freeCDict(dictPtr);
  95     printf("All %u files compressed. \n", argc-2);
  96     return 0;
  97 }