From: Eric Biggers Date: Sun, 16 Dec 2012 02:50:30 +0000 (-0600) Subject: Various cleanups X-Git-Tag: v1.2.1~57 X-Git-Url: https://wimlib.net/git/?p=wimlib;a=commitdiff_plain;h=034fd063f956d0806e557680a36a69a42556a776 Various cleanups - wim_timestamp_to_str() instead of asctime(gmtime(wim_timestamp_to_unix())) - inode_stream_lte() non-inline to save space - new errors WIMLIB_ERR_CHAR_CONVERSION and WIMLIB_ERR_INVALID_PART_NUMBER - write_xml_data(): Have libxml2 write directly to output file with UTF-16LE encoding (no need for in-memory buffer and converting to UTF-16LE afterwards) - read_header(): Validate total_parts and part_number - Fix a couple potential integer overflows - Fix some indentation and phase out more 'uint' uses - Remove some unneeded DEBUG() statements --- diff --git a/src/comp.h b/src/comp.h index 76f958c2..37b1f992 100644 --- a/src/comp.h +++ b/src/comp.h @@ -20,7 +20,7 @@ struct output_bitstream { output_bitbuf_t bitbuf; /* Number of free bits in @bitbuf */ - uint free_bits; + unsigned free_bits; u8 *bit_output; u8 *next_bit_output; @@ -56,27 +56,27 @@ static inline int bitstream_put_two_bytes(struct output_bitstream *ostream, } struct lz_params { - uint min_match; - uint max_match; - uint nice_match; - uint good_match; - uint max_chain_len; - uint max_lazy_match; - uint too_far; + unsigned min_match; + unsigned max_match; + unsigned nice_match; + unsigned good_match; + unsigned max_chain_len; + unsigned max_lazy_match; + unsigned too_far; }; -typedef uint (*lz_record_match_t)(uint, uint, void *, void *); -typedef uint (*lz_record_literal_t)(u8, void *); +typedef unsigned (*lz_record_match_t)(unsigned, unsigned, void *, void *); +typedef unsigned (*lz_record_literal_t)(u8, void *); -extern uint lz_analyze_block(const u8 uncompressed_data[], - uint uncompressed_len, - u32 match_tab[], - lz_record_match_t record_match, - lz_record_literal_t record_literal, - void *record_match_arg1, - void *record_match_arg2, - void *record_literal_arg, - const struct lz_params *params); +extern unsigned lz_analyze_block(const u8 uncompressed_data[], + unsigned uncompressed_len, + u32 match_tab[], + lz_record_match_t record_match, + lz_record_literal_t record_literal, + void *record_match_arg1, + void *record_match_arg2, + void *record_literal_arg, + const struct lz_params *params); extern int bitstream_put_bits(struct output_bitstream *ostream, output_bitbuf_t bits, unsigned num_bits); @@ -86,8 +86,10 @@ extern void init_output_bitstream(struct output_bitstream *ostream, extern int flush_output_bitstream(struct output_bitstream *ostream); -extern void make_canonical_huffman_code(uint num_syms, uint max_codeword_len, - const u32 freq_tab[], u8 lens[], +extern void make_canonical_huffman_code(unsigned num_syms, + unsigned max_codeword_len, + const u32 freq_tab[], + u8 lens[], u16 codewords[]); #endif /* _WIMLIB_COMP_H */ diff --git a/src/decomp.h b/src/decomp.h index 3ecf9d2f..0bd2a69f 100644 --- a/src/decomp.h +++ b/src/decomp.h @@ -25,15 +25,15 @@ struct input_bitstream { const u8 *data; /* Number of bits in @bitbuf that are valid. */ - uint bitsleft; + unsigned bitsleft; /* Number of words of data that are left. */ - uint data_bytes_left; + unsigned data_bytes_left; }; /* Initializes a bitstream to receive its input from @data. */ static inline void init_input_bitstream(struct input_bitstream *istream, - const void *data, uint num_data_bytes) + const void *data, unsigned num_data_bytes) { istream->bitbuf = 0; istream->bitsleft = 0; @@ -43,7 +43,7 @@ static inline void init_input_bitstream(struct input_bitstream *istream, /* Ensures that the bit buffer contains @num_bits bits. */ static inline int bitstream_ensure_bits(struct input_bitstream *istream, - uint num_bits) + unsigned num_bits) { wimlib_assert(num_bits <= 16); @@ -62,8 +62,8 @@ static inline int bitstream_ensure_bits(struct input_bitstream *istream, if (istream->data_bytes_left < 2) return 1; - uint shift = sizeof(input_bitbuf_t) * 8 - 16 - - istream->bitsleft; + unsigned shift = sizeof(input_bitbuf_t) * 8 - 16 - + istream->bitsleft; istream->bitbuf |= (input_bitbuf_t)le16_to_cpu( *(u16*)istream->data) << shift; istream->data += 2; @@ -75,8 +75,8 @@ static inline int bitstream_ensure_bits(struct input_bitstream *istream, /* Returns the next @num_bits bits in the bit buffer. It must contain at least * @num_bits bits to call this function. */ -static inline uint bitstream_peek_bits(const struct input_bitstream *istream, - uint num_bits) +static inline unsigned +bitstream_peek_bits(const struct input_bitstream *istream, unsigned num_bits) { if (num_bits == 0) return 0; @@ -86,7 +86,7 @@ static inline uint bitstream_peek_bits(const struct input_bitstream *istream, /* Removes @num_bits bits from the bit buffer. It must contain at least * @num_bits bits to call this function. */ static inline void bitstream_remove_bits(struct input_bitstream *istream, - uint num_bits) + unsigned num_bits) { istream->bitbuf <<= num_bits; istream->bitsleft -= num_bits; @@ -94,7 +94,7 @@ static inline void bitstream_remove_bits(struct input_bitstream *istream, /* Reads and returns @num_bits bits from the input bitstream. */ static inline int bitstream_read_bits(struct input_bitstream *istream, - uint num_bits, uint *n) + unsigned num_bits, unsigned *n) { int ret; ret = bitstream_ensure_bits(istream, num_bits); @@ -129,10 +129,10 @@ static inline int bitstream_read_byte(struct input_bitstream *istream) /* Reads @num_bits bits from the bit buffer without checking to see if that many * bits are in the buffer or not. */ -static inline uint bitstream_read_bits_nocheck(struct input_bitstream *istream, - uint num_bits) +static inline unsigned +bitstream_read_bits_nocheck(struct input_bitstream *istream, unsigned num_bits) { - uint n = bitstream_peek_bits(istream, num_bits); + unsigned n = bitstream_peek_bits(istream, num_bits); bitstream_remove_bits(istream, num_bits); return n; } @@ -156,11 +156,11 @@ extern int read_huffsym(struct input_bitstream *stream, const u8 lengths[], unsigned num_symbols, unsigned table_bits, - uint *n, + unsigned *n, unsigned max_codeword_len); -extern int make_huffman_decode_table(u16 decode_table[], uint num_syms, - uint num_bits, const u8 lengths[], - uint max_codeword_len); +extern int make_huffman_decode_table(u16 decode_table[], unsigned num_syms, + unsigned num_bits, const u8 lengths[], + unsigned max_codeword_len); #endif /* _WIMLIB_DECOMP_H */ diff --git a/src/dentry.c b/src/dentry.c index 5ea26d3e..54601493 100644 --- a/src/dentry.c +++ b/src/dentry.c @@ -756,49 +756,37 @@ int print_dentry(struct dentry *dentry, void *lookup_table) const u8 *hash; struct lookup_table_entry *lte; const struct inode *inode = dentry->d_inode; - time_t time; - char *p; + char buf[50]; printf("[DENTRY]\n"); printf("Length = %"PRIu64"\n", dentry->length); printf("Attributes = 0x%x\n", inode->attributes); - for (unsigned i = 0; i < ARRAY_LEN(file_attr_flags); i++) + for (size_t i = 0; i < ARRAY_LEN(file_attr_flags); i++) if (file_attr_flags[i].flag & inode->attributes) printf(" FILE_ATTRIBUTE_%s is set\n", file_attr_flags[i].name); printf("Security ID = %d\n", inode->security_id); printf("Subdir offset = %"PRIu64"\n", dentry->subdir_offset); - /* Translate the timestamps into something readable */ - time = wim_timestamp_to_unix(inode->creation_time); - p = asctime(gmtime(&time)); - *(strrchr(p, '\n')) = '\0'; - printf("Creation Time = %s UTC\n", p); + wim_timestamp_to_str(inode->creation_time, buf, sizeof(buf)); + printf("Creation Time = %s\n", buf); - time = wim_timestamp_to_unix(inode->last_access_time); - p = asctime(gmtime(&time)); - *(strrchr(p, '\n')) = '\0'; - printf("Last Access Time = %s UTC\n", p); + wim_timestamp_to_str(inode->last_access_time, buf, sizeof(buf)); + printf("Last Access Time = %s\n", buf); - time = wim_timestamp_to_unix(inode->last_write_time); - p = asctime(gmtime(&time)); - *(strrchr(p, '\n')) = '\0'; - printf("Last Write Time = %s UTC\n", p); + wim_timestamp_to_str(inode->last_write_time, buf, sizeof(buf)); + printf("Last Write Time = %s\n", buf); printf("Reparse Tag = 0x%"PRIx32"\n", inode->reparse_tag); printf("Hard Link Group = 