diff options
| author | Eric Andersen | 1999-10-05 16:24:54 +0000 |
|---|---|---|
| committer | Eric Andersen | 1999-10-05 16:24:54 +0000 |
| commit | cc8ed39b240180b58810784f844e253263594ac3 (patch) | |
| tree | 15feebbb4be9a9168209609f48f0b100f9364420 /archival | |
| download | busybox-cc8ed39b240180b58810784f844e253263594ac3.zip busybox-cc8ed39b240180b58810784f844e253263594ac3.tar.gz | |
Initial revision0_29alpha2
Diffstat (limited to 'archival')
| -rw-r--r-- | archival/gzip.c | 3231 | ||||
| -rw-r--r-- | archival/tar.c | 1425 |
2 files changed, 4656 insertions, 0 deletions
diff --git a/archival/gzip.c b/archival/gzip.c new file mode 100644 index 0000000..6fd2e39 --- /dev/null +++ b/archival/gzip.c @@ -0,0 +1,3231 @@ +/* gzip.c -- this is a stripped down version of gzip I put into busybox, it does + * only standard in to standard out with -9 compression. It also requires the + * zcat module for some important functions. + * + * Charles P. Wright <cpw@unix.asb.com> + */ +#include "internal.h" +#ifdef BB_GZIP + +#ifndef BB_ZCAT +error: you need zcat to have gzip support! +#endif + +const char gzip_usage[] = "gzip\nignores all command line arguments\ncompress stdin to stdout with -9 compression\n"; + +/* gzip.h -- common declarations for all gzip modules + * Copyright (C) 1992-1993 Jean-loup Gailly. + * This is free software; you can redistribute it and/or modify it under the + * terms of the GNU General Public License, see the file COPYING. + */ + +#if defined(__STDC__) || defined(PROTO) +# define OF(args) args +#else +# define OF(args) () +#endif + +#ifdef __STDC__ + typedef void *voidp; +#else + typedef char *voidp; +#endif + +/* I don't like nested includes, but the string and io functions are used + * too often + */ +#include <stdio.h> +#if !defined(NO_STRING_H) || defined(STDC_HEADERS) +# include <string.h> +# if !defined(STDC_HEADERS) && !defined(NO_MEMORY_H) && !defined(__GNUC__) +# include <memory.h> +# endif +# define memzero(s, n) memset ((voidp)(s), 0, (n)) +#else +# include <strings.h> +# define strchr index +# define strrchr rindex +# define memcpy(d, s, n) bcopy((s), (d), (n)) +# define memcmp(s1, s2, n) bcmp((s1), (s2), (n)) +# define memzero(s, n) bzero((s), (n)) +#endif + +#ifndef RETSIGTYPE +# define RETSIGTYPE void +#endif + +#define local static + +typedef unsigned char uch; +typedef unsigned short ush; +typedef unsigned long ulg; + +/* Return codes from gzip */ +#define OK 0 +#define ERROR 1 +#define WARNING 2 + +/* Compression methods (see algorithm.doc) */ +#define STORED 0 +#define COMPRESSED 1 +#define PACKED 2 +#define LZHED 3 +/* methods 4 to 7 reserved */ +#define DEFLATED 8 +#define MAX_METHODS 9 +extern int method; /* compression method */ + +/* To save memory for 16 bit systems, some arrays are overlaid between + * the various modules: + * deflate: prev+head window d_buf l_buf outbuf + * unlzw: tab_prefix tab_suffix stack inbuf outbuf + * inflate: window inbuf + * unpack: window inbuf prefix_len + * unlzh: left+right window c_table inbuf c_len + * For compression, input is done in window[]. For decompression, output + * is done in window except for unlzw. + */ + +#ifndef INBUFSIZ +# ifdef SMALL_MEM +# define INBUFSIZ 0x2000 /* input buffer size */ +# else +# define INBUFSIZ 0x8000 /* input buffer size */ +# endif +#endif +#define INBUF_EXTRA 64 /* required by unlzw() */ + +#ifndef OUTBUFSIZ +# ifdef SMALL_MEM +# define OUTBUFSIZ 8192 /* output buffer size */ +# else +# define OUTBUFSIZ 16384 /* output buffer size */ +# endif +#endif +#define OUTBUF_EXTRA 2048 /* required by unlzw() */ + +#ifndef DIST_BUFSIZE +# ifdef SMALL_MEM +# define DIST_BUFSIZE 0x2000 /* buffer for distances, see trees.c */ +# else +# define DIST_BUFSIZE 0x8000 /* buffer for distances, see trees.c */ +# endif +#endif + +#ifdef DYN_ALLOC +# define EXTERN(type, array) extern type * near array +# define DECLARE(type, array, size) type * near array +# define ALLOC(type, array, size) { \ + array = (type*)fcalloc((size_t)(((size)+1L)/2), 2*sizeof(type)); \ + if (array == NULL) error("insufficient memory"); \ + } +# define FREE(array) {if (array != NULL) fcfree(array), array=NULL;} +#else +# define EXTERN(type, array) extern type array[] +# define DECLARE(type, array, size) type array[size] +# define ALLOC(type, array, size) +# define FREE(array) +#endif + +EXTERN(uch, inbuf); /* input buffer */ +EXTERN(uch, outbuf); /* output buffer */ +EXTERN(ush, d_buf); /* buffer for distances, see trees.c */ +EXTERN(uch, window); /* Sliding window and suffix table (unlzw) */ +#define tab_suffix window +#ifndef MAXSEG_64K +# define tab_prefix prev /* hash link (see deflate.c) */ +# define head (prev+WSIZE) /* hash head (see deflate.c) */ + EXTERN(ush, tab_prefix); /* prefix code (see unlzw.c) */ +#else +# define tab_prefix0 prev +# define head tab_prefix1 + EXTERN(ush, tab_prefix0); /* prefix for even codes */ + EXTERN(ush, tab_prefix1); /* prefix for odd codes */ +#endif + +extern unsigned insize; /* valid bytes in inbuf */ +extern unsigned inptr; /* index of next byte to be processed in inbuf */ +extern unsigned outcnt; /* bytes in output buffer */ + +extern long bytes_in; /* number of input bytes */ +extern long bytes_out; /* number of output bytes */ +extern long header_bytes;/* number of bytes in gzip header */ + +#define isize bytes_in +/* for compatibility with old zip sources (to be cleaned) */ + +extern int ifd; /* input file descriptor */ +extern int ofd; /* output file descriptor */ +extern char ifname[]; /* input file name or "stdin" */ +extern char ofname[]; /* output file name or "stdout" */ +extern char *progname; /* program name */ + +extern long time_stamp; /* original time stamp (modification time) */ +extern long ifile_size; /* input file size, -1 for devices (debug only) */ + +typedef int file_t; /* Do not use stdio */ +#define NO_FILE (-1) /* in memory compression */ + + +#define PACK_MAGIC "\037\036" /* Magic header for packed files */ +#define GZIP_MAGIC "\037\213" /* Magic header for gzip files, 1F 8B */ +#define OLD_GZIP_MAGIC "\037\236" /* Magic header for gzip 0.5 = freeze 1.x */ +#define LZH_MAGIC "\037\240" /* Magic header for SCO LZH Compress files*/ +#define PKZIP_MAGIC "\120\113\003\004" /* Magic