liblzma/doc/examples/11_file_info.c
2017-04-24 20:20:11 +03:00

207 lines
5.3 KiB
C

///////////////////////////////////////////////////////////////////////////////
//
/// \file 11_file_info.c
/// \brief Get uncmopressed size of .xz file(s)
///
/// Usage: ./11_file_info INFILE1.xz [INFILEn.xz]...
///
/// Example: ./11_file_info foo.xz
//
// Author: Lasse Collin
//
// This file has been put into the public domain.
// You can do whatever you want with this file.
//
///////////////////////////////////////////////////////////////////////////////
#include <stdbool.h>
#include <inttypes.h>
#include <stdlib.h>
#include <stdio.h>
#include <string.h>
#include <errno.h>
#include <lzma.h>
static bool
print_file_size(lzma_stream *strm, FILE *infile, const char *filename)
{
// Get the file size. In standard C it can be done by seeking to
// the end of the file and then getting the file position.
// In POSIX one can use fstat() and then st_size from struct stat.
// Also note that fseek() and ftell() use long and thus don't support
// large files on 32-bit systems (POSIX versions fseeko() and
// ftello() can support large files).
if (fseek(infile, 0, SEEK_END)) {
fprintf(stderr, "Error seeking the file `%s': %s\n",
filename, strerror(errno));
return false;
}
const long file_size = ftell(infile);
// The decoder wants to start from the beginning of the .xz file.
rewind(infile);
// Initialize the decoder.
lzma_index *i;
lzma_ret ret = lzma_file_info_decoder(strm, &i, UINT64_MAX,
(uint64_t)file_size);
switch (ret) {
case LZMA_OK:
// Initialization succeeded.
break;
case LZMA_MEM_ERROR:
fprintf(stderr, "Out of memory when initializing "
"the .xz file info decoder\n");
return false;
case LZMA_PROG_ERROR:
default:
fprintf(stderr, "Unknown error, possibly a bug\n");
return false;
}
// This example program reuses the same lzma_stream structure
// for multiple files, so we need to reset this when starting
// a new file.
strm->avail_in = 0;
// Buffer for input data.
uint8_t inbuf[BUFSIZ];
// Pass data to the decoder and seek when needed.
while (true) {
if (strm->avail_in == 0) {
strm->next_in = inbuf;
strm->avail_in = fread(inbuf, 1, sizeof(inbuf),
infile);
if (ferror(infile)) {
fprintf(stderr,
"Error reading from `%s': %s\n",
filename, strerror(errno));
return false;
}
// We don't need to care about hitting the end of
// the file so no need to check for feof().
}
ret = lzma_code(strm, LZMA_RUN);
switch (ret) {
case LZMA_OK:
break;
case LZMA_SEEK_NEEDED:
// The cast is safe because liblzma won't ask us to
// seek past the known size of the input file which
// did fit into a long.
//
// NOTE: Remember to change these to off_t if you
// switch fseeko() or lseek().
if (fseek(infile, (long)(strm->seek_pos), SEEK_SET)) {
fprintf(stderr, "Error seeking the "
"file `%s': %s\n",
filename, strerror(errno));
return false;
}
// The old data in the inbuf is useless now. Set
// avail_in to zero so that we will read new input
// from the new file position on the next iteration
// of this loop.
strm->avail_in = 0;
break;
case LZMA_STREAM_END:
// File information was successfully decoded.
// See <lzma/index.h> for functions that can be
// used on it. In this example we just print
// the uncompressed size (in bytes) of
// the .xz file followed by its file name.
printf("%10" PRIu64 " %s\n",
lzma_index_uncompressed_size(i),
filename);
// Free the memory of the lzma_index structure.
lzma_index_end(i, NULL);
return true;
case LZMA_FORMAT_ERROR:
// .xz magic bytes weren't found.
fprintf(stderr, "The file `%s' is not "
"in the .xz format\n", filename);
return false;
case LZMA_OPTIONS_ERROR:
fprintf(stderr, "The file `%s' has .xz headers that "
"are not supported by this liblzma "
"version\n", filename);
return false;
case LZMA_DATA_ERROR:
fprintf(stderr, "The file `%s' is corrupt\n",
filename);
return false;
case LZMA_MEM_ERROR:
fprintf(stderr, "Memory allocation failed when "
"decoding the file `%s'\n", filename);
return false;
// LZMA_MEMLIMIT_ERROR shouldn't happen because we used
// UINT64_MAX as the limit.
//
// LZMA_BUF_ERROR shouldn't happen because we always provide
// new input when the input buffer is empty. The decoder
// knows the input file size and thus won't try to read past
// the end of the file.
case LZMA_MEMLIMIT_ERROR:
case LZMA_BUF_ERROR:
case LZMA_PROG_ERROR:
default:
fprintf(stderr, "Unknown error, possibly a bug\n");
return false;
}
}
// This line is never reached.
}
extern int
main(int argc, char **argv)
{
bool success = true;
lzma_stream strm = LZMA_STREAM_INIT;
for (int i = 1; i < argc; ++i) {
FILE *infile = fopen(argv[i], "rb");
if (infile == NULL) {
fprintf(stderr, "Cannot open the file `%s': %s\n",
argv[i], strerror(errno));
success = false;
}
success &= print_file_size(&strm, infile, argv[i]);
(void)fclose(infile);
}
lzma_end(&strm);
// Close stdout to catch possible write errors that can occur
// when pending data is flushed from the stdio buffers.
if (fclose(stdout)) {
fprintf(stderr, "Write error: %s\n", strerror(errno));
success = false;
}
return success ? EXIT_SUCCESS : EXIT_FAILURE;
}