diff options
Diffstat (limited to 'gzio.c')
| -rw-r--r-- | gzio.c | 370 |
1 files changed, 210 insertions, 160 deletions
@@ -1,5 +1,5 @@ /* gzio.c -- IO on .gz files - * Copyright (C) 1995 Jean-loup Gailly. + * Copyright (C) 1995-1996 Jean-loup Gailly. * For conditions of distribution and use, see copyright notice in zlib.h */ @@ -16,8 +16,7 @@ struct internal_state {int dummy;}; /* for buggy compilers */ #define ALLOC(size) malloc(size) #define TRYFREE(p) {if (p) free(p);} -#define GZ_MAGIC_1 0x1f -#define GZ_MAGIC_2 0x8b +static int gz_magic[2] = {0x1f, 0x8b}; /* gzip magic header */ /* gzip flag byte */ #define ASCII_FLAG 0x01 /* bit 0 set: file probably ascii text */ @@ -27,10 +26,6 @@ struct internal_state {int dummy;}; /* for buggy compilers */ #define COMMENT 0x10 /* bit 4 set: file comment present */ #define RESERVED 0xE0 /* bits 5..7: reserved */ -#ifndef SEEK_CUR -# define SEEK_CUR 1 -#endif - typedef struct gz_stream { z_stream stream; int z_err; /* error code for last stream operation */ @@ -46,44 +41,16 @@ typedef struct gz_stream { } gz_stream; -local int destroy OF((gz_stream *s)); -local gzFile gz_open OF((char *path, char *mode, int fd)); -local void putLong OF((FILE *file, uLong x)); -local uLong getLong OF((Bytef *buf)); - - /* =========================================================================== - * Cleanup then free the given gz_stream. Return a zlib error code. - */ -local int destroy (s) - gz_stream *s; -{ - int err = Z_OK; - - if (!s) return Z_STREAM_ERROR; - - TRYFREE(s->inbuf); - TRYFREE(s->outbuf); - TRYFREE(s->path); - TRYFREE(s->msg); - - if (s->stream.state != NULL) { - if (s->mode == 'w') { - err = deflateEnd(&(s->stream)); - } else if (s->mode == 'r') { - err = inflateEnd(&(s->stream)); - } - } - if (s->file != NULL && fclose(s->file)) { - err = Z_ERRNO; - } - if (s->z_err < 0) err = s->z_err; - TRYFREE(s); - return err; -} +local gzFile gz_open OF((const char *path, const char *mode, int fd)); +local int get_byte OF((gz_stream *s)); +local void check_header OF((gz_stream *s)); +local int destroy OF((gz_stream *s)); +local void putLong OF((FILE *file, uLong x)); +local uLong getLong OF((gz_stream *s)); /* =========================================================================== Opens a gzip (.gz) file for reading or writing. The mode parameter - is as in fopen ("rb" or "wb"). The file is given either by file descritor + is as in fopen ("rb" or "wb"). The file is given either by file descriptor or path name (if fd == -1). gz_open return NULL if the file could not be opened or if there was insufficient memory to allocate the (de)compression state; errno @@ -91,19 +58,25 @@ local int destroy (s) zlib error is Z_MEM_ERROR). */ local gzFile gz_open (path, mode, fd) - char *path; - char *mode; + const char *path; + const char *mode; int fd; { int err; int level = Z_DEFAULT_COMPRESSION; /* compression level */ - char *p = mode; - gz_stream *s = (gz_stream *)ALLOC(sizeof(gz_stream)); + char *p = (char*)mode; + gz_stream *s; + char fmode[80]; /* copy of mode, without the compression level */ + char *m = fmode; + + if (!path || !mode) return Z_NULL; + s = (gz_stream *)ALLOC(sizeof(gz_stream)); if (!s) return Z_NULL; s->stream.zalloc = (alloc_func)0; s->stream.zfree = (free_func)0; + s->stream.opaque = (voidpf)0; s->stream.next_in = s->inbuf = Z_NULL; s->stream.next_out = s->outbuf = Z_NULL; s->stream.avail_in = s->stream.avail_out = 0; @@ -123,14 +96,18 @@ local gzFile gz_open (path, mode, fd) s->mode = '\0'; do { if (*p == 'r') s->mode = 'r'; - if (*p == 'w') s->mode = 'w'; - if (*p >= '1' && *p <= '9') level = *p - '0'; - } while (*p++); + if (*p == 'w' || *p == 'a') s->mode = 'w'; + if (*p >= '0' && *p <= '9') { + level = *p - '0'; + } else { + *m++ = *p; /* copy the mode */ + } + } while (*p++ && m != fmode + sizeof(fmode)); if (s->mode == '\0') return destroy(s), (gzFile)Z_NULL; if (s->mode == 'w') { err = deflateInit2(&(s->stream), level, - DEFLATED, -MAX_WBITS, DEF_MEM_LEVEL, 0); + Z_DEFLATED, -MAX_WBITS, DEF_MEM_LEVEL, 0); /* windowBits is passed < 0 to suppress zlib header */ s->stream.next_out = s->outbuf = (Byte*)ALLOC(Z_BUFSIZE); @@ -149,7 +126,7 @@ local gzFile gz_open (path, mode, fd) s->stream.avail_out = Z_BUFSIZE; errno = 0; - s->file = fd < 0 ? FOPEN(path, mode) : fdopen(fd, mode); + s->file = fd < 0 ? FOPEN(path, fmode) : (FILE*)fdopen(fd, fmode); if (s->file == NULL) { return destroy(s), (gzFile)Z_NULL; @@ -157,50 +134,10 @@ local gzFile gz_open (path, mode, fd) if (s->mode == 'w') { /* Write a very simple .gz header: */ - fprintf(s->file, "%c%c%c%c%c%c%c%c%c%c", GZ_MAGIC_1, GZ_MAGIC_2, - DEFLATED, 0 /*flags*/, 0,0,0,0 /*time*/, 0 /*xflags*/, OS_CODE); + fprintf(s->file, "%c%c%c%c%c%c%c%c%c%c", gz_magic[0], gz_magic[1], + Z_DEFLATED, 0 /*flags*/, 0,0,0,0 /*time*/, 0 /*xflags*/, OS_CODE); } else { - /* Check and skip the header: - */ - Byte c1 = 0, c2 = 0; - Byte method = 0; - Byte flags = 0; - Byte xflags = 0; - Byte time[4]; - Byte osCode; - int c; - - s->stream.avail_in = fread(s->inbuf, 1, 2, s->file); - if (s->stream.avail_in != 2 || s->inbuf[0] != GZ_MAGIC_1 - || s->inbuf[1] != GZ_MAGIC_2) { - s->transparent = 1; - return (gzFile)s; - } - s->stream.avail_in = 0; - fscanf(s->file,"%c%c%4c%c%c", &method, &flags, time, &xflags, &osCode); - - if (method != DEFLATED || feof(s->file) || (flags & RESERVED) != 0) { - s->z_err = Z_DATA_ERROR; - return (gzFile)s; - } - if ((flags & EXTRA_FIELD) != 0) { /* skip the extra field */ - long len; - fscanf(s->file, "%c%c", &c1, &c2); - len = c1 + ((long)c2<<8); - fseek(s->file, len, SEEK_CUR); - } - if ((flags & ORIG_NAME) != 0) { /* skip the original file name */ - while ((c = getc(s->file)) != 0 && c != EOF) ; - } - if ((flags & COMMENT) != 0) { /* skip the .gz file comment */ - while ((c = getc(s->file)) != 0 && c != EOF) ; - } - if ((flags & HEAD_CRC) != 0) { /* skip the header crc */ - fscanf(s->file, "%c%c", &c1, &c2); - } - if (feof(s->file)) { - s->z_err = Z_DATA_ERROR; - } + check_header(s); /* skip the .gz header */ } return (gzFile)s; } @@ -209,26 +146,139 @@ local gzFile gz_open (path, mode, fd) Opens a gzip (.gz) file for reading or writing. */ gzFile gzopen (path, mode) - char *path; - char *mode; + const char *path; + const char *mode; { return gz_open (path, mode, -1); } /* =========================================================================== - Associate a gzFile with the file descriptor fd. + Associate a gzFile with the file descriptor fd. fd is not dup'ed here + to mimic the behavio(u)r of fdopen. */ gzFile gzdopen (fd, mode) int fd; - char *mode; + const char *mode; { char name[20]; + + if (fd < 0) return (gzFile)Z_NULL; sprintf(name, "<fd:%d>", fd); /* for debugging */ return gz_open (name, mode, fd); } /* =========================================================================== + Read a byte from a gz_stream; update next_in and avail_in. Return EOF + for end of file. + IN assertion: the stream s has been sucessfully opened for reading. +*/ +local int get_byte(s) + gz_stream *s; +{ + if (s->z_eof) return EOF; + if (s->stream.avail_in == 0) { + errno = 0; + s->stream.avail_in = fread(s->inbuf, 1, Z_BUFSIZE, s->file); + if (s->stream.avail_in == 0) { + s->z_eof = 1; + if (ferror(s->file)) s->z_err = Z_ERRNO; + return EOF; + } + s->stream.next_in = s->inbuf; + } + s->stream.avail_in--; + return *(s->stream.next_in)++; +} + +/* =========================================================================== + Check the gzip header of a gz_stream opened for reading. Set the stream + mode to transparent if the gzip magic header is not