0x%"PRIx64"\n", inode->ino); printf("Hard Link Group Size = %"PRIu32"\n", inode->link_count); printf("Number of Alternate Data Streams = %hu\n", inode->num_ads); - printf("Filename = \""); - print_string(dentry->file_name, dentry->file_name_len); - puts("\""); - printf("Filename Length = %hu\n", dentry->file_name_len); printf("Filename (UTF-8) = \"%s\"\n", dentry->file_name_utf8); - printf("Filename (UTF-8) Length = %hu\n", dentry->file_name_utf8_len); - printf("Short Name = \""); + /*printf("Filename (UTF-8) Length = %hu\n", dentry->file_name_utf8_len);*/ + printf("Short Name (UTF-16LE) = \""); print_string(dentry->short_name, dentry->short_name_len); puts("\""); - printf("Short Name Length = %hu\n", dentry->short_name_len); + /*printf("Short Name Length = %hu\n", dentry->short_name_len);*/ printf("Full Path (UTF-8) = \"%s\"\n", dentry->full_path_utf8); lte = inode_stream_lte(dentry->d_inode, 0, lookup_table); if (lte) { diff --git a/src/header.c b/src/header.c index 2a7fe32b..bb172091 100644 --- a/src/header.c +++ b/src/header.c @@ -26,6 +26,7 @@ #include "wimlib_internal.h" #include "io.h" +#include /* First 8 bytes in every WIM file. */ static const u8 wim_magic_chars[WIM_MAGIC_LEN] = { @@ -104,6 +105,15 @@ int read_header(FILE *fp, struct wim_header *hdr, int open_flags) p = get_u16(p, &hdr->part_number); p = get_u16(p, &hdr->total_parts); + if (hdr->total_parts == 0 || + hdr->part_number == 0 || + hdr->part_number > hdr->total_parts) + { + ERROR("Invalid WIM part number: %hu of %hu", + hdr->part_number, hdr->total_parts); + return WIMLIB_ERR_INVALID_PART_NUMBER; + } + if (!(open_flags & WIMLIB_OPEN_FLAG_SPLIT_OK) && (hdr->part_number != 1 || hdr->total_parts != 1)) { @@ -117,6 +127,11 @@ int read_header(FILE *fp, struct wim_header *hdr, int open_flags) DEBUG("part_number = %u, total_parts = %u, image_count = %u", hdr->part_number, hdr->total_parts, hdr->image_count); + if (hdr->image_count >= INT_MAX) { + ERROR("Invalid image count (%u)", hdr->image_count); + return WIMLIB_ERR_IMAGE_COUNT; + } + /* Byte 48 */ p = get_resource_entry(p, &hdr->lookup_table_res_entry); @@ -240,14 +255,13 @@ struct hdr_flag hdr_flags[] = { WIMLIBAPI void wimlib_print_header(const WIMStruct *w) { const struct wim_header *hdr = &w->hdr; - uint i; printf("Magic Characters = MSWIM\\000\\000\\000\n"); printf("Header Size = %u\n", WIM_HEADER_DISK_SIZE); printf("Version = 0x%x\n", WIM_VERSION); printf("Flags = 0x%x\n", hdr->flags); - for (i = 0; i < ARRAY_LEN(hdr_flags); i++) + for (size_t i = 0; i < ARRAY_LEN(hdr_flags); i++) if (hdr_flags[i].flag & hdr->flags) printf(" WIM_HDR_FLAG_%s is set\n", hdr_flags[i].name); diff --git a/src/lookup_table.c b/src/lookup_table.c index 364f469e..d3bc8b24 100644 --- a/src/lookup_table.c +++ b/src/lookup_table.c @@ -686,6 +686,24 @@ int dentry_unresolve_ltes(struct dentry *dentry, void *ignore) return 0; } +/* + * Returns the lookup table entry for stream @stream_idx of the inode, where + * stream_idx = 0 means the default un-named file stream, and stream_idx >= 1 + * corresponds to an alternate data stream. + * + * This works for both resolved and un-resolved dentries. + */ +struct lookup_table_entry * +inode_stream_lte(const struct inode *inode, unsigned stream_idx, + const struct lookup_table *table) +{ + if (inode->resolved) + return inode_stream_lte_resolved(inode, stream_idx); + else + return inode_stream_lte_unresolved(inode, stream_idx, table); +} + + /* Return the lookup table entry for the unnamed data stream of an inode, or * NULL if there is none. * diff --git a/src/lookup_table.h b/src/lookup_table.h index 592116bc..7038cc5d 100644 --- a/src/lookup_table.h +++ b/src/lookup_table.h @@ -325,23 +325,10 @@ inode_stream_lte_unresolved(const struct inode *inode, unsigned stream_idx, inode->ads_entries[ stream_idx - 1].hash); } -/* - * Returns the lookup table entry for stream @stream_idx of the inode, where - * stream_idx = 0 means the default un-named file stream, and stream_idx >= 1 - * corresponds to an alternate data stream. - * - * This works for both resolved and un-resolved dentries. - */ -static inline struct lookup_table_entry * -inode_stream_lte(const struct inode *inode, unsigned stream_idx, - const struct lookup_table *table) -{ - if (inode->resolved) - return inode_stream_lte_resolved(inode, stream_idx); - else - return inode_stream_lte_unresolved(inode, stream_idx, table); -} +extern struct lookup_table_entry * +inode_stream_lte(const struct inode *inode, unsigned stream_idx, + const struct lookup_table *table); static inline const u8 *inode_stream_hash_unresolved(const struct inode *inode, unsigned stream_idx) diff --git a/src/lzx-comp.c b/src/lzx-comp.c index 34251e34..f96d90c3 100644 --- a/src/lzx-comp.c +++ b/src/lzx-comp.c @@ -83,9 +83,6 @@ struct lzx_freq_tables { u32 aligned_freq_table[LZX_ALIGNEDTREE_NUM_SYMBOLS]; }; - - - /* Returns the LZX position slot that corresponds to a given formatted offset. * * Logically, this returns the smallest i such that @@ -146,19 +143,19 @@ static inline unsigned lzx_get_num_extra_bits(unsigned position_slot) * the frequency of symbols in the main, length, and aligned offset alphabets. * The return value is a 32-bit number that provides the match in an * intermediate representation documented below. */ -static u32 lzx_record_match(uint match_offset, uint match_len, +static u32 lzx_record_match(unsigned match_offset, unsigned match_len, void *__freq_tabs, void *__queue) { struct lzx_freq_tables *freq_tabs = __freq_tabs; struct lru_queue *queue = __queue; - uint formatted_offset; - uint position_slot; - uint position_footer = 0; + unsigned formatted_offset; + unsigned position_slot; + unsigned position_footer = 0; u32 match; u32 len_header; u32 len_pos_header; - uint len_footer; - uint adjusted_match_len; + unsigned len_footer; + unsigned adjusted_match_len; wimlib_assert(match_len >= LZX_MIN_MATCH && match_len <= LZX_MAX_MATCH); wimlib_assert(match_offset != 0); @@ -261,21 +258,21 @@ static u32 lzx_record_match(uint match_offset, uint match_len, * main, length, and aligned offset Huffman codes. */ static int lzx_write_match(struct output_bitstream *out, int block_type, - u32 match, const struct lzx_codes *codes) + u32 match, const struct lzx_codes *codes) { /* low 8 bits are the match length minus 2 */ - uint match_len_minus_2 = match & 0xff; + unsigned match_len_minus_2 = match & 0xff; /* Next 17 bits are the position footer */ - uint position_footer = (match >> 8) & 0x1ffff; /* 17 bits */ + unsigned position_footer = (match >> 8) & 0x1ffff; /* 17 bits */ /* Next 6 bits are the position slot. */ - uint position_slot = (match >> 25) & 0x3f; /* 6 bits */ - uint len_header; - uint len_footer; - uint len_pos_header; - uint main_symbol; - uint num_extra_bits; - uint verbatim_bits; - uint aligned_bits; + unsigned position_slot = (match >> 25) & 0x3f; /* 6 bits */ + unsigned len_header; + unsigned len_footer; + unsigned len_pos_header; + unsigned main_symbol; + unsigned num_extra_bits; + unsigned verbatim_bits; + unsigned aligned_bits; int ret; /* If the match length is less than MIN_MATCH (= 2) + @@ -291,7 +288,7 @@ static int lzx_write_match(struct output_bitstream *out, int block_type, len_header = match_len_minus_2; /* No length footer-- mark it with a special * value. */ - len_footer = (uint)(-1); + len_footer = (unsigned)(-1); } else { len_header = LZX_NUM_PRIMARY_LENS; len_footer = match_len_minus_2 - LZX_NUM_PRIMARY_LENS; @@ -315,7 +312,7 @@ static int lzx_write_match(struct output_bitstream *out, int block_type, /* If there is a length footer, output it using the * length Huffman code. */ - if (len_footer != (uint)(-1)) { + if (len_footer != (unsigned)(-1)) { ret = bitstream_put_bits(out, codes->len_codewords[len_footer], codes->len_lens[len_footer]); if (ret != 0) @@ -369,10 +366,10 @@ static int lzx_write_match(struct output_bitstream *out, int block_type, static int lzx_write_compressed_literals(struct output_bitstream *ostream, int block_type, const u32 match_tab[], - uint num_compressed_literals, + unsigned num_compressed_literals, const struct lzx_codes *codes) { - uint i; + unsigned i; u32 match; int ret; @@ -417,20 +414,19 @@ static int lzx_write_compressed_literals(struct output_bitstream *ostream, * @num_symbols: The number of symbols in the code. */ static int lzx_write_compressed_tree(struct output_bitstream *out, - const u8 lens[], - uint num_symbols) + const u8 lens[], unsigned num_symbols) { /* Frequencies of the length symbols, including the RLE symbols (NOT the * actual lengths themselves). */ - uint pretree_freqs[LZX_PRETREE_NUM_SYMBOLS]; + unsigned pretree_freqs[LZX_PRETREE_NUM_SYMBOLS]; u8 pretree_lens[LZX_PRETREE_NUM_SYMBOLS]; u16 pretree_codewords[LZX_PRETREE_NUM_SYMBOLS]; u8 output_syms[num_symbols * 2]; - uint output_syms_idx; - uint cur_run_len; - uint i; - uint len_in_run; - uint additional_bits; + unsigned output_syms_idx; + unsigned cur_run_len; + unsigned i; + unsigned len_in_run; + unsigned additional_bits; char delta; u8 pretree_sym; @@ -609,7 +605,7 @@ static void lzx_make_huffman_codes(const struct lzx_freq_tables *freq_tabs, * format as used in other file formats such as the cabinet format, where a bit * is reserved for that purpose. */ static void do_call_insn_preprocessing(u8 uncompressed_data[], - uint uncompressed_data_len) + unsigned uncompressed_data_len) { int i = 0; int file_size = LZX_MAGIC_FILESIZE; @@ -671,39 +667,37 @@ static const struct lz_params lzx_lz_params = { * not reduce its size, and @compressed_data will not contain the full * compressed data. */ -int lzx_compress(const void *__uncompressed_data, uint uncompressed_len, - void *compressed_data, uint *compressed_len_ret) +int lzx_compress(const void *__uncompressed_data, unsigned uncompressed_len, + void *compressed_data, unsigned *compressed_len_ret) { struct output_bitstream ostream; u8 uncompressed_data[uncompressed_len + LZX_MAX_MATCH]; struct lzx_freq_tables freq_tabs; struct lzx_codes codes; u32 match_tab[uncompressed_len]; - struct lru_queue queue = {.R0 = 1, .R1 = 1, .R2 = 1}; - uint num_matches; - uint compressed_len; - uint i; + struct lru_queue queue; + unsigned num_matches; + unsigned compressed_len; + unsigned i; int ret; int block_type = LZX_BLOCKTYPE_ALIGNED; - LZX_DEBUG("uncompressed_len = %u", uncompressed_len); - if (uncompressed_len < 100) return 1; - memset(&freq_tabs, 0, sizeof(freq_tabs)); + queue.R0 = 1; + queue.R1 = 1; + queue.R2 = 1; /* The input data must be preprocessed. To avoid changing the original * input, copy it to a temporary buffer. */ memcpy(uncompressed_data, __uncompressed_data, uncompressed_len); - /* Before doing any actual compression, do the call instruction (0xe8 * byte) translation on the uncompressed data. */ do_call_insn_preprocessing(uncompressed_data, uncompressed_len); - /* Determine the sequence of matches and literals that will be output, * and in the process, keep counts of the number of times each symbol * will be output, so that the Huffman trees can be made. */ @@ -714,8 +708,6 @@ int lzx_compress(const void *__uncompressed_data, uint uncompressed_len, &queue, freq_tabs.main_freq_table, &lzx_lz_params); - LZX_DEBUG("using %u matches", num_matches); - lzx_make_huffman_codes(&freq_tabs, &codes); /* Initialize the output bitstream. */ @@ -776,14 +768,10 @@ int lzx_compress(const void *__uncompressed_data, uint uncompressed_len, compressed_len = ostream.bit_output - (u8*)compressed_data; - LZX_DEBUG("Compressed %u => %u bytes", - uncompressed_len, compressed_len); - *compressed_len_ret = compressed_len; #ifdef ENABLE_VERIFY_COMPRESSION /* Verify that we really get the same thing back when decompressing. */ - LZX_DEBUG("Verifying the compressed data."); u8 buf[uncompressed_len]; ret = lzx_decompress(compressed_data, compressed_len, buf, uncompressed_len); @@ -800,8 +788,6 @@ int lzx_compress(const void *__uncompressed_data, uint uncompressed_len, abort(); } } - LZX_DEBUG("Compression verified to be correct."); #endif - return 0; } diff --git a/src/lzx-decomp.c b/src/lzx-decomp.c index fb8b17e9..73e68127 100644 --- a/src/lzx-decomp.c +++ b/src/lzx-decomp.c @@ -135,7 +135,7 @@ struct lzx_tables { */ static inline int read_huffsym_using_pretree(struct input_bitstream *istream, const u16 pretree_decode_table[], - const u8 pretree_lens[], uint *n) + const u8 pretree_lens[], unsigned *n) { return read_huffsym(istream, pretree_decode_table, pretree_lens, LZX_PRETREE_NUM_SYMBOLS, LZX_PRETREE_TABLEBITS, n, @@ -145,7 +145,7 @@ static inline int read_huffsym_using_pretree(struct input_bitstream *istream, /* Reads a Huffman-encoded symbol using the main tree. */ static inline int read_huffsym_using_maintree(struct input_bitstream *istream, const struct lzx_tables *tables, - uint *n) + unsigned *n) { return read_huffsym(istream, tables->maintree_decode_table, tables->maintree_lens, LZX_MAINTREE_NUM_SYMBOLS, @@ -155,7 +155,7 @@ static inline int read_huffsym_using_maintree(struct input_bitstream *istream, /* Reads a Huffman-encoded symbol using the length tree. */ static inline int read_huffsym_using_lentree(struct input_bitstream *istream, const struct lzx_tables *tables, - uint *n) + unsigned *n) { return read_huffsym(istream, tables->lentree_decode_table, tables->lentree_lens, LZX_LENTREE_NUM_SYMBOLS, @@ -165,7 +165,7 @@ static inline int read_huffsym_using_lentree(struct input_bitstream *istream, /* Reads a Huffman-encoded symbol using the aligned offset tree. */ static inline int read_huffsym_using_alignedtree(struct input_bitstream *istream, const struct lzx_tables *tables, - uint *n) + unsigned *n) { return read_huffsym(istream, tables->alignedtree_decode_table, tables->alignedtree_lens, @@ -186,14 +186,14 @@ static inline int read_huffsym_using_alignedtree(struct input_bitstream *istream * */ static int lzx_read_code_lens(struct input_bitstream *istream, u8 lens[], - uint num_lens) + unsigned num_lens) { /* Declare the decoding table and length table for the pretree. */ u16 pretree_decode_table[(1 << LZX_PRETREE_TABLEBITS) + (LZX_PRETREE_NUM_SYMBOLS * 2)]; u8 pretree_lens[LZX_PRETREE_NUM_SYMBOLS]; - uint i; - uint len; + unsigned i; + unsigned len; int ret; /* Read the code lengths of the pretree codes. There are 20 lengths of @@ -226,10 +226,10 @@ static int lzx_read_code_lens(struct input_bitstream *istream, u8 lens[], * some number of the next lengths are all 0, or some number of * the next lengths are all equal to the next symbol in the * input. */ - uint tree_code; - uint num_zeroes; - uint code; - uint num_same; + unsigned tree_code; + unsigned num_zeroes; + unsigned code; + unsigned num_same; char value; ret = read_huffsym_using_pretree(istream, pretree_decode_table, @@ -307,16 +307,17 @@ static int lzx_read_code_lens(struct input_bitstream *istream, u8 lens[], * blocks, which contain this information in the header) */ static int lzx_read_block_header(struct input_bitstream *istream, - int *block_size_ret, int *block_type_ret, + unsigned *block_size_ret, + unsigned *block_type_ret, struct lzx_tables *tables, struct lru_queue *queue) { int ret; int block_type; - uint block_size; + unsigned block_size; int s; int i; - uint len; + unsigned len; int32_t R[3]; ret = bitstream_ensure_bits(istream, 4); @@ -492,14 +493,14 @@ static int lzx_decode_match(int main_element, int block_type, struct lru_queue *queue, struct input_bitstream *istream) { - uint length_header; - uint position_slot; - uint match_len; - uint match_offset; - uint additional_len; - uint num_extra_bits; - uint verbatim_bits; - uint aligned_bits; + unsigned length_header; + unsigned position_slot; + unsigned match_len; + unsigned match_offset; + unsigned additional_len; + unsigned num_extra_bits; + unsigned verbatim_bits; + unsigned aligned_bits; int ret; int i; u8 *match_dest; @@ -652,7 +653,7 @@ static int lzx_decode_match(int main_element, int block_type, * format as used in other file formats, where a bit is reserved for that * purpose. */ static void undo_call_insn_preprocessing(u8 uncompressed_data[], - uint uncompressed_data_len) + unsigned uncompressed_data_len) { int i = 0; int file_size = LZX_MAGIC_FILESIZE; @@ -704,8 +705,8 @@ static int lzx_decompress_block(int block_type, int block_size, u8 *window, struct lru_queue *queue, struct input_bitstream *istream) { - uint bytes_remaining; - uint main_element; + unsigned