header for pkzip files */ + +/* gzip flag byte */ +#define ASCII_FLAG 0x01 /* bit 0 set: file probably ascii text */ +#define CONTINUATION 0x02 /* bit 1 set: continuation of multi-part gzip file */ +#define EXTRA_FIELD 0x04 /* bit 2 set: extra field present */ +#define ORIG_NAME 0x08 /* bit 3 set: original file name present */ +#define COMMENT 0x10 /* bit 4 set: file comment present */ +#define ENCRYPTED 0x20 /* bit 5 set: file is encrypted */ +#define RESERVED 0xC0 /* bit 6,7: reserved */ + +/* internal file attribute */ +#define UNKNOWN 0xffff +#define BINARY 0 +#define ASCII 1 + +#ifndef WSIZE +# define WSIZE 0x8000 /* window size--must be a power of two, and */ +#endif /* at least 32K for zip's deflate method */ + +#define MIN_MATCH 3 +#define MAX_MATCH 258 +/* The minimum and maximum match lengths */ + +#define MIN_LOOKAHEAD (MAX_MATCH+MIN_MATCH+1) +/* Minimum amount of lookahead, except at the end of the input file. + * See deflate.c for comments about the MIN_MATCH+1. + */ + +#define MAX_DIST (WSIZE-MIN_LOOKAHEAD) +/* In order to simplify the code, particularly on 16 bit machines, match + * distances are limited to MAX_DIST instead of WSIZE. + */ + +extern int decrypt; /* flag to turn on decryption */ +extern int exit_code; /* program exit code */ +extern int verbose; /* be verbose (-v) */ +extern int quiet; /* be quiet (-q) */ +extern int test; /* check .z file integrity */ +extern int to_stdout; /* output to stdout (-c) */ +extern int save_orig_name; /* set if original name must be saved */ + +#define get_byte() (inptr < insize ? inbuf[inptr++] : fill_inbuf(0)) +#define try_byte() (inptr < insize ? inbuf[inptr++] : fill_inbuf(1)) + +/* put_byte is used for the compressed output, put_ubyte for the + * uncompressed output. However unlzw() uses window for its + * suffix table instead of its output buffer, so it does not use put_ubyte + * (to be cleaned up). + */ +#define put_byte(c) {outbuf[outcnt++]=(uch)(c); if (outcnt==OUTBUFSIZ)\ + flush_outbuf();} +#define put_ubyte(c) {window[outcnt++]=(uch)(c); if (outcnt==WSIZE)\ + flush_window();} + +/* Output a 16 bit value, lsb first */ +#define put_short(w) \ +{ if (outcnt < OUTBUFSIZ-2) { \ + outbuf[outcnt++] = (uch) ((w) & 0xff); \ + outbuf[outcnt++] = (uch) ((ush)(w) >> 8); \ + } else { \ + put_byte((uch)((w) & 0xff)); \ + put_byte((uch)((ush)(w) >> 8)); \ + } \ +} + +/* Output a 32 bit value to the bit stream, lsb first */ +#define put_long(n) { \ + put_short((n) & 0xffff); \ + put_short(((ulg)(n)) >> 16); \ +} + +#define seekable() 0 /* force sequential output */ +#define translate_eol 0 /* no option -a yet */ + +#define tolow(c) (isupper(c) ? (c)-'A'+'a' : (c)) /* force to lower case */ + +/* Macros for getting two-byte and four-byte header values */ +#define SH(p) ((ush)(uch)((p)[0]) | ((ush)(uch)((p)[1]) << 8)) +#define LG(p) ((ulg)(SH(p)) | ((ulg)(SH((p)+2)) << 16)) + +/* Diagnostic functions */ +#ifdef DEBUG +# define Assert(cond,msg) {if(!(cond)) error(msg);} +# define Trace(x) fprintf x +# define Tracev(x) {if (verbose) fprintf x ;} +# define Tracevv(x) {if (verbose>1) fprintf x ;} +# define Tracec(c,x) {if (verbose && (c)) fprintf x ;} +# define Tracecv(c,x) {if (verbose>1 && (c)) fprintf x ;} +#else +# define Assert(cond,msg) +# define Trace(x) +# define Tracev(x) +# define Tracevv(x) +# define Tracec(c,x) +# define Tracecv(c,x) +#endif + +#define WARN(msg) {if (!quiet) fprintf msg ; \ + if (exit_code == OK) exit_code = WARNING;} + +local void do_exit(int exitcode); + + /* in zip.c: */ +extern int zip OF((int in, int out)); +extern int file_read OF((char *buf, unsigned size)); + + /* in unzip.c */ +extern int unzip OF((int in, int out)); +extern int check_zipfile OF((int in)); + + /* in unpack.c */ +extern int unpack OF((int in, int out)); + + /* in unlzh.c */ +extern int unlzh OF((int in, int out)); + + /* in gzip.c */ +RETSIGTYPE abort_gzip OF((void)); + + /* in deflate.c */ +void lm_init OF((ush *flags)); +ulg deflate OF((void)); + + /* in trees.c */ +void ct_init OF((ush *attr, int *method)); +int ct_tally OF((int dist, int lc)); +ulg flush_block OF((char *buf, ulg stored_len, int eof)); + + /* in bits.c */ +void bi_init OF((file_t zipfile)); +void send_bits OF((int value, int length)); +unsigned bi_reverse OF((unsigned value, int length)); +void bi_windup OF((void)); +void copy_block OF((char *buf, unsigned len, int header)); +extern int (*read_buf) OF((char *buf, unsigned size)); + + /* in util.c: */ +extern int copy OF((int in, int out)); +extern ulg updcrc OF((uch *s, unsigned n)); +extern void clear_bufs OF((void)); +extern int fill_inbuf OF((int eof_ok)); +extern void flush_outbuf OF((void)); +extern void flush_window OF((void)); +extern void write_buf OF((int fd, voidp buf, unsigned cnt)); +extern char *strlwr OF((char *s)); +extern char *add_envopt OF((int *argcp, char ***argvp, char *env)); +extern void error OF((char *m)); +extern void warn OF((char *a, char *b)); +extern void read_error OF((void)); +extern void write_error OF((void)); +extern void display_ratio OF((long num, long den, FILE *file)); +extern voidp xmalloc OF((unsigned int size)); + + /* in inflate.c */ +extern int inflate OF((void)); +/* lzw.h -- define the lzw functions. + * Copyright (C) 1992-1993 Jean-loup Gailly. + * This is free software; you can redistribute it and/or modify it under the + * terms of the GNU General Public License, see the file COPYING. + */ + +#if !defined(OF) && defined(lint) +# include "gzip.h" +#endif + +#ifndef BITS +# define BITS 16 +#endif +#define INIT_BITS 9 /* Initial number of bits per code */ + +#define BIT_MASK 0x1f /* Mask for 'number of compression bits' */ +/* Mask 0x20 is reserved to mean a fourth header byte, and 0x40 is free. + * It's a pity that old uncompress does not check bit 0x20. That makes + * extension of the format actually undesirable because old compress + * would just crash on the new format instead of giving a meaningful + * error message. It does check the number of bits, but it's more + * helpful to say "unsupported format, get a new version" than + * "can only handle 16 bits". + */ + +#define BLOCK_MODE 0x80 +/* Block compression: if table is full and compression rate is dropping, + * clear the dictionary. + */ + +#define LZW_RESERVED 0x60 /* reserved bits */ + +#define CLEAR 256 /* flush the dictionary */ +#define FIRST (CLEAR+1) /* first free entry */ + +extern int maxbits; /* max bits per code for LZW */ +extern int block_mode; /* block compress mode -C compatible with 2.0 */ + +/* revision.h -- define the version number + * Copyright (C) 1992-1993 Jean-loup Gailly. + * This is free software; you can redistribute it and/or modify it under the + * terms of the GNU General Public License, see the file COPYING. + */ + +#define VERSION "1.2.4" +#define PATCHLEVEL 0 +#define REVDATE "18 Aug 93" + +/* This version does not support compression into old compress format: */ +#ifdef LZW +# undef LZW +#endif + +/* $Id: gzip.c,v 1.1 1999/10/05 16:24:56 andersen Exp $ */ +/* tailor.h -- target dependent definitions + * Copyright (C) 1992-1993 Jean-loup Gailly. + * This is free software; you can redistribute it and/or modify it under the + * terms of the GNU General Public License, see the file COPYING. + */ + +/* The target dependent definitions should be defined here only. + * The target dependent functions should be defined in tailor.c. + */ + +/* $Id: gzip.c,v 1.1 1999/10/05 16:24:56 andersen Exp $ */ + +#if defined(__MSDOS__) && !defined(MSDOS) +# define MSDOS +#endif + +#if defined(__OS2__) && !defined(OS2) +# define OS2 +#endif + +#if defined(OS2) && defined(MSDOS) /* MS C under OS/2 */ +# undef MSDOS +#endif + +#ifdef MSDOS +# ifdef __GNUC__ + /* DJGPP version 1.09+ on MS-DOS. + * The DJGPP 1.09 stat() function must be upgraded before gzip will + * fully work. + * No need for DIRENT, since <unistd.h> defines POSIX_SOURCE which + * implies DIRENT. + */ +# define near +# else +# define MAXSEG_64K +# ifdef __TURBOC__ +# define NO_OFF_T +# ifdef __BORLANDC__ +# define DIRENT +# else +# define NO_UTIME +# endif +# else /* MSC */ +# define HAVE_SYS_UTIME_H +# define NO_UTIME_H +# endif +# endif +# define PATH_SEP2 '\\' +# define PATH_SEP3 ':' +# define MAX_PATH_LEN 128 +# define NO_MULTIPLE_DOTS +# define MAX_EXT_CHARS 3 +# define Z_SUFFIX "z" +# define NO_CHOWN +# define PROTO +# define STDC_HEADERS +# define NO_SIZE_CHECK +# define casemap(c) tolow(c) /* Force file names to lower case */ +# include <io.h> +# define OS_CODE 0x00 +# define SET_BINARY_MODE(fd) setmode(fd, O_BINARY) +# if !defined(NO_ASM) && !defined(ASMV) +# define ASMV +# endif +#else +# define near +#endif + +#ifdef OS2 +# define PATH_SEP2 '\\' +# define PATH_SEP3 ':' +# define MAX_PATH_LEN 260 +# ifdef OS2FAT +# define NO_MULTIPLE_DOTS +# define MAX_EXT_CHARS 3 +# define Z_SUFFIX "z" +# define casemap(c) tolow(c) +# endif +# define NO_CHOWN +# define PROTO +# define STDC_HEADERS +# include <io.h> +# define OS_CODE 0x06 +# define SET_BINARY_MODE(fd) setmode(fd, O_BINARY) +# ifdef _MSC_VER +# define HAVE_SYS_UTIME_H +# define NO_UTIME_H +# define MAXSEG_64K +# undef near +# define near _near +# endif +# ifdef __EMX__ +# define HAVE_SYS_UTIME_H +# define NO_UTIME_H +# define DIRENT +# define EXPAND(argc,argv) \ + {_response(&argc, &argv); _wildcard(&argc, &argv);} +# endif +# ifdef __BORLANDC__ +# define DIRENT +# endif +# ifdef __ZTC__ +# define NO_DIR +# define NO_UTIME_H +# include <dos.h> +# define EXPAND(argc,argv) \ + {response_expand(&argc, &argv);} +# endif +#endif + +#ifdef WIN32 /* Windows NT */ +# define HAVE_SYS_UTIME_H +# define NO_UTIME_H +# define PATH_SEP2 '\\' +# define PATH_SEP3 ':' +# define MAX_PATH_LEN 260 +# define NO_CHOWN +# define PROTO +# define STDC_HEADERS +# define SET_BINARY_MODE(fd) setmode(fd, O_BINARY) +# include <io.h> +# include <malloc.h> +# ifdef NTFAT +# define NO_MULTIPLE_DOTS +# define MAX_EXT_CHARS 3 +# define Z_SUFFIX "z" +# define casemap(c) tolow(c) /* Force file names to lower case */ +# endif +# define OS_CODE 0x0b +#endif + +#ifdef MSDOS +# ifdef __TURBOC__ +# include <alloc.h> +# define DYN_ALLOC + /* Turbo C 2.0 does not accept static allocations of large arrays */ + void * fcalloc (unsigned items, unsigned size); + void fcfree (void *ptr); +# else /* MSC */ +# include <malloc.h> +# define fcalloc(nitems,itemsize) halloc((long)(nitems),(itemsize)) +# define fcfree(ptr) hfree(ptr) +# endif +#else +# ifdef MAXSEG_64K +# define fcalloc(items,size) calloc((items),(size)) +# else +# define fcalloc(items,size) malloc((size_t)(items)*(size_t)(size)) +# endif +# define fcfree(ptr) free(ptr) +#endif + +#if defined(VAXC) || defined(VMS) +# define PATH_SEP ']' +# define PATH_SEP2 ':' +# define SUFFIX_SEP ';' +# define NO_MULTIPLE_DOTS +# define Z_SUFFIX "-gz" +# define RECORD_IO 1 +# define casemap(c) tolow(c) +# define OS_CODE 0x02 +# define OPTIONS_VAR "GZIP_OPT" +# define STDC_HEADERS +# define NO_UTIME +# define EXPAND(argc,argv) vms_expand_args(&argc,&argv); +# include <file.h> +# define unlink delete +# ifdef VAXC +# define NO_FCNTL_H +# include <unixio.h> +# endif +#endif + +#ifdef AMIGA +# define PATH_SEP2 ':' +# define STDC_HEADERS +# define OS_CODE 0x01 +# define ASMV +# ifdef __GNUC__ +# define DIRENT +# define HAVE_UNISTD_H +# else /* SASC */ +# define NO_STDIN_FSTAT +# define SYSDIR +# define NO_SYMLINK +# define NO_CHOWN +# define NO_FCNTL_H +# include <fcntl.h> /* for read() and write() */ +# define direct dirent + extern void _expand_args(int *argc, char ***argv); +# define EXPAND(argc,argv) _expand_args(&argc,&argv); +# undef O_BINARY /* disable useless --ascii option */ +# endif +#endif + +#if defined(ATARI) || defined(atarist) +# ifndef STDC_HEADERS +# define STDC_HEADERS +# define HAVE_UNISTD_H +# define DIRENT +# endif +# define ASMV +# define OS_CODE 0x05 +# ifdef TOSFS +# define PATH_SEP2 '\\' +# define PATH_SEP3 ':' +# define MAX_PATH_LEN 128 +# define NO_MULTIPLE_DOTS +# define MAX_EXT_CHARS 3 +# define Z_SUFFIX "z" +# define NO_CHOWN +# define casemap(c) tolow(c) /* Force file names to lower case */ +# define NO_SYMLINK +# endif +#endif + +#ifdef MACOS +# define PATH_SEP ':' +# define DYN_ALLOC +# define PROTO +# define NO_STDIN_FSTAT +# define NO_CHOWN +# define NO_UTIME +# define chmod(file, mode) (0) +# define OPEN(name, flags, mode) open(name, flags) +# define OS_CODE 0x07 +# ifdef MPW +# define isatty(fd) ((fd) <= 2) +# endif +#endif + +#ifdef __50SERIES /* Prime/PRIMOS */ +# define