From 3a0c5378abefaf86aa39a62a7c9682bdb21568a1 Mon Sep 17 00:00:00 2001 From: Lasse Collin Date: Thu, 14 Jun 2012 10:52:33 +0300 Subject: Docs: Add new example programs. These have more comments than the old examples and human-readable error messages. More tutorial-like examples are needed but these are a start. --- doc/examples/02_decompress.c | 287 +++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 287 insertions(+) create mode 100644 doc/examples/02_decompress.c (limited to 'doc/examples/02_decompress.c') diff --git a/doc/examples/02_decompress.c b/doc/examples/02_decompress.c new file mode 100644 index 00000000..4c0f37cb --- /dev/null +++ b/doc/examples/02_decompress.c @@ -0,0 +1,287 @@ +/////////////////////////////////////////////////////////////////////////////// +// +/// \file 02_decompress.c +/// \brief Decompress .xz files to stdout +/// +/// Usage: ./02_decompress INPUT_FILES... > OUTFILE +/// +/// Example: ./02_decompress foo.xz bar.xz > foobar +// +// Author: Lasse Collin +// +// This file has been put into the public domain. +// You can do whatever you want with this file. +// +/////////////////////////////////////////////////////////////////////////////// + +#include +#include +#include +#include +#include +#include + + +static bool +init_decoder(lzma_stream *strm) +{ + // Initialize a .xz decoder. The decoder supports a memory usage limit + // and a set of flags. + // + // The memory usage of the decompressor depends on the settings used + // to compress a .xz file. It can vary from less than a megabyte to + // a few gigabytes, but in practice (at least for now) it rarely + // exceeds 65 MiB because that's how much memory is required to + // decompress files created with "xz -9". Settings requiring more + // memory take extra effort to use and don't (at least for now) + // provide significantly better compression in most cases. + // + // Memory usage limit is useful if it is important that the + // decompressor won't consume gigabytes of memory. The need + // for limiting depends on the application. In this example, + // no memory usage limiting is used. This is done by setting + // the limit to UINT64_MAX. + // + // The .xz format allows concatenating compressed files as is: + // + // echo foo | xz > foobar.xz + // echo bar | xz >> foobar.xz + // + // When decompressing normal standalone .xz files, LZMA_CONCATENATED + // should always be used to support decompression of concatenated + // .xz files. If LZMA_CONCATENATED isn't used, the decoder will stop + // after the first .xz stream. This can be useful when .xz data has + // been embedded inside another file format. + // + // Flags other than LZMA_CONCATENATED are supported too, and can + // be combined with bitwise-or. See lzma/container.h + // (src/liblzma/api/lzma/container.h in the source package or e.g. + // /usr/include/lzma/container.h depending on the install prefix) + // for details. + lzma_ret ret = lzma_stream_decoder( + strm, UINT64_MAX, LZMA_CONCATENATED); + + // Return successfully if the initialization went fine. + if (ret == LZMA_OK) + return true; + + // Something went wrong. The possible errors are documented in + // lzma/container.h (src/liblzma/api/lzma/container.h in the source + // package or e.g. /usr/include/lzma/container.h depending on the + // install prefix). + // + // Note that LZMA_MEMLIMIT_ERROR is never possible here. If you + // specify a very tiny limit, the error will be delayed until + // the first headers have been parsed by a call to lzma_code(). + const char *msg; + switch (ret) { + case LZMA_MEM_ERROR: + msg = "Memory allocation failed"; + break; + + case LZMA_OPTIONS_ERROR: + msg = "Unsupported decompressor flags"; + break; + + default: + // This is most likely LZMA_PROG_ERROR indicating a bug in + // this program or in liblzma. It is inconvenient to have a + // separate error message for errors that should be impossible + // to occur, but knowing the error code is important for + // debugging. That's why it is good to print the error code + // at least when there is no good error message to show. + msg = "Unknown error, possibly a bug"; + break; + } + + fprintf(stderr, "Error initializing the decoder: %s (error code %u)\n", + msg, ret); + return false; +} + + +static bool +decompress(lzma_stream *strm, const char *inname, FILE *infile, FILE *outfile) +{ + // When LZMA_CONCATENATED flag was used when initializing the decoder, + // we need to tell lzma_code() when there will be no more input. + // This is done by setting action to LZMA_FINISH instead of LZMA_RUN + // in the same way as it is done when encoding. + // + // When LZMA_CONCATENATED isn't used, there is no need to use + // LZMA_FINISH to tell when all the input has been read, but it + // is still OK