bytes_remaining; + unsigned main_element; int match_len; int ret; @@ -749,16 +750,16 @@ static int lzx_decompress_block(int block_type, int block_size, u8 *window, * * Return non-zero on failure. */ -int lzx_decompress(const void *compressed_data, uint compressed_len, - void *uncompressed_data, uint uncompressed_len) +int lzx_decompress(const void *compressed_data, unsigned compressed_len, + void *uncompressed_data, unsigned uncompressed_len) { - struct lzx_tables tables; - struct input_bitstream istream; - struct lru_queue queue; - uint bytes_remaining; + struct lzx_tables tables; + struct input_bitstream istream; + struct lru_queue queue; + unsigned bytes_remaining; + unsigned block_size; + unsigned block_type; int ret; - int block_size; - int block_type; LZX_DEBUG("lzx_decompress (compressed_data = %p, compressed_len = %d, " "uncompressed_data = %p, uncompressed_len = %d).", @@ -784,17 +785,17 @@ int lzx_decompress(const void *compressed_data, uint compressed_len, while (bytes_remaining != 0) { LZX_DEBUG("Reading block header."); - ret = lzx_read_block_header(&istream, &block_size, &block_type, - &tables, &queue); + ret = lzx_read_block_header(&istream, &block_size, + &block_type, &tables, &queue); if (ret != 0) return ret; - LZX_DEBUG("block_size = %d, bytes_remaining = %d.", + LZX_DEBUG("block_size = %u, bytes_remaining = %u", block_size, bytes_remaining); if (block_size > bytes_remaining) { ERROR("lzx_decompress(): Expected a block size of at " - "most %d bytes (found %d bytes)", + "most %u bytes (found %u bytes)", bytes_remaining, block_size); return 1; } @@ -836,7 +837,6 @@ int lzx_decompress(const void *compressed_data, uint compressed_len, if (bytes_remaining != 0) LZX_DEBUG("%d bytes remaining.", bytes_remaining); - } if (uncompressed_len >= 10) diff --git a/src/timestamp.h b/src/timestamp.h index 6b54954d..406f6a5d 100644 --- a/src/timestamp.h +++ b/src/timestamp.h @@ -4,6 +4,7 @@ #include "util.h" #include #include +#include #define intervals_per_second (1000000000ULL / 100ULL) #define intervals_per_microsecond (10) @@ -51,7 +52,7 @@ static inline u64 timespec_to_wim_timestamp(const struct timespec *ts) + (u64)ts->tv_nsec / nanoseconds_per_interval; } - extern u64 get_wim_timestamp(); +extern void wim_timestamp_to_str(u64 timestamp, char *buf, size_t len); #endif diff --git a/src/util.c b/src/util.c index 5509a434..01ba8f36 100644 --- a/src/util.c +++ b/src/util.c @@ -112,6 +112,8 @@ static const char *error_strings[] = { = "Success", [WIMLIB_ERR_ALREADY_LOCKED] = "The WIM is already locked for writing", + [WIMLIB_ERR_CHAR_CONVERSION] + = "Failed to perform a conversion between UTF-8 and UTF-16LE", [WIMLIB_ERR_COMPRESSED_LOOKUP_TABLE] = "Lookup table is compressed", [WIMLIB_ERR_DECOMPRESSION] @@ -152,6 +154,8 @@ static const char *error_strings[] = { = "An entry in the WIM's lookup table is invalid", [WIMLIB_ERR_INVALID_PARAM] = "An invalid parameter was given", + [WIMLIB_ERR_INVALID_PART_NUMBER] + = "The part number or total parts of the WIM is invalid", [WIMLIB_ERR_INVALID_RESOURCE_HASH] = "The SHA1 message digest of a WIM resource did not match the expected value", [WIMLIB_ERR_INVALID_RESOURCE_SIZE] @@ -533,4 +537,10 @@ u64 get_wim_timestamp() return timeval_to_wim_timestamp(&tv); } - +void wim_timestamp_to_str(u64 timestamp, char *buf, size_t len) +{ + struct tm tm; + time_t t = wim_timestamp_to_unix(timestamp); + gmtime_r(&t, &tm); + strftime(buf, len, "%a %b %d %H:%M:%S %Y UTC", &tm); +} diff --git a/src/wim.c b/src/wim.c index 117361c1..cbda0061 100644 --- a/src/wim.c +++ b/src/wim.c @@ -28,6 +28,7 @@ #include #include #include +#include #include "dentry.h" #include @@ -433,7 +434,7 @@ static int begin_read(WIMStruct *w, const char *in_wim_path, int open_flags, wimlib_progress_func_t progress_func) { int ret; - uint xml_num_images; + int xml_num_images; DEBUG("Reading the WIM file `%s'", in_wim_path); @@ -456,7 +457,10 @@ static int begin_read(WIMStruct *w, const char *in_wim_path, int open_flags, w->filename = realpath(in_wim_path, NULL); if (!w->filename) { ERROR_WITH_ERRNO("Failed to resolve WIM filename"); - ret = WIMLIB_ERR_NOMEM; + if (errno == ENOMEM) + ret = WIMLIB_ERR_NOMEM; + else + ret = WIMLIB_ERR_OPEN; goto out_close; } @@ -464,7 +468,7 @@ static int begin_read(WIMStruct *w, const char *in_wim_path, int open_flags, if (ret != 0) goto out_close; - DEBUG("Wim file contains %u images", w->hdr.image_count); + DEBUG("According to header, WIM contains %u images", w->hdr.image_count); /* If the boot index is invalid, print a warning and set it to 0 */ if (w->hdr.boot_idx > w->hdr.image_count) { @@ -564,7 +568,7 @@ static int begin_read(WIMStruct *w, const char *in_wim_path, int open_flags, } DEBUG("Done beginning read of WIM file `%s'.", in_wim_path); - return 0; + /*return 0;*/ // // Everything is freed in wimlib_free() anyway, so no need to roll back @@ -585,7 +589,6 @@ out_free_lookup_table: out_close: /*fclose(w->fp);*/ /*w->fp = NULL;*/ -out: return ret; } diff --git a/src/wimlib.h b/src/wimlib.h index a547d3a8..084f7049 100644 --- a/src/wimlib.h +++ b/src/wimlib.h @@ -678,6 +678,7 @@ typedef int (*wimlib_progress_func_t)(enum wimlib_progress_msg msg_type, enum wimlib_error_code { WIMLIB_ERR_SUCCESS = 0, WIMLIB_ERR_ALREADY_LOCKED, + WIMLIB_ERR_CHAR_CONVERSION, WIMLIB_ERR_COMPRESSED_LOOKUP_TABLE, WIMLIB_ERR_DECOMPRESSION, WIMLIB_ERR_DELETE_STAGING_DIR, @@ -696,6 +697,7 @@ enum wimlib_error_code { WIMLIB_ERR_INVALID_INTEGRITY_TABLE, WIMLIB_ERR_INVALID_LOOKUP_TABLE_ENTRY, WIMLIB_ERR_INVALID_PARAM, + WIMLIB_ERR_INVALID_PART_NUMBER, WIMLIB_ERR_INVALID_RESOURCE_HASH, WIMLIB_ERR_INVALID_RESOURCE_SIZE, WIMLIB_ERR_INVALID_SECURITY_DATA, diff --git a/src/xml.c b/src/xml.c index 9a0bded2..a1e6f596 100644 --- a/src/xml.c +++ b/src/xml.c @@ -23,16 +23,18 @@ * along with wimlib; if not, see http://www.gnu.org/licenses/. */ -#include "wimlib_internal.h" #include "dentry.h" -#include "xml.h" -#include "timestamp.h" #include "lookup_table.h" +#include "timestamp.h" +#include "wimlib_internal.h" +#include "xml.h" + #include -#include #include #include #include +#include +#include /* Structures used to form an in-memory representation of the XML data (other * than the raw parse tree from libxml). */ @@ -55,28 +57,28 @@ struct windows_info { char *product_suite; char **languages; char *default_language; - u64 num_languages; + size_t num_languages; char *system_root; bool windows_version_exists; struct windows_version windows_version; }; struct image_info { - u64 index; - u64 dir_count; - u64 file_count; - u64 total_bytes; - u64 hard_link_bytes; - u64 creation_time; - u64 last_modification_time; - bool windows_info_exists; + int index; + bool windows_info_exists; + u64 dir_count; + u64 file_count; + u64 total_bytes; + u64 hard_link_bytes; + u64 creation_time; + u64 last_modification_time; struct windows_info windows_info; char *name; char *description; - char *display_name; - char *display_description; + char *display_name; + char *display_description; union { - char *flags; + char *flags; struct lookup_table *lookup_table; }; }; @@ -117,11 +119,6 @@ static inline bool node_is_text(xmlNode *node) return node->type == XML_TEXT_NODE; } -static inline bool node_is_attribute(xmlNode *node) -{ - return node->type == XML_ATTRIBUTE_NODE; -} - static inline bool node_name_is(xmlNode *node, const char *name) { /* For now, both upper case and lower case element names are accepted. */ @@ -158,10 +155,8 @@ static int node_get_string(const xmlNode *string_node, char **str) for_node_child(string_node, child) { if (node_is_text(child) && child->content) { p = STRDUP(child->content); - if (!p) { - ERROR("Out of memory"); + if (!p) return WIMLIB_ERR_NOMEM; - } break; } } @@ -190,8 +185,8 @@ static u64 node_get_timestamp(const xmlNode *time_node) /* Used to sort an array of struct image_infos by their image indices. */ static int sort_by_index(const void *p1, const void *p2) { - u64 index_1 = ((struct image_info*)p1)->index; - u64 index_2 = ((struct image_info*)p1)->index; + int index_1 = ((struct image_info*)p1)->index; + int index_2 = ((struct image_info*)p1)->index; if (index_1 < index_2) return -1; else if (index_1 > index_2) @@ -204,15 +199,13 @@ static int sort_by_index(const void *p1, const void *p2) /* Frees memory allocated inside a struct windows_info structure. */ static void destroy_windows_info(struct windows_info *windows_info) { - uint i; - FREE(windows_info->product_name); FREE(windows_info->edition_id); FREE(windows_info->installation_type); FREE(windows_info->hal); FREE(windows_info->product_type); FREE(windows_info->product_suite); - for (i = 0; i < windows_info->num_languages; i++) + for (size_t i = 0; i < windows_info->num_languages; i++) FREE(windows_info->languages[i]); FREE(windows_info->languages); FREE(windows_info->default_language); @@ -233,10 +226,9 @@ static void destroy_image_info(struct image_info *image_info) void free_wim_info(struct wim_info *info) { - uint i; if (info) { if (info->images) { - for (i = 0; i < info->num_images; i++) + for (int i = 0; i < info->num_images; i++) destroy_image_info(&info->images[i]); FREE(info->images); } @@ -254,11 +246,11 @@ static void xml_read_windows_version(const xmlNode *version_node, if (!node_is_element(child)) continue; if (node_name_is(child, "MAJOR")) - windows_version->major = node_get_u64(child); + windows_version->major = node_get_u64(child); else if (node_name_is(child, "MINOR")) - windows_version->minor = node_get_u64(child); + windows_version->minor = node_get_u64(child); else if (node_name_is(child, "BUILD")) - windows_version->build = node_get_u64(child); + windows_version->build = node_get_u64(child); else if (node_name_is(child, "SPBUILD")) windows_version->sp_build = node_get_u64(child); else if (node_name_is(child, "SPLEVEL")) @@ -270,40 +262,35 @@ static void xml_read_windows_version(const xmlNode *version_node, * */ static int xml_read_languages(const xmlNode *languages_node, char ***languages_ret, - u64 *num_languages_ret, + size_t *num_languages_ret, char **default_language_ret) { xmlNode *child; - uint i; - uint num_languages; + size_t num_languages = 0; char **languages; int ret; - num_languages = 0; for_node_child(languages_node, child) if (node_is_element(child) && node_name_is(child, "LANGUAGE")) num_languages++; - languages = CALLOC(num_languages, sizeof(char*)); - if (!languages) { - ERROR("Out of memory"); + languages = CALLOC(num_languages, sizeof(languages[0])); + if (!languages) return WIMLIB_ERR_NOMEM; - } *languages_ret = languages; *num_languages_ret = num_languages; - i = 0; ret = 0; for_node_child(languages_node, child) { if (!node_is_element(child)) continue; if (node_name_is(child, "LANGUAGE")) - ret = node_get_string(child, &languages[i++]); + ret = node_get_string(child, languages++); else if (node_name_is(child, "DEFAULT")) ret = node_get_string(child, default_language_ret); if (ret != 0) - return ret; + break; } return ret; } @@ -349,7 +336,6 @@ static int xml_read_windows_info(const xmlNode *windows_node, } else if (node_name_is(child, "HAL")) { ret = node_get_string(child, &windows_info->hal); } - if (ret != 0) return ret; } @@ -366,11 +352,10 @@ static int xml_read_image_info(xmlNode *image_node, index_prop = xmlGetProp(image_node, "INDEX"); if (index_prop) { - char *tmp; - image_info->index = strtoul(index_prop, &tmp, 10); + image_info->index = atoi(index_prop); FREE(index_prop); } else { - image_info->index = 0; + image_info->index = 1; } ret = 0; @@ -409,14 +394,13 @@ static int xml_read_image_info(xmlNode *image_node, return ret; } if (!image_info->name) { - WARNING("Image with index %"PRIu64" has no name", - image_info->index); - image_info->name = MALLOC(1); - if (!image_info->name) { - ERROR("Out of memory"); + char *empty_name; + WARNING("Image with index %d has no name", image_info->index); + empty_name = MALLOC(1); + if (!empty_name) return WIMLIB_ERR_NOMEM; - } - image_info->name[0] = '\0'; + *empty_name = '\0'; + image_info->name = empty_name; } return ret; } @@ -429,50 +413,63 @@ static int xml_read_wim_info(const xmlNode *wim_node, struct wim_info *wim_info; xmlNode *child; int ret; - uint num_images; - struct image_info *cur_image_info; + int num_images; + int i; wim_info = CALLOC(1, sizeof(struct wim_info)); - if (!wim_info) { - ERROR("Out of memory"); + if (!wim_info) return WIMLIB_ERR_NOMEM; - } /* Count how many images there are. */ num_images = 0; - for_node_child(wim_node, child) - if (node_is_element(child) && node_name_is(child, "IMAGE")) + for_node_child(wim_node, child) { + if (node_is_element(child) && node_name_is(child, "IMAGE")) { + if (num_images == INT_MAX) { + return WIMLIB_ERR_IMAGE_COUNT; + } num_images++; + } + } - if (num_images == 0) - goto done; + if (num_images > 0) { + /* Allocate the array of struct image_infos and fill them in. */ + wim_info->images = CALLOC(num_images, sizeof(wim_info->images[0])); + if (!wim_info->images) { + ret = WIMLIB_ERR_NOMEM; + goto err; + } + wim_info->num_images = num_images; + i = 0; + for_node_child(wim_node, child) { + if (!node_is_element(child)) + continue; + if (node_name_is(child, "IMAGE")) { + DEBUG("Found tag"); + ret = xml_read_image_info(child, + &wim_info->images[i]); + if (ret != 0) + goto err; + i++; + } else if (node_name_is(child, "TOTALBYTES")) { + wim_info->total_bytes = node_get_u64(child); + } + } - /* Allocate the array of struct image_infos and fill them in. */ - wim_info->images = CALLOC(num_images, sizeof(wim_info->images[0])); - if (!wim_info->images) { - ret = WIMLIB_ERR_NOMEM; - ERROR("Out of memory!"); - goto err; - } - wim_info->num_images = num_images; - cur_image_info = wim_info->images; - for_node_child(wim_node, child) { - if (!node_is_element(child)) - continue; - if (node_name_is(child, "IMAGE")) { - DEBUG("Found tag"); - ret = xml_read_image_info(child, cur_image_info++); - if (ret != 0) - goto err; - } else if (node_name_is(child, "TOTALBYTES")) { - wim_info->total_bytes = node_get_u64(child); + /* Sort the array of image info by image index. */ + qsort(wim_info->images, num_images, + sizeof(struct image_info), sort_by_index); + + /* Make sure the image indices make sense */ + for (i = 0; i < num_images; i++) { + if (wim_info->images[i].index != i + 1) { + ERROR("WIM images are not indexed [1...%d] " + "in XML data as expected", + num_images); + return WIMLIB_ERR_IMAGE_COUNT; + } } - } - /* Sort the array of struct image_infos by image index. */ - qsort(wim_info->images, wim_info->num_images, - sizeof(struct image_info), sort_by_index); -done: + } *wim_info_ret = wim_info; return 0; err: @@ -480,10 +477,9 @@ err: return ret; } -/* Prints the information contained in a struct windows_info structure. */ +/* Prints the information contained in a `struct windows_info'. */ static void print_windows_info(const struct windows_info *windows_info) { - uint i; const struct windows_version *windows_version; printf("Architecture: %s\n", get_arch(windows_info->arch)); @@ -501,7 +497,8 @@ static void print_windows_info(const struct windows_info *windows_info) windows_info->installation_type); if (windows_info->hal) - printf("HAL: %s\n", windows_info->hal); + printf("HAL: %s\n", + windows_info->hal); if (windows_info->product_type) printf("Product Type: %s\n", @@ -510,8 +507,9 @@ static void print_windows_info(const struct windows_info *windows_info) if (windows_info->product_suite) printf("Product Suite: %s\n", windows_info->product_suite); + printf("Languages: "); - for (i = 0; i < windows_info->num_languages; i++) { + for (size_t i = 0; i < windows_info->num_languages; i++) { fputs(windows_info->languages[i], stdout); putchar(' '); } @@ -525,22 +523,22 @@ static void print_windows_info(const struct windows_info *windows_info) if (windows_info->windows_version_exists) { windows_version = &windows_info->windows_version; printf("Major Version: %"PRIu64"\n", - windows_version->major); + windows_version->major); printf("Minor Version: %"PRIu64"\n", - windows_version->minor); + windows_version->minor); printf("Build: %"PRIu64"\n", - windows_version->build); + windows_version->build); printf("Service Pack Build: %"PRIu64"\n", - windows_version->sp_build); + windows_version->sp_build); printf("Service Pack Level: %"PRIu64"\n", - windows_version->sp_level); + windows_version->sp_level); } } -/* Writes the information