PATH_SEP '>' +# define STDC_HEADERS +# define NO_MEMORY_H +# define NO_UTIME_H +# define NO_UTIME +# define NO_CHOWN +# define NO_STDIN_FSTAT +# define NO_SIZE_CHECK +# define NO_SYMLINK +# define RECORD_IO 1 +# define casemap(c) tolow(c) /* Force file names to lower case */ +# define put_char(c) put_byte((c) & 0x7F) +# define get_char(c) ascii2pascii(get_byte()) +# define OS_CODE 0x0F /* temporary, subject to change */ +# ifdef SIGTERM +# undef SIGTERM /* We don't want a signal handler for SIGTERM */ +# endif +#endif + +#if defined(pyr) && !defined(NOMEMCPY) /* Pyramid */ +# define NOMEMCPY /* problem with overlapping copies */ +#endif + +#ifdef TOPS20 +# define OS_CODE 0x0a +#endif + +#ifndef unix +# define NO_ST_INO /* don't rely on inode numbers */ +#endif + + + /* Common defaults */ + +#ifndef OS_CODE +# define OS_CODE 0x03 /* assume Unix */ +#endif + +#ifndef PATH_SEP +# define PATH_SEP '/' +#endif + +#ifndef casemap +# define casemap(c) (c) +#endif + +#ifndef OPTIONS_VAR +# define OPTIONS_VAR "GZIP" +#endif + +#ifndef Z_SUFFIX +# define Z_SUFFIX ".gz" +#endif + +#ifdef MAX_EXT_CHARS +# define MAX_SUFFIX MAX_EXT_CHARS +#else +# define MAX_SUFFIX 30 +#endif + +#ifndef MAKE_LEGAL_NAME +# ifdef NO_MULTIPLE_DOTS +# define MAKE_LEGAL_NAME(name) make_simple_name(name) +# else +# define MAKE_LEGAL_NAME(name) +# endif +#endif + +#ifndef MIN_PART +# define MIN_PART 3 + /* keep at least MIN_PART chars between dots in a file name. */ +#endif + +#ifndef EXPAND +# define EXPAND(argc,argv) +#endif + +#ifndef RECORD_IO +# define RECORD_IO 0 +#endif + +#ifndef SET_BINARY_MODE +# define SET_BINARY_MODE(fd) +#endif + +#ifndef OPEN +# define OPEN(name, flags, mode) open(name, flags, mode) +#endif + +#ifndef get_char +# define get_char() get_byte() +#endif + +#ifndef put_char +# define put_char(c) put_byte(c) +#endif +/* bits.c -- output variable-length bit strings + * Copyright (C) 1992-1993 Jean-loup Gailly + * This is free software; you can redistribute it and/or modify it under the + * terms of the GNU General Public License, see the file COPYING. + */ + + +/* + * PURPOSE + * + * Output variable-length bit strings. Compression can be done + * to a file or to memory. (The latter is not supported in this version.) + * + * DISCUSSION + * + * The PKZIP "deflate" file format interprets compressed file data + * as a sequence of bits. Multi-bit strings in the file may cross + * byte boundaries without restriction. + * + * The first bit of each byte is the low-order bit. + * + * The routines in this file allow a variable-length bit value to + * be output right-to-left (useful for literal values). For + * left-to-right output (useful for code strings from the tree routines), + * the bits must have been reversed first with bi_reverse(). + * + * For in-memory compression, the compressed bit stream goes directly + * into the requested output buffer. The input data is read in blocks + * by the mem_read() function. The buffer is limited to 64K on 16 bit + * machines. + * + * INTERFACE + * + * void bi_init (FILE *zipfile) + * Initialize the bit string routines. + * + * void send_bits (int value, int length) + * Write out a bit string, taking the source bits right to + * left. + * + * int bi_reverse (int value, int length) + * Reverse the bits of a bit string, taking the source bits left to + * right and emitting them right to left. + * + * void bi_windup (void) + * Write out any remaining bits in an incomplete byte. + * + * void copy_block(char *buf, unsigned len, int header) + * Copy a stored block to the zip file, storing first the length and + * its one's complement if requested. + * + */ + +#ifdef DEBUG +# include <stdio.h> +#endif + +#ifdef RCSID +static char rcsid[] = "$Id: gzip.c,v 1.1 1999/10/05 16:24:56 andersen Exp $"; +#endif + +/* =========================================================================== + * Local data used by the "bit string" routines. + */ + +local file_t zfile; /* output gzip file */ + +local unsigned short bi_buf; +/* Output buffer. bits are inserted starting at the bottom (least significant + * bits). + */ + +#define Buf_size (8 * 2*sizeof(char)) +/* Number of bits used within bi_buf. (bi_buf might be implemented on + * more than 16 bits on some systems.) + */ + +local int bi_valid; +/* Number of valid bits in bi_buf. All bits above the last valid bit + * are always zero. + */ + +int (*read_buf) OF((char *buf, unsigned size)); +/* Current input function. Set to mem_read for in-memory compression */ + +#ifdef DEBUG + ulg bits_sent; /* bit length of the compressed data */ +#endif + +/* =========================================================================== + * Initialize the bit string routines. + */ +void bi_init (zipfile) + file_t zipfile; /* output zip file, NO_FILE for in-memory compression */ +{ + zfile = zipfile; + bi_buf = 0; + bi_valid = 0; +#ifdef DEBUG + bits_sent = 0L; +#endif + + /* Set the defaults for file compression. They are set by memcompress + * for in-memory compression. + */ + if (zfile != NO_FILE) { + read_buf = file_read; + } +} + +/* =========================================================================== + * Send a value on a given number of bits. + * IN assertion: length <= 16 and value fits in length bits. + */ +void send_bits(value, length) + int value; /* value to send */ + int length; /* number of bits */ +{ +#ifdef DEBUG + Tracev((stderr," l %2d v %4x ", length, value)); + Assert(length > 0 && length <= 15, "invalid length"); + bits_sent += (ulg)length; +#endif + /* If not enough room in bi_buf, use (valid) bits from bi_buf and + * (16 - bi_valid) bits from value, leaving (width - (16-bi_valid)) + * unused bits in value. + */ + if (bi_valid > (int)Buf_size - length) { + bi_buf |= (value << bi_valid); + put_short(bi_buf); + bi_buf = (ush)value >> (Buf_size - bi_valid); + bi_valid += length - Buf_size; + } else { + bi_buf |= value << bi_valid; + bi_valid += length; + } +} + +/* =========================================================================== + * Reverse the first len bits of a code, using straightforward code (a faster + * method would use a table) + * IN assertion: 1 <= len <= 15 + */ +unsigned bi_reverse(code, len) + unsigned code; /* the value to invert */ + int len; /* its bit length */ +{ + register unsigned res = 0; + do { + res |= code & 1; + code >>= 1, res <<= 1; + } while (--len > 0); + return res >> 1; +} + +/* =========================================================================== + * Write out any remaining bits in an incomplete byte. + */ +void bi_windup() +{ + if (bi_valid > 8) { + put_short(bi_buf); + } else if (bi_valid > 0) { + put_byte(bi_buf); + } + bi_buf = 0; + bi_valid = 0; +#ifdef DEBUG + bits_sent = (bits_sent+7) & ~7; +#endif +} + +/* =========================================================================== + * Copy a stored block to the zip file, storing first the length and its + * one's complement if requested. + */ +void copy_block(buf, len, header) + char *buf; /* the input data */ + unsigned len; /* its length */ + int header; /* true if block header must be written */ +{ + bi_windup(); /* align on byte boundary */ + + if (header) { + put_short((ush)len); + put_short((ush)~len); +#ifdef DEBUG + bits_sent += 2*16; +#endif + } +#ifdef DEBUG + bits_sent += (ulg)len<<3; +#endif + while (len--) { +#ifdef CRYPT + int t; + if (key) zencode(*buf, t); +#endif + put_byte(*buf++); + } +} +/* deflate.c -- compress data using the deflation algorithm + * Copyright (C) 1992-1993 Jean-loup Gailly + * This is free software; you can redistribute it and/or modify it under the + * terms of the GNU General Public License, see the file COPYING. + */ + +/* + * PURPOSE + * + * Identify new text as repetitions of old text within a fixed- + * length sliding window trailing behind the new text. + * + * DISCUSSION + * + * The "deflation" process depends on being able to identify portions + * of the input text which are identical to earlier input (within a + * sliding window trailing behind the input currently being processed). + * + * The most straightforward technique turns out to be the fastest for + * most input files: try all possible matches and select the longest. + * The key feature of this algorithm is that insertions into the string + * dictionary are very simple and thus fast, and deletions are avoided + * completely. Insertions are performed at each input character, whereas + * string matches are performed only when the previous match ends. So it + * is preferable to spend more time in matches to allow very fast string + * insertions and avoid deletions. The matching algorithm for small + * strings is inspired from that of Rabin & Karp. A brute force approach + * is used to find longer strings when a small match has been found. + * A similar algorithm is used in comic (by Jan-Mark Wams) and freeze + * (by Leonid Broukhis). + * A previous version of this file used a more sophisticated algorithm + * (by Fiala and Greene) which is guaranteed to run in linear amortized + * time, but has a larger average cost, uses more memory and is patented. + * However the F&G algorithm may be faster for some highly redundant + * files if the parameter max_chain_length (described below) is too large. + * + * ACKNOWLEDGEMENTS + * + * The idea of lazy evaluation of matches is due to Jan-Mark Wams, and + * I found it in 'freeze' written by Leonid Broukhis. + * Thanks to many info-zippers for bug reports and testing. + * + * REFERENCES + * + * APPNOTE.TXT documentation file in PKZIP 1.93a distribution. + * + * A description of the Rabin and Karp algorithm is given in the book + * "Algorithms" by R. Sedgewick, Addison-Wesley, p252. + * + * Fiala,E.R., and Greene,D.H. + * Data Compression with Finite Windows, Comm.ACM, 32,4 (1989) 490-595 + * + * INTERFACE + * + * void lm_init (int pack_level, ush *flags) + * Initialize the "longest match" routines for a new file + * + * ulg deflate (void) + * Processes a new input file and return its compressed length. Sets + * the compressed length, crc, deflate flags and internal file + * attributes. + */ + +#include <stdio.h> + +#ifdef RCSID +static char rcsid[] = "$Id: gzip.c,v 1.1 1999/10/05 16:24:56 andersen Exp $"; +#endif + +/* =========================================================================== + * Configuration parameters + */ + +/* Compile with MEDIUM_MEM to reduce the memory requirements or + * with SMALL_MEM to use as little memory as possible. Use BIG_MEM if the + * entire input file can be held in memory (not possible on 16 bit systems). + * Warning: defining these symbols affects HASH_BITS (see below) and thus + * affects the compression ratio. The compressed output + * is still correct, and might even be smaller in some cases. + */ + +#ifdef SMALL_MEM +# define HASH_BITS 13 /* Number of bits used to hash strings */ +#endif +#ifdef MEDIUM_MEM +# define HASH_BITS 14 +#endif +#ifndef HASH_BITS +# define HASH_BITS 15 + /* For portability to 16 bit machines, do not use values above 15. */ +#endif + +/* To save space (see unlzw.c), we overlay prev+head with tab_prefix and + * window with tab_suffix. Check that we can do this: + */ +#if (WSIZE<<1) > (1<<BITS) + error: cannot overlay window with tab_suffix and prev with tab_prefix0 +#endif +#if HASH_BITS > BITS-1 + error: cannot overlay head with tab_prefix1 +#endif + +#define HASH_SIZE (unsigned)(1<<HASH_BITS) +#define HASH_MASK (HASH_SIZE-1) +#define WMASK (WSIZE-1) +/* HASH_SIZE and WSIZE must be powers of two */ + +#define NIL 0 +/* Tail of hash chains */ + +#define FAST 4 +#define SLOW 2 +/* speed options for the general purpose bit flag */ + +#ifndef TOO_FAR +# define TOO_FAR 4096 +#endif +/* Matches of length 3 are discarded if their distance exceeds TOO_FAR */ + +/* =========================================================================== + * Local data used by the "longest match" routines. + */ + +typedef ush Pos; +typedef unsigned IPos; +/* A Pos is an index in the character window. We use short instead of int to + * save space in the various tables. IPos is used only for parameter passing. + */ + +/* DECLARE(uch, window, 2L*WSIZE); */ +/* Sliding window. Input bytes are read into the second half of the window, + * and move to the first half later to keep a dictionary of at least WSIZE + * bytes. With this organization, matches are limited to a distance of + * WSIZE-MAX_MATCH bytes, but this ensures that IO is always + * performed with a length multiple