to use it if you want. When LZMA_CONCATENATED isn't + // used, the decoder will stop after the first .xz stream. In that + // case some unused data may be left in strm->next_in. + lzma_action action = LZMA_RUN; + + uint8_t inbuf[BUFSIZ]; + uint8_t outbuf[BUFSIZ]; + + strm->next_in = NULL; + strm->avail_in = 0; + strm->next_out = outbuf; + strm->avail_out = sizeof(outbuf); + + while (true) { + if (strm->avail_in == 0 && !feof(infile)) { + strm->next_in = inbuf; + strm->avail_in = fread(inbuf, 1, sizeof(inbuf), + infile); + + if (ferror(infile)) { + fprintf(stderr, "%s: Read error: %s\n", + inname, strerror(errno)); + return false; + } + + // Once the end of the input file has been reached, + // we need to tell lzma_code() that no more input + // will be coming. As said before, this isn't required + // if the LZMA_CONATENATED flag isn't used when + // initializing the decoder. + if (feof(infile)) + action = LZMA_FINISH; + } + + lzma_ret ret = lzma_code(strm, action); + + if (strm->avail_out == 0 || ret == LZMA_STREAM_END) { + size_t write_size = sizeof(outbuf) - strm->avail_out; + + if (fwrite(outbuf, 1, write_size, outfile) + != write_size) { + fprintf(stderr, "Write error: %s\n", + strerror(errno)); + return false; + } + + strm->next_out = outbuf; + strm->avail_out = sizeof(outbuf); + } + + if (ret != LZMA_OK) { + // Once everything has been decoded successfully, the + // return value of lzma_code() will be LZMA_STREAM_END. + // + // It is important to check for LZMA_STREAM_END. Do not + // assume that getting ret != LZMA_OK would mean that + // everything has gone well or that when you aren't + // getting more output it must have successfully + // decoded everything. + if (ret == LZMA_STREAM_END) + return true; + + // It's not LZMA_OK nor LZMA_STREAM_END, + // so it must be an error code. See lzma/base.h + // (src/liblzma/api/lzma/base.h in the source package + // or e.g. /usr/include/lzma/base.h depending on the + // install prefix) for the list and documentation of + // possible values. Many values listen in lzma_ret + // enumeration aren't possible in this example, but + // can be made possible by enabling memory usage limit + // or adding flags to the decoder initialization. + const char *msg; + switch (ret) { + case LZMA_MEM_ERROR: + msg = "Memory allocation failed"; + break; + + case LZMA_FORMAT_ERROR: + // .xz magic bytes weren't found. + msg = "The input is not in the .xz format"; + break; + + case LZMA_OPTIONS_ERROR: + // For example, the headers specify a filter + // that isn't supported by this liblzma + // version (or it hasn't been enabled when + // building liblzma, but no-one sane does + // that unless building liblzma for an + // embedded system). Upgrading to a newer + // liblzma might help. + // + // Note that it is unlikely that the file has + // accidentally became corrupt if you get this + // error. The integrity of the .xz headers is + // always verified with a CRC32, so + // unintentionally corrupt files can be + // distinguished from unsupported files. + msg = "Unsupported compression options"; + break; + + case LZMA_DATA_ERROR: + msg = "Compressed file is corrupt"; + break; + + case LZMA_BUF_ERROR: + // Typically this error means that a valid + // file has got truncated, but it might also + // be a damaged part in the file that makes + // the decoder think the file is truncated. + // If you prefer, you can use the same error + // message for this as for LZMA_DATA_ERROR. + msg = "Compressed file is truncated or " + "otherwise corrupt"; + break; + + default: + // This is most likely LZMA_PROG_ERROR. + msg = "Unknown error, possibly a bug"; + break; + } + + fprintf(stderr, "%s: Decoder error: " + "%s (error code %u)\n", + inname, msg, ret); + return false; + } + } +} + + +extern int +main(int argc, char **argv) +{ + if (argc <= 1) { + fprintf(stderr, "Usage: %s FILES...\n", argv[0]); + return EXIT_FAILURE; + } + + lzma_stream strm = LZMA_STREAM_INIT; + + bool success = true; + + // Try to decompress all files. + for (int i = 1; i < argc; ++i) { + if (!init_decoder(&strm)) { + // Decoder initialization failed. There's no point + // to retry it so we need to exit. + success = false; + break; + } + + FILE *infile = fopen(argv[i], "rb"); + + if (infile == NULL) { + fprintf(stderr, "%s: Error opening the " + "input file: %s\n", + argv[i], strerror(errno)); + success = false; + } else { + success &= decompress(&strm, argv[i], infile, stdout); + fclose(infile); + } + } + + // Free the memory allocated for the decoder. This only needs to be + // done after the last file. + lzma_end(&strm); + + if (fclose(stdout)) { + fprintf(stderr, "Write error: %s\n", strerror(errno)); + success = false; + } + + return success ? EXIT_SUCCESS : EXIT_FAILURE; +} -- cgit v1.2.3