contained in a struct windows_version structure to the XML - * document being constructed in memory. This is the element inside - * the element. */ +/* Writes the information contained in a `struct windows_version' to the XML + * document being written. This is the element inside the + * element. */ static int xml_write_windows_version(xmlTextWriter *writer, const struct windows_version *version) { @@ -550,35 +548,35 @@ static int xml_write_windows_version(xmlTextWriter *writer, return rc; rc = xmlTextWriterWriteFormatElement(writer, "MAJOR", "%"PRIu64, - version->major); + version->major); if (rc < 0) return rc; rc = xmlTextWriterWriteFormatElement(writer, "MINOR", "%"PRIu64, - version->minor); + version->minor); if (rc < 0) return rc; rc = xmlTextWriterWriteFormatElement(writer, "BUILD", "%"PRIu64, - version->build); + version->build); if (rc < 0) return rc; rc = xmlTextWriterWriteFormatElement(writer, "SPBUILD", "%"PRIu64, - version->sp_build); + version->sp_build); if (rc < 0) return rc; rc = xmlTextWriterWriteFormatElement(writer, "SPLEVEL", "%"PRIu64, - version->sp_level); + version->sp_level); if (rc < 0) return rc; return xmlTextWriterEndElement(writer); /* */ } -/* Writes the information contained in a struct windows_info structure to the XML - * document being constructed in memory. This is the element. */ +/* Writes the information contained in a `struct windows_info' to the XML + * document being written. This is the element. */ static int xml_write_windows_info(xmlTextWriter *writer, const struct windows_info *windows_info) { @@ -587,52 +585,51 @@ static int xml_write_windows_info(xmlTextWriter *writer, if (rc < 0) return rc; - rc = xmlTextWriterWriteFormatElement(writer, "ARCH", "%"PRIu64, - windows_info->arch); + windows_info->arch); if (rc < 0) return rc; if (windows_info->product_name) { rc = xmlTextWriterWriteElement(writer, "PRODUCTNAME", - windows_info->product_name); + windows_info->product_name); if (rc < 0) return rc; } if (windows_info->edition_id) { rc = xmlTextWriterWriteElement(writer, "EDITIONID", - windows_info->edition_id); + windows_info->edition_id); if (rc < 0) return rc; } if (windows_info->installation_type) { rc = xmlTextWriterWriteElement(writer, "INSTALLATIONTYPE", - windows_info->installation_type); + windows_info->installation_type); if (rc < 0) return rc; } if (windows_info->hal) { rc = xmlTextWriterWriteElement(writer, "HAL", - windows_info->hal); + windows_info->hal); if (rc < 0) return rc; } if (windows_info->product_type) { rc = xmlTextWriterWriteElement(writer, "PRODUCTTYPE", - windows_info->product_type); + windows_info->product_type); if (rc < 0) return rc; } if (windows_info->product_suite) { rc = xmlTextWriterWriteElement(writer, "PRODUCTSUITE", - windows_info->product_suite); - if (rc < 0) - return rc; + windows_info->product_suite); + if (rc < 0) + return rc; } if (windows_info->num_languages) { @@ -640,14 +637,14 @@ static int xml_write_windows_info(xmlTextWriter *writer, if (rc < 0) return rc; - for (int i = 0; i < windows_info->num_languages; i++) { + for (size_t i = 0; i < windows_info->num_languages; i++) { rc = xmlTextWriterWriteElement(writer, "LANGUAGE", - windows_info->languages[i]); + windows_info->languages[i]); if (rc < 0) return rc; } rc = xmlTextWriterWriteElement(writer, "DEFAULT", - windows_info->default_language); + windows_info->default_language); if (rc < 0) return rc; @@ -664,9 +661,9 @@ static int xml_write_windows_info(xmlTextWriter *writer, if (windows_info->system_root) { rc = xmlTextWriterWriteElement(writer, "SYSTEMROOT", - windows_info->system_root); - if (rc < 0) - return rc; + windows_info->system_root); + if (rc < 0) + return rc; } return xmlTextWriterEndElement(writer); /* */ @@ -697,7 +694,6 @@ static int xml_write_time(xmlTextWriter *writer, const char *element_name, return 0; } - /* Writes an element to the XML document. */ static int xml_write_image_info(xmlTextWriter *writer, const struct image_info *image_info) @@ -707,38 +703,37 @@ static int xml_write_image_info(xmlTextWriter *writer, if (rc < 0) return rc; - rc = xmlTextWriterWriteFormatAttribute(writer, "INDEX", "%"PRIu64, - image_info->index); + rc = xmlTextWriterWriteFormatAttribute(writer, "INDEX", "%d", + image_info->index); if (rc < 0) return rc; rc = xmlTextWriterWriteFormatElement(writer, "DIRCOUNT", "%"PRIu64, - image_info->dir_count); + image_info->dir_count); if (rc < 0) return rc; rc = xmlTextWriterWriteFormatElement(writer, "FILECOUNT", "%"PRIu64, - image_info->file_count); + image_info->file_count); if (rc < 0) return rc; rc = xmlTextWriterWriteFormatElement(writer, "TOTALBYTES", "%"PRIu64, - image_info->total_bytes); + image_info->total_bytes); if (rc < 0) return rc; rc = xmlTextWriterWriteFormatElement(writer, "HARDLINKBYTES", "%"PRIu64, - image_info->hard_link_bytes); + image_info->hard_link_bytes); if (rc < 0) return rc; - rc = xml_write_time(writer, "CREATIONTIME", - image_info->creation_time); + rc = xml_write_time(writer, "CREATIONTIME", image_info->creation_time); if (rc < 0) return rc; rc = xml_write_time(writer, "LASTMODIFICATIONTIME", - image_info->last_modification_time); + image_info->last_modification_time); if (rc < 0) return rc; @@ -746,37 +741,37 @@ static int xml_write_image_info(xmlTextWriter *writer, rc = xml_write_windows_info(writer, &image_info->windows_info); if (rc < 0) return rc; - } else { - DEBUG(" tag does not exist."); } if (image_info->name) { - rc = xmlTextWriterWriteElement(writer, "NAME", image_info->name); + rc = xmlTextWriterWriteElement(writer, "NAME", + image_info->name); if (rc < 0) return rc; } + if (image_info->description) { rc = xmlTextWriterWriteElement(writer, "DESCRIPTION", - image_info->description); + image_info->description); if (rc < 0) return rc; } if (image_info->display_name) { rc = xmlTextWriterWriteElement(writer, "DISPLAYNAME", - image_info->display_name); + image_info->display_name); if (rc < 0) return rc; } if (image_info->display_description) { rc = xmlTextWriterWriteElement(writer, "DISPLAYDESCRIPTION", - image_info->display_description); + image_info->display_description); if (rc < 0) return rc; } if (image_info->flags) { rc = xmlTextWriterWriteElement(writer, "FLAGS", - image_info->flags); + image_info->flags); if (rc < 0) return rc; } @@ -806,8 +801,6 @@ static struct image_info *add_image_info_struct(struct wim_info *wim_info) static int clone_windows_info(const struct windows_info *old, struct windows_info *new) { - uint i; - if (old->product_name && !(new->product_name = STRDUP(old->product_name))) return WIMLIB_ERR_NOMEM; if (old->edition_id && !(new->edition_id = STRDUP(old->edition_id))) @@ -827,7 +820,7 @@ static int clone_windows_info(const struct windows_info *old, if (!new->languages) return WIMLIB_ERR_NOMEM; new->num_languages = old->num_languages; - for (i = 0; i < new->num_languages; i++) { + for (size_t i = 0; i < new->num_languages; i++) { if (!old->languages[i]) continue; new->languages[i] = STRDUP(old->languages[i]); @@ -905,6 +898,7 @@ int xml_export_image(const struct wim_info *old_wim_info, DEBUG("Copying XML data between WIM files for source image %d.", image); + wimlib_assert(old_wim_info != NULL); wimlib_assert(image >= 1 && image <= old_wim_info->num_images); if (*new_wim_info_p) { @@ -921,7 +915,7 @@ int xml_export_image(const struct wim_info *old_wim_info, ret = clone_image_info(&old_wim_info->images[image - 1], image_info); if (ret != 0) - goto err; + goto err_destroy_image_info; image_info->index = new_wim_info->num_images; @@ -929,19 +923,21 @@ int xml_export_image(const struct wim_info *old_wim_info, FREE(image_info->name); image_info->name = STRDUP(dest_image_name); if (!image_info->name) - goto err; + goto err_destroy_image_info; } if (dest_image_description) { FREE(image_info->description); image_info->description = STRDUP(dest_image_description); if (!image_info->description) - goto err; + goto err_destroy_image_info; } *new_wim_info_p = new_wim_info; return 0; +err_destroy_image_info: + destroy_image_info(image_info); err: - ERROR("Out of memory"); - free_wim_info(new_wim_info); + if (new_wim_info != *new_wim_info_p) + free_wim_info(new_wim_info); return