of the block size. Also, it limits + * the window size to 64K, which is quite useful on MSDOS. + * To do: limit the window size to WSIZE+BSZ if SMALL_MEM (the code would + * be less efficient). + */ + +/* DECLARE(Pos, prev, WSIZE); */ +/* Link to older string with same hash index. To limit the size of this + * array to 64K, this link is maintained only for the last 32K strings. + * An index in this array is thus a window index modulo 32K. + */ + +/* DECLARE(Pos, head, 1<<HASH_BITS); */ +/* Heads of the hash chains or NIL. */ + +ulg window_size = (ulg)2*WSIZE; +/* window size, 2*WSIZE except for MMAP or BIG_MEM, where it is the + * input file length plus MIN_LOOKAHEAD. + */ + +long block_start; +/* window position at the beginning of the current output block. Gets + * negative when the window is moved backwards. + */ + +local unsigned ins_h; /* hash index of string to be inserted */ + +#define H_SHIFT ((HASH_BITS+MIN_MATCH-1)/MIN_MATCH) +/* Number of bits by which ins_h and del_h must be shifted at each + * input step. It must be such that after MIN_MATCH steps, the oldest + * byte no longer takes part in the hash key, that is: + * H_SHIFT * MIN_MATCH >= HASH_BITS + */ + +unsigned int near prev_length; +/* Length of the best match at previous step. Matches not greater than this + * are discarded. This is used in the lazy match evaluation. + */ + + unsigned near strstart; /* start of string to insert */ + unsigned near match_start; /* start of matching string */ +local int eofile; /* flag set at end of input file */ +local unsigned lookahead; /* number of valid bytes ahead in window */ + +unsigned near max_chain_length; +/* To speed up deflation, hash chains are never searched beyond this length. + * A higher limit improves compression ratio but degrades the speed. + */ + +local unsigned int max_lazy_match; +/* Attempt to find a better match only when the current match is strictly + * smaller than this value. This mechanism is used only for compression + * levels >= 4. + */ +#define max_insert_length max_lazy_match +/* Insert new strings in the hash table only if the match length + * is not greater than this length. This saves time but degrades compression. + * max_insert_length is used only for compression levels <= 3. + */ + +unsigned near good_match; +/* Use a faster search when the previous match is longer than this */ + + +/* Values for max_lazy_match, good_match and max_chain_length, depending on + * the desired pack level (0..9). The values given below have been tuned to + * exclude worst case performance for pathological files. Better values may be + * found for specific files. + */ + +typedef struct config { + ush good_length; /* reduce lazy search above this match length */ + ush max_lazy; /* do not perform lazy search above this match length */ + ush nice_length; /* quit search above this match length */ + ush max_chain; +} config; + +#ifdef FULL_SEARCH +# define nice_match MAX_MATCH +#else + int near nice_match; /* Stop searching when current match exceeds this */ +#endif + +local config configuration_table = +/* 9 */ {32, 258, 258, 4096}; /* maximum compression */ + +/* Note: the deflate() code requires max_lazy >= MIN_MATCH and max_chain >= 4 + * For deflate_fast() (levels <= 3) good is ignored and lazy has a different + * meaning. + */ + +#define EQUAL 0 +/* result of memcmp for equal strings */ + +/* =========================================================================== + * Prototypes for local functions. + */ +local void fill_window OF((void)); + + int longest_match OF((IPos cur_match)); +#ifdef ASMV + void match_init OF((void)); /* asm code initialization */ +#endif + +#ifdef DEBUG +local void check_match OF((IPos start, IPos match, int length)); +#endif + +/* =========================================================================== + * Update a hash value with the given input byte + * IN assertion: all calls to to UPDATE_HASH are made with consecutive + * input characters, so that a running hash key can be computed from the + * previous key instead of complete recalculation each time. + */ +#define UPDATE_HASH(h,c) (h = (((h)<<H_SHIFT) ^ (c)) & HASH_MASK) + +/* =========================================================================== + * Insert string s in the dictionary and set match_head to the previous head + * of the hash chain (the most recent string with same hash key). Return + * the previous length of the hash chain. + * IN assertion: all calls to to INSERT_STRING are made with consecutive + * input characters and the first MIN_MATCH bytes of s are valid + * (except for the last MIN_MATCH-1 bytes of the input file). + */ +#define INSERT_STRING(s, match_head) \ + (UPDATE_HASH(ins_h, window[(s) + MIN_MATCH-1]), \ + prev[(s) & WMASK] = match_head = head[ins_h], \ + head[ins_h] = (s)) + +/* =========================================================================== + * Initialize the "longest match" routines for a new file + */ +void lm_init (flags) + ush *flags; /* general purpose bit flag */ +{ + register unsigned j; + + /* Initialize the hash table. */ +#if defined(MAXSEG_64K) && HASH_BITS == 15 + for (j = 0; j < HASH_SIZE; j++) head[j] = NIL; +#else + memzero((char*)head, HASH_SIZE*sizeof(*head)); +#endif + /* prev will be initialized on the fly */ + + /* Set the default configuration parameters: + */ + max_lazy_match = configuration_table.max_lazy; + good_match = configuration_table.good_length; +#ifndef FULL_SEARCH + nice_match = configuration_table.nice_length; +#endif + max_chain_length = configuration_table.max_chain; + *flags |= SLOW; + /* ??? reduce max_chain_length for binary files */ + + strstart = 0; + block_start = 0L; +#ifdef ASMV + match_init(); /* initialize the asm code */ +#endif + + lookahead = read_buf((char*)window, + sizeof(int) <= 2 ? (unsigned)WSIZE : 2*WSIZE); + + if (lookahead == 0 || lookahead == (unsigned)EOF) { + eofile = 1, lookahead = 0; + return; + } + eofile = 0; + /* Make sure that we always have enough lookahead. This is important + * if input comes from a device such as a tty. + */ + while (lookahead < MIN_LOOKAHEAD && !eofile) fill_window(); + + ins_h = 0; + for (j=0; j<MIN_MATCH-1; j++) UPDATE_HASH(ins_h, window[j]); + /* If lookahead < MIN_MATCH, ins_h is garbage, but this is + * not important since only literal bytes will be emitted. + */ +} + +/* =========================================================================== + * Set match_start to the longest match starting at the given string and + * return its length. Matches shorter or equal to prev_length are discarded, + * in which case the result is equal to prev_length and match_start is + * garbage. + * IN assertions: cur_match is the head of the hash chain for the current + * string (strstart) and its distance is <= MAX_DIST, and prev_length >= 1 + */ +#ifndef ASMV +/* For MSDOS, OS/2 and 386 Unix, an optimized version is in match.asm or + * match.s. The code is functionally equivalent, so you can use the C version + * if desired. + */ +int longest_match(cur_match) + IPos cur_match; /* current match */ +{ + unsigned chain_length = max_chain_length; /* max hash chain length */ + register uch *scan = window + strstart; /* current string */ + register uch *match; /* matched string */ + register int len; /* length of current match */ + int best_len = prev_length; /* best match length so far */ + IPos limit = strstart > (IPos)MAX_DIST ? strstart - (IPos)MAX_DIST : NIL; + /* Stop when cur_match becomes <= limit. To simplify the code, + * we prevent matches with the string of window index 0. + */ + +/* The code is optimized for HASH_BITS >= 8 and MAX_MATCH-2 multiple of 16. + * It is easy to get rid of this optimization if necessary. + */ +#if HASH_BITS < 8 || MAX_MATCH != 258 + error: Code too clever +#endif + +#ifdef UNALIGNED_OK + /* Compare two bytes at a time. Note: this is not always beneficial. + * Try with and without -DUNALIGNED_OK to check. + */ + register uch *strend = window + strstart + MAX_MATCH - 1; + register ush scan_start = *(ush*)scan; + register ush scan_end = *(ush*)(scan+best_len-1); +#else + register uch *strend = window + strstart + MAX_MATCH; + register uch scan_end1 = scan[best_len-1]; + register uch scan_end = scan[best_len]; +#endif + + /* Do not waste too much time if we already have a good match: */ + if (prev_length >= good_match) { + chain_length >>= 2; + } + Assert(strstart <= window_size-MIN_LOOKAHEAD, "insufficient lookahead"); + + do { + Assert(cur_match < strstart, "no future"); + match = window + cur_match; + + /* Skip to next match if the match length cannot increase + * or if the match length is less than 2: + */ +#if (defined(UNALIGNED_OK) && MAX_MATCH == 258) + /* This code assumes sizeof(unsigned short) == 2. Do not use + * UNALIGNED_OK if your compiler uses a different size. + */ + if (*(ush*)(match+best_len-1) != scan_end || + *(ush*)match != scan_start) continue; + + /* It is not necessary to compare scan[2] and match[2] since they are + * always equal when the other bytes match, given that the hash keys + * are equal and that HASH_BITS >= 8. Compare 2 bytes at a time at + * strstart+3, +5, ... up to strstart+257. We check for insufficient + * lookahead only every 4th comparison; the 128th check will be made + * at strstart+257. If MAX_MATCH-2 is not a multiple of 8, it is + * necessary to put more guard bytes at the end of the window, or + * to check more often for insufficient lookahead. + */ + scan++, match++; + do { + } while (*(ush*)(scan+=2) == *(ush*)(match+=2) && + *(ush*)(scan+=2) == *(ush*)(match+=2) && + *(ush*)(scan+=2) == *(ush*)(match+=2) && + *(ush*)(scan+=2) == *(ush*)(match+=2) && + scan < strend); + /* The funny "do {}" generates better code on most compilers */ + + /* Here, scan <= window+strstart+257 */ + Assert(scan <= window+(unsigned)(window_size-1), "wild scan"); + if (*scan == *match) scan++; + + len = (MAX_MATCH - 1) - (int)(strend-scan); + scan = strend - (MAX_MATCH-1); + +#else /* UNALIGNED_OK */ + + if (match[best_len] != scan_end || + match[best_len-1] != scan_end1 || + *match != *scan || + *++match != scan[1]) continue; + + /* The check at best_len-1 can be removed because it will be made + * again later. (This heuristic is not always a win.) + * It is not necessary to compare scan[2] and match[2] since they + * are always equal when the other bytes match, given that + * the hash keys are equal and that HASH_BITS >= 8. + */ + scan += 2, match++; + + /* We check for insufficient lookahead only every 8th comparison; + * the 256th check will be made at strstart+258. + */ + do { + } while (*++scan == *++match && *++scan == *++match && + *++scan == *++match && *++scan == *++match && + *++scan == *++match && *++scan == *++match && + *++scan == *++match && *++scan == *++match && + scan < strend); + + len = MAX_MATCH - (int)(strend - scan); + scan = strend - MAX_MATCH; + +#endif /* UNALIGNED_OK */ + + if (len > best_len) { + match_start = cur_match; + best_len = len; + if (len >= nice_match) break; +#ifdef UNALIGNED_OK + scan_end = *(ush*)(scan+best_len-1); +#else + scan_end1 = scan[best_len-1]; + scan_end = scan[best_len]; +#endif + } + } while ((cur_match = prev[cur_match & WMASK]) > limit + && --chain_length != 0); + + return best_len; +} +#endif /* ASMV */ + +#ifdef DEBUG +/* =========================================================================== + * Check that the match at match_start is indeed a match. + */ +local void check_match(start, match, length) + IPos start, match; + int length; +{ + /* check that the match is indeed a match */ + if (memcmp((char*)window + match, + (char*)window + start, length) != EQUAL) { + fprintf(stderr, + " start %d, match %d, length %d\n", + start, match, length); + error("invalid match"); + } + if (verbose > 1) { + fprintf(stderr,"\\[%d,%d]", start-match, length); + do { putc(window[start++], stderr); } while (--length != 0); + } +} +#else +# define check_match(start, match, length) +#endif + +/* =========================================================================== + * Fill the window when the lookahead becomes insufficient. + * Updates strstart and lookahead, and sets eofile if end of input file. + * IN assertion: lookahead < MIN_LOOKAHEAD && strstart + lookahead > 0 + * OUT assertions: at least one byte has been read, or eofile is set; + * file reads are performed for at least two bytes (required for the + * translate_eol option). + */ +local void fill_window() +{ + register