present; set s->err + to Z_DATA_ERROR if the magic header is present but the rest of the header + is incorrect. + IN assertion: the stream s has already been created sucessfully; + s->stream.avail_in is zero for the first time, but may be non-zero + for concatenated .gz files. +*/ +local void check_header(s) + gz_stream *s; +{ + int method = 0; + int flags = 0; + uInt len; + int c; + + /* Check the gzip magic header */ + for (len = 0; len < 2; len++) { + c = get_byte(s); + if (c != gz_magic[len]) { + s->transparent = 1; + if (c != EOF) s->stream.avail_in++, s->stream.next_in--; + s->z_err = s->stream.avail_in != 0 ? Z_OK : Z_STREAM_END; + return; + } + } + method = get_byte(s); + flags = get_byte(s); + if (method != Z_DEFLATED || (flags & RESERVED) != 0) { + s->z_err = Z_DATA_ERROR; + return; + } + + /* Discard time, xflags and OS code: */ + for (len = 0; len < 6; len++) (void)get_byte(s); + + if ((flags & EXTRA_FIELD) != 0) { /* skip the extra field */ + len = (uInt)get_byte(s); + len += ((uInt)get_byte(s))<<8; + /* len is garbage if EOF but the loop below will quit anyway */ + while (len-- != 0 && get_byte(s) != EOF) ; + } + if ((flags & ORIG_NAME) != 0) { /* skip the original file name */ + while ((c = get_byte(s)) != 0 && c != EOF) ; + } + if ((flags & COMMENT) != 0) { /* skip the .gz file comment */ + while ((c = get_byte(s)) != 0 && c != EOF) ; + } + if ((flags & HEAD_CRC) != 0) { /* skip the header crc */ + for (len = 0; len < 2; len++) (void)get_byte(s); + } + s->z_err = s->z_eof ? Z_DATA_ERROR : Z_OK; +} + + /* =========================================================================== + * Cleanup then free the given gz_stream. Return a zlib error code. + Try freeing in the reverse order of allocations. + */ +local int destroy (s) + gz_stream *s; +{ + int err = Z_OK; + + if (!s) return Z_STREAM_ERROR; + + TRYFREE(s->msg); + + if (s->stream.state != NULL) { + if (s->mode == 'w') { + err = deflateEnd(&(s->stream)); + } else if (s->mode == 'r') { + err = inflateEnd(&(s->stream)); + } + } + if (s->file != NULL && fclose(s->file)) { + err = Z_ERRNO; + } + if (s->z_err < 0) err = s->z_err; + + TRYFREE(s->inbuf); + TRYFREE(s->outbuf); + TRYFREE(s->path); + TRYFREE(s); + return err; +} + +/* =========================================================================== Reads the given number of uncompressed bytes from the compressed file. gzread returns the number of bytes actually read (0 for end of file). */ @@ -238,52 +288,71 @@ int gzread (file, buf, len) unsigned len; { gz_stream *s = (gz_stream*)file; + Byte *start = buf; /* starting point for crc computation */ if (s == NULL || s->mode != 'r') return Z_STREAM_ERROR; - if (s->transparent) { - int n = 0; - Byte *b = (Byte*)buf; - /* Copy the first two (non-magic) bytes if not done already */ - while (s->stream.avail_in > 0 && len > 0) { - *b++ = *s->stream.next_in++; - s->stream.avail_in--; - len--; n++; - } - if (len == 0) return n; - return n + fread(b, 1, len, s->file); - } - if (s->z_err == Z_DATA_ERROR) return -1; /* bad .gz file */ - if (s->z_err == Z_STREAM_END) return 0; /* don't read crc as data */ + if (s->z_err == Z_DATA_ERROR || s->z_err == Z_ERRNO) return -1; + if (s->z_err == Z_STREAM_END) return 0; /* EOF */ s->stream.next_out = buf; s->stream.avail_out = len; while (s->stream.avail_out != 0) { + if (s->transparent) { + /* Copy first the lookahead bytes: */ + uInt n = s->stream.avail_in; + if (n > s->stream.avail_out) n = s->stream.avail_out; + if (n > 0) { + zmemcpy(s->stream.next_out, s->stream.next_in, n); + s->stream.next_out += n; + s->stream.next_in += n; + s->stream.avail_out -= n; + s->stream.avail_in -= n; + } + if (s->stream.avail_out > 0) { + s->stream.avail_out -= fread(s->stream.next_out, + 1, s->stream.avail_out, s->file); + } + return (int)(len - s->stream.avail_out); + } if (s->stream.avail_in == 0 && !s->z_eof) { errno = 0; - s->stream.avail_in = - fread(s->inbuf, 1, Z_BUFSIZE, s->file); + s->stream.avail_in = fread(s->inbuf, 1, Z_BUFSIZE, s->file); if (s->stream.avail_in == 0) { s->z_eof = 1; - } else if (s->stream.avail_in == (uInt)EOF) { - s->stream.avail_in = 0; - s->z_eof = 1; - s->z_err = Z_ERRNO; - break; + if (ferror(s->file)) { + s->z_err = Z_ERRNO; + break; + } } s->stream.next_in = s->inbuf; } s->z_err = inflate(&(s->stream), Z_NO_FLUSH); - if (s->z_err == Z_STREAM_END || - s->z_err != Z_OK || s->z_eof) break; + if (s->z_err == Z_STREAM_END) { + /* Check CRC and original size */ + s->crc = crc32(s->crc, start, (uInt)(s->stream.next_out - start)); + start = s->stream.next_out; + + if (getLong(s) != s->crc || getLong(s) != s->stream.total_out) { + s->z_err = Z_DATA_ERROR; + } else { + /* Check for concatenated .gz files: */ + check_header(s); + if (s->z_err == Z_OK) { + inflateReset(&(s->stream)); + s->crc = crc32(0L, Z_NULL, 0); + } + } + } + if (s->z_err != Z_OK || s->z_eof) break; } - len -= s->stream.avail_out; - s->crc = crc32(s->crc, buf, len); - return (int)len; + s->crc = crc32(s->crc, start, (uInt)(s->stream.next_out - start)); + + return (int)(len - s->stream.avail_out); } /* =========================================================================== @@ -292,7 +361,7 @@ int gzread (file, buf, len) */ int gzwrite (file, buf, len) gzFile file; - voidp buf; + const voidp buf; unsigned len; { gz_stream *s = (gz_stream*)file; @@ -343,7 +412,7 @@ int gzflush (file, flush) len = Z_BUFSIZE - s->stream.avail_out; if (len != 0) { - if (fwrite(s->outbuf, 1, len, s->file) != len) { + if ((uInt)fwrite(s->outbuf, 1, len, s->file) != len) { s->z_err = Z_ERRNO; return Z_ERRNO; } @@ -379,18 +448,19 @@ local void putLong (file, x) } /* =========================================================================== - Reads a long in LSB order from the given buffer + Reads a long in LSB order from the given gz_stream. Sets */ -local uLong getLong (buf) - Bytef *buf; +local uLong getLong (s) + gz_stream *s; { - uLong x = 0; - Bytef *p = buf+4; - - do { - x <<= 8; - x |= *--p; - } while (p != buf); + uLong x = (uLong)get_byte(s); + int c; + + x += ((uLong)get_byte(s))<<8; + x += ((uLong)get_byte(s))<<16; + c = get_byte(s); + if (c == EOF) s->z_err = Z_DATA_ERROR; + x += ((uLong)c)<<24; return x; } @@ -401,7 +471,6 @@ local uLong getLong (buf) int gzclose (file) gzFile file; { - uInt n; int err; gz_stream *s = (gz_stream*)file; @@ -414,25 +483,6 @@ int gzclose (file) putLong (s->file, s->crc); putLong (s->file, s->stream.total_in); - } else if (s->mode == 'r' && s->z_err == Z_STREAM_END) { - - /* slide CRC and original size if they are at the end of inbuf */ - if ((n = s->stream.avail_in) < 8 && !s->z_eof) { - Byte *p = s->inbuf; - Bytef *q = s->stream.next_in; - while (n--) { *p++ = *q++; }; - - n = s->stream.avail_in; - n += fread(p, 1, 8, s->file); - s->stream.next_in = s->inbuf; - } - /* check CRC and original size */ - if (n < 8 || - getLong(s->stream.next_in) != s->crc || - getLong(s->stream.next_in + 4) != s->stream.total_out) { - - s->z_err = Z_DATA_ERROR; - } } return destroy(file); } @@ -453,14 +503,14 @@ char* gzerror (file, errnum) if (s == NULL) { *errnum = Z_STREAM_ERROR; - return z_errmsg[1-Z_STREAM_ERROR]; + return ERR_MSG(Z_STREAM_ERROR); } *errnum = s->z_err; - if (*errnum == Z_OK) return ""; + if (*errnum == Z_OK) return (char*)""; - m = *errnum == Z_ERRNO ? zstrerror(errno) : s->stream.msg; + m = (char*)(*errnum == Z_ERRNO ? zstrerror(errno) : s->stream.msg); - if (m == NULL || *m == '\0') m = z_errmsg[1-s->z_err]; + if (m == NULL || *m == '\0') m = (char*)z_errmsg[1-s->z_err]; TRYFREE(s->msg); s->msg = (char*)ALLOC(strlen(s->path) + strlen(m) + 3); |