WIMLIB_ERR_NOMEM; } @@ -949,37 +945,41 @@ err: void xml_delete_image(struct wim_info **wim_info_p, int image) { struct wim_info *wim_info; - int i; DEBUG("Deleting image %d from the XML data.", image); wim_info = *wim_info_p; - wimlib_assert(wim_info); + wimlib_assert(wim_info != NULL); wimlib_assert(image >= 1 && image <= wim_info->num_images); destroy_image_info(&wim_info->images[image - 1]); - for (i = image - 1; i < wim_info->num_images - 1; i++) { - memcpy(&wim_info->images[i], &wim_info->images[i + 1], - sizeof(struct image_info)); - wim_info->images[i].index--; - } + memcpy(&wim_info->images[image - 1], + &wim_info->images[image], + (wim_info->num_images - image) * sizeof(struct image_info)); if (--wim_info->num_images == 0) { free_wim_info(wim_info); *wim_info_p = NULL; + } else { + for (int i = image - 1; i < wim_info->num_images; i++) + wim_info->images[i].index--; } } size_t xml_get_max_image_name_len(const WIMStruct *w) { - size_t len = 0; - uint i; - uint num_images = w->wim_info->num_images; - for (i = 0; i < num_images; i++) - len = max(len, strlen(w->wim_info->images[i].name)); - return len; + size_t max_len = 0; + if (w->wim_info) { + size_t len; + for (int i = 0; i < w->wim_info->num_images; i++) { + len = strlen(w->wim_info->images[i].name); + if (len > max_len) + max_len = len; + } + } + return max_len; } #ifdef ENABLE_CUSTOM_MEMORY_ALLOCATOR @@ -1062,6 +1062,13 @@ static int calculate_dentry_statistics(struct dentry *dentry, void *arg) return 0; } +/* + * Calculate what to put in the , , , and + * elements of each . + * + * Please note there is no official documentation for exactly how this is done. + * But, see calculate_dentry_statistics(). + */ void xml_update_image_info(WIMStruct *w, int image) { struct image_info *image_info; @@ -1078,12 +1085,9 @@ void xml_update_image_info(WIMStruct *w, int image) flags_save = image_info->flags; image_info->lookup_table = w->lookup_table; - for_dentry_in_tree(w->image_metadata[image - 1].root_dentry, calculate_dentry_statistics, image_info); - - image_info->lookup_table = NULL; image_info->flags = flags_save; image_info->last_modification_time = get_wim_timestamp(); } @@ -1094,22 +1098,16 @@ int xml_add_image(WIMStruct *w, const char *name) struct wim_info *wim_info; struct image_info *image_info; - wimlib_assert(name); - - DEBUG("Adding image: name = %s", name); + wimlib_assert(name != NULL); /* If this is the first image, allocate the struct wim_info. Otherwise * use the existing struct wim_info. */ if (w->wim_info) { wim_info = w->wim_info; } else { - DEBUG("Allocing struct wim_info with 1 image"); wim_info = CALLOC(1, sizeof(struct wim_info)); - if (!wim_info) { - ERROR("Could not allocate WIM information struct--- " - "out of memory"); + if (!wim_info) return WIMLIB_ERR_NOMEM; - } } image_info = add_image_info_struct(wim_info); @@ -1131,7 +1129,6 @@ out_destroy_image_info: out_free_wim_info: if (wim_info != w->wim_info) FREE(wim_info); - ERROR("Out of memory"); return WIMLIB_ERR_NOMEM; } @@ -1141,14 +1138,13 @@ void print_image_info(const struct wim_info *wim_info, int image) { const struct image_info *image_info; const char *desc; - time_t time; - char *p; + char buf[50]; wimlib_assert(image >= 1 && image <= wim_info->num_images); image_info = &wim_info->images[image - 1]; - printf("Index: %"PRIu64"\n", image_info->index); + printf("Index: %d\n", image_info->index); printf("Name: %s\n", image_info->name); /* Always print the Description: part even if there is no @@ -1172,17 +1168,11 @@ void print_image_info(const struct wim_info *wim_info, int image) printf("Total Bytes: %"PRIu64"\n", image_info->total_bytes); printf("Hard Link Bytes: %"PRIu64"\n", image_info->hard_link_bytes); - time = wim_timestamp_to_unix(image_info->creation_time); - p = asctime(gmtime(&time)); - *(strrchr(p, '\n')) = '\0'; - - printf("Creation Time: %s UTC\n", p); + wim_timestamp_to_str(image_info->creation_time, buf, sizeof(buf)); + printf("Creation Time: %s\n", buf); - time = wim_timestamp_to_unix(image_info->last_modification_time); - p = asctime(gmtime(&time)); - *(strrchr(p, '\n')) = '\0'; - - printf("Last Modification Time: %s UTC\n", p); + wim_timestamp_to_str(image_info->creation_time, buf, sizeof(buf)); + printf("Last Modification Time: %s\n", buf); if (image_info->windows_info_exists) print_windows_info(&image_info->windows_info); if (image_info->flags) @@ -1193,7 +1183,7 @@ void print_image_info(const struct wim_info *wim_info, int image) /* * Reads the XML data from a WIM file. */ -int read_xml_data(FILE *fp, const struct resource_entry *res, +int read_xml_data(FILE *fp, const struct resource_entry *res_entry, u8 **xml_data_ret, struct wim_info **info_ret) { u8 *xml_data; @@ -1202,35 +1192,39 @@ int read_xml_data(FILE *fp, const struct resource_entry *res, int ret; DEBUG("XML data is %"PRIu64" bytes at offset %"PRIu64"", - (u64)res->size, res->offset); + (u64)res_entry->size, res_entry->offset); - if (resource_is_compressed(res)) { + if (resource_is_compressed(res_entry)) { ERROR("XML data is supposed to be uncompressed"); ret = WIMLIB_ERR_XML; goto out_cleanup_parser; } - if (res->size < 2) { - ERROR("XML data must be at least 2 bytes"); + + if (res_entry->size < 2) { + ERROR("XML data must be at least 2 bytes long"); ret = WIMLIB_ERR_XML; goto out_cleanup_parser; } - xml_data = MALLOC(res->size + 2); + xml_data = MALLOC(res_entry->size + 2); if (!xml_data) { ret = WIMLIB_ERR_NOMEM; goto out_cleanup_parser; } - ret = read_uncompressed_resource(fp, res->offset, res->size, xml_data); + + ret = read_uncompressed_resource(fp, res_entry->offset, + res_entry->size, xml_data); if (ret != 0) goto out_free_xml_data; - xml_data[res->size] = 0; - xml_data[res->size + 1] = 0; - - DEBUG("Parsing XML using libxml2 to create XML tree."); + /* Null-terminate just in case */ + xml_data[res_entry->size] = 0; + xml_data[res_entry->size + 1] = 0; - doc = xmlReadMemory(xml_data, res->size, "noname.xml", "UTF-16", 0); + DEBUG("Parsing XML using libxml2 to create XML tree"); + doc = xmlReadMemory(xml_data, res_entry->size, + "noname.xml", "UTF-16", 0); if (!doc) { ERROR("Failed to parse XML data"); @@ -1242,7 +1236,7 @@ int read_xml_data(FILE *fp, const struct resource_entry *res, root = xmlDocGetRootElement(doc); if (!root) { - ERROR("Empty XML document"); + ERROR("WIM XML data is an empty XML document"); ret = WIMLIB_ERR_XML; goto out_free_doc; } @@ -1260,10 +1254,8 @@ int read_xml_data(FILE *fp, const struct resource_entry *res, DEBUG("Freeing XML tree."); - xmlFreeDoc(doc); - xmlCleanupParser(); *xml_data_ret = xml_data; - return 0; + xml_data = NULL; out_free_doc: xmlFreeDoc(doc); out_free_xml_data: @@ -1289,15 +1281,12 @@ out_cleanup_parser: int write_xml_data(const struct wim_info *wim_info, int image, FILE *out, u64 total_bytes, struct resource_entry *out_res_entry) { - xmlBuffer *buf; + xmlCharEncodingHandler *encoding_handler; + xmlOutputBuffer *out_buffer; xmlTextWriter *writer; - char *utf16_str; int ret; - const xmlChar *content; - size_t len; - size_t utf16_len; - size_t bytes_written; - off_t start_offset, end_offset; + off_t start_offset; + off_t end_offset; wimlib_assert(image == WIMLIB_ALL_IMAGES || (wim_info != NULL && image >= 1 && @@ -1307,42 +1296,55 @@ int write_xml_data(const struct wim_info *wim_info, int image, FILE *out, if (start_offset == -1) return WIMLIB_ERR_WRITE; + DEBUG("Writing XML data for image %d at offset %"PRIu64, + image, start_offset); + + /* 2 bytes endianness marker for UTF-16LE. This is _required_ for WIM + * XML data. */ + if ((putc(0xff, out)) == EOF || (putc(0xfe, out) == EOF)) { + ERROR_WITH_ERRNO("Error writing XML data"); + return WIMLIB_ERR_WRITE; + } + /* The contents of the element in the XML data, under the - * element not the element, is (for non-spit WIMs) the - * size of the WIM file excluding the XML data and integrity table, - * which is the current offset, since the XML data goes at the end of - * the WIM file before the integrity table. */ - if (total_bytes == 0) { - total_bytes = ftello(out); - if (total_bytes == (u64)-1) - return WIMLIB_ERR_WRITE; + * element (not the element), is for non-split WIMs the + * size of the WIM file excluding the XML data and integrity table. + * This should be equal to the current position in the output stream, + * since the XML data and integrity table are the last elements of the + * WIM. + * + * For split WIMs, takes into account the entire WIM, not + * just the current part. In that case, @total_bytes should be passed + * in to this function. */ + if (total_bytes == 0) + total_bytes = start_offset; + + xmlInitCharEncodingHandlers(); + + /* The encoding of the XML data must be UTF-16LE. */ + encoding_handler = xmlGetCharEncodingHandler(XML_CHAR_ENCODING_UTF16LE); + if (!encoding_handler) { + ERROR("Failed to get XML character encoding handler for UTF-16LE"); + ret = WIMLIB_ERR_CHAR_CONVERSION; + goto out_cleanup_char_encoding_handlers; } - DEBUG("Creating XML buffer and text writer."); - buf = xmlBufferCreate(); - if (!buf) { - ERROR("Failed to allocate XML buffer"); + out_buffer = xmlOutputBufferCreateFile(out, encoding_handler); + if (!out_buffer) { + ERROR("Failed to allocate xmlOutputBuffer"); ret = WIMLIB_ERR_NOMEM; - goto out; + goto out_cleanup_char_encoding_handlers; } - writer = xmlNewTextWriterMemory(buf, 0); + + writer = xmlNewTextWriter(out_buffer); if (!writer) { - ERROR("Failed to allocate XML writer"); + ERROR("Failed to allocate xmlTextWriter"); ret = WIMLIB_ERR_NOMEM; - goto out_buffer_free; + goto out_output_buffer_close; } - /* XXX */ - /* M$'s WIM files do not have XML declarations, so do not write one. - * I'm not sure how we can force the document to be written in UTF-16 - * without calling xmlTextWriterStartDocument(), though, so currently it - * is composed in a buffer UTF-8, then converted to UTF-16. */ -#if 0 - ret = xmlTextWriterStartDocument(writer, NULL, "UTF-16", NULL); - CHECK_RET; -#endif - DEBUG("Writing element"); + ret = xmlTextWriterStartElement(writer, "WIM"); CHECK_RET; @@ -1351,11 +1353,16 @@ int write_xml_data(const struct wim_info *wim_info, int image, FILE *out, CHECK_RET; if (wim_info != NULL) { - DEBUG("Writing %d elements", (int)wim_info->num_images); - for (int i = 1; i <= (int)wim_info->num_images; i++) { - if (image != WIMLIB_ALL_IMAGES && i != image) - continue; - DEBUG("Writing element for image %d", i); + int first, last; + if (image == WIMLIB_ALL_IMAGES) { + first = 1; + last = wim_info->num_images; + } else { + first = image; + last = image; + } + DEBUG("Writing %d elements", last - first + 1); + for (int i = first; i <= last; i++) { ret = xml_write_image_info(writer, &wim_info->images[i - 1]); CHECK_RET; } @@ -1367,62 +1374,43 @@ int write_xml_data(const struct wim_info *wim_info, int image, FILE *out, ret = xmlTextWriterEndDocument(writer); CHECK_RET; + DEBUG("Ended XML document"); + + /* Call xmlFreeTextWriter() before ftello() because the former will + * flush the file stream. */ xmlFreeTextWriter(writer); writer = NULL; - DEBUG("Done composing XML document. Now converting to UTF-16 and " - "writing it to the output file."); - - content = xmlBufferContent(buf); - len = xmlBufferLength(buf); - - DEBUG("XML UTF-8 length = %zu", len); - - utf16_str = utf8_to_utf16(content, len, &utf16_len); - if (!utf16_str) { - ret = WIMLIB_ERR_NOMEM; - goto out_free_text_writer; - } - - DEBUG("XML UTF-16 length = %zu", utf16_len); - - if ((putc(0xff, out)) == EOF || (putc(0xfe, out) == EOF) || - ((bytes_written = fwrite(utf16_str, 1, utf16_len, out)) - != utf16_len)) { - ERROR_WITH_ERRNO("Error writing XML data"); - ret = WIMLIB_ERR_WRITE; - goto out_free_utf16_str; - } - - DEBUG("Cleaning up."); end_offset = ftello(out); if (end_offset == -1) { ret = WIMLIB_ERR_WRITE; - goto out_free_utf16_str; + } else { + ret = 0; + out_res_entry->offset = start_offset; + out_res_entry->size = end_offset - start_offset; + out_res_entry->original_size = end_offset - start_offset; + out_res_entry->flags = WIM_RESHDR_FLAG_METADATA; } - - out_res_entry->offset = start_offset; - out_res_entry->size = end_offset - start_offset; - out_res_entry->original_size = end_offset - start_offset; - out_res_entry->flags = WIM_RESHDR_FLAG_METADATA; - ret = 0; -out_free_utf16_str: - FREE(utf16_str); out_free_text_writer: + /* xmlFreeTextWriter will free the attached xmlOutputBuffer. */ xmlFreeTextWriter(writer); -out_buffer_free: - xmlBufferFree(buf); + out_buffer = NULL; +out_output_buffer_close: + if (out_buffer != NULL) + xmlOutputBufferClose(out_buffer); +out_cleanup_char_encoding_handlers: + xmlCleanupCharEncodingHandlers(); out: + if (ret == 0) + DEBUG("Successfully wrote XML data"); return ret; } /* Returns the name of the specified image. */ WIMLIBAPI const char *wimlib_get_image_name(const WIMStruct *w, int image) { - DEBUG("Getting the name of image %d", image); if (image < 1 || image > w->hdr.image_count) return NULL; - return w->wim_info->images[image - 1].name; } @@ -1430,34 +1418,27 @@ WIMLIBAPI const char *wimlib_get_image_name(const WIMStruct *w, int image) WIMLIBAPI const char *wimlib_get_image_description(const WIMStruct *w, int image) { - DEBUG("Getting the description of image %d", image); if (image < 1 || image > w->hdr.image_count) return NULL; - return w->wim_info->images[image - 1].description; } /* Determines if an image name is already used by some image in the WIM. */ WIMLIBAPI bool wimlib_image_name_in_use(const WIMStruct *w, const char *name) { - int i; - - DEBUG("Checking to see if the image name `%s' is already in use", name); - if (!name || !*name || !w->wim_info) + if (!name || !*name) return false; - for (i = 1; i <= w->wim_info->num_images; i++) + for (int i = 1; i <= w->hdr.image_count; i++) if (strcmp(w->wim_info->images[i - 1].name, name) == 0) return true; - return false; } +/* Extracts the raw XML data to a file stream. */ WIMLIBAPI int wimlib_extract_xml_data(WIMStruct *w, FILE *fp) { - DEBUG("Extracting the XML data."); if (!w->xml_data) return WIMLIB_ERR_INVALID_PARAM; - if (fwrite(w->xml_data, 1, w->hdr.xml_res_entry.size, fp) != w->hdr.xml_res_entry.size) { ERROR_WITH_ERRNO("Failed to extract XML data"); @@ -1474,13 +1455,11 @@ WIMLIBAPI int wimlib_set_image_name(WIMStruct *w, int image, const char *name) DEBUG("Setting the name of image %d to %s", image, name); - if (!w) - return WIMLIB_ERR_INVALID_PARAM; - if (!name || !*name) { ERROR("Must specify a non-empty string for the image name"); return WIMLIB_ERR_INVALID_PARAM; } + if (image < 1 || image > w->hdr.image_count) { ERROR("%d is not a valid image", image); return WIMLIB_ERR_INVALID_IMAGE; @@ -1497,10 +1476,9 @@ WIMLIBAPI int wimlib_set_image_name(WIMStruct *w, int image, const char *name) } p = STRDUP(name); - if (!p) { - ERROR("Out of memory"); + if (!p) return WIMLIB_ERR_NOMEM; - } + FREE(w->wim_info->images[image - 1].name); w->wim_info->images[image - 1].name = p; return 0; @@ -1512,21 +1490,14 @@ WIMLIBAPI int wimlib_set_image_descripton(WIMStruct *w, int image, { char *p; - DEBUG("Setting the description of image %d to %s", image, description); - - if (!w) - return WIMLIB_ERR_INVALID_PARAM; - if (image < 1 || image > w->hdr.image_count) { ERROR("%d is not a valid image", image); return WIMLIB_ERR_INVALID_IMAGE; } if (description) { p = STRDUP(description); - if (!p) { - ERROR("Out of memory"); + if (!p) return WIMLIB_ERR_NOMEM; - } } else { p = NULL; } @@ -1535,23 +1506,20 @@ WIMLIBAPI int wimlib_set_image_descripton(WIMStruct *w, int image, return 0; } +/* Set the element of a WIM image */ WIMLIBAPI int wimlib_set_image_flags(WIMStruct *w, int image, const char *flags) { char *p; - DEBUG("Setting the flags of image %d to %s", image, flags); - if (image < 1 || image > w->hdr.image_count) { ERROR("%d is not a valid image", image); return WIMLIB_ERR_INVALID_IMAGE; } if (flags) { p = STRDUP(flags); - if (!p) { - ERROR("Out of memory"); + if (!p) return WIMLIB_ERR_NOMEM; - } } else { p = NULL; } diff --git a/src/xml.h b/src/xml.h index 5a875758..5eb24c9c 100644 --- a/src/xml.h +++ b/src/xml.h @@ -6,7 +6,7 @@ /* A struct wim_info structure corresponds to the entire XML data for a WIM file. */ struct wim_info { u64 total_bytes; - u64 num_images; + int num_images; /* Array of `struct image_info's, one for each image in the WIM that is * mentioned in the XML data. */ struct image_info *images;