unsigned n, m; + unsigned more = (unsigned)(window_size - (ulg)lookahead - (ulg)strstart); + /* Amount of free space at the end of the window. */ + + /* If the window is almost full and there is insufficient lookahead, + * move the upper half to the lower one to make room in the upper half. + */ + if (more == (unsigned)EOF) { + /* Very unlikely, but possible on 16 bit machine if strstart == 0 + * and lookahead == 1 (input done one byte at time) + */ + more--; + } else if (strstart >= WSIZE+MAX_DIST) { + /* By the IN assertion, the window is not empty so we can't confuse + * more == 0 with more == 64K on a 16 bit machine. + */ + Assert(window_size == (ulg)2*WSIZE, "no sliding with BIG_MEM"); + + memcpy((char*)window, (char*)window+WSIZE, (unsigned)WSIZE); + match_start -= WSIZE; + strstart -= WSIZE; /* we now have strstart >= MAX_DIST: */ + + block_start -= (long) WSIZE; + + for (n = 0; n < HASH_SIZE; n++) { + m = head[n]; + head[n] = (Pos)(m >= WSIZE ? m-WSIZE : NIL); + } + for (n = 0; n < WSIZE; n++) { + m = prev[n]; + prev[n] = (Pos)(m >= WSIZE ? m-WSIZE : NIL); + /* If n is not on any hash chain, prev[n] is garbage but + * its value will never be used. + */ + } + more += WSIZE; + } + /* At this point, more >= 2 */ + if (!eofile) { + n = read_buf((char*)window+strstart+lookahead, more); + if (n == 0 || n == (unsigned)EOF) { + eofile = 1; + } else { + lookahead += n; + } + } +} + +/* =========================================================================== + * Flush the current block, with given end-of-file flag. + * IN assertion: strstart is set to the end of the current match. + */ +#define FLUSH_BLOCK(eof) \ + flush_block(block_start >= 0L ? (char*)&window[(unsigned)block_start] : \ + (char*)NULL, (long)strstart - block_start, (eof)) + +/* =========================================================================== + * Same as above, but achieves better compression. We use a lazy + * evaluation for matches: a match is finally adopted only if there is + * no better match at the next window position. + */ +ulg deflate() +{ + IPos hash_head; /* head of hash chain */ + IPos prev_match; /* previous match */ + int flush; /* set if current block must be flushed */ + int match_available = 0; /* set if previous match exists */ + register unsigned match_length = MIN_MATCH-1; /* length of best match */ +#ifdef DEBUG + extern long isize; /* byte length of input file, for debug only */ +#endif + + /* Process the input block. */ + while (lookahead != 0) { + /* Insert the string window[strstart .. strstart+2] in the + * dictionary, and set hash_head to the head of the hash chain: + */ + INSERT_STRING(strstart, hash_head); + + /* Find the longest match, discarding those <= prev_length. + */ + prev_length = match_length, prev_match = match_start; + match_length = MIN_MATCH-1; + + if (hash_head != NIL && prev_length < max_lazy_match && + strstart - hash_head <= MAX_DIST) { + /* To simplify the code, we prevent matches with the string + * of window index 0 (in particular we have to avoid a match + * of the string with itself at the start of the input file). + */ + match_length = longest_match (hash_head); + /* longest_match() sets match_start */ + if (match_length > lookahead) match_length = lookahead; + + /* Ignore a length 3 match if it is too distant: */ + if (match_length == MIN_MATCH && strstart-match_start > TOO_FAR){ + /* If prev_match is also MIN_MATCH, match_start is garbage + * but we will ignore the current match anyway. + */ + match_length--; + } + } + /* If there was a match at the previous step and the current + * match is not better, output the previous match: + */ + if (prev_length >= MIN_MATCH && match_length <= prev_length) { + + check_match(strstart-1, prev_match, prev_length); + + flush = ct_tally(strstart-1-prev_match, prev_length - MIN_MATCH); + + /* Insert in hash table all strings up to the end of the match. + * strstart-1 and strstart are already inserted. + */ + lookahead -= prev_length-1; + prev_length -= 2; + do { + strstart++; + INSERT_STRING(strstart, hash_head); + /* strstart never exceeds WSIZE-MAX_MATCH, so there are + * always MIN_MATCH bytes ahead. If lookahead < MIN_MATCH + * these bytes are garbage, but it does not matter since the + * next lookahead bytes will always be emitted as literals. + */ + } while (--prev_length != 0); + match_available = 0; + match_length = MIN_MATCH-1; + strstart++; + if (flush) FLUSH_BLOCK(0), block_start = strstart; + + } else if (match_available) { + /* If there was no match at the previous position, output a + * single literal. If there was a match but the current match + * is longer, truncate the previous match to a single literal. + */ + Tracevv((stderr,"%c",window[strstart-1])); + if (ct_tally (0, window[strstart-1])) { + FLUSH_BLOCK(0), block_start = strstart; + } + strstart++; + lookahead--; + } else { + /* There is no previous match to compare with, wait for + * the next step to decide. + */ + match_available = 1; + strstart++; + lookahead--; + } + Assert (strstart <= isize && lookahead <= isize, "a bit too far"); + + /* Make sure that we always have enough lookahead, except + * at the end of the input file. We need MAX_MATCH bytes + * for the next match, plus MIN_MATCH bytes to insert the + * string following the next match. + */ + while (lookahead < MIN_LOOKAHEAD && !eofile) fill_window(); + } + if (match_available) ct_tally (0, window[strstart-1]); + + return FLUSH_BLOCK(1); /* eof */ +} +/* gzip (GNU zip) -- compress files with zip algorithm and 'compress' interface + * Copyright (C) 1992-1993 Jean-loup Gailly + * The unzip code was written and put in the public domain by Mark Adler. + * Portions of the lzw code are derived from the public domain 'compress' + * written by Spencer Thomas, Joe Orost, James Woods, Jim McKie, Steve Davies, + * Ken Turkowski, Dave Mack and Peter Jannesen. + * + * See the license_msg below and the file COPYING for the software license. + * See the file algorithm.doc for the compression algorithms and file formats. + */ + +/* Compress files with zip algorithm and 'compress' interface. + * See usage() and help() functions below for all options. + * Outputs: + * file.gz: compressed file with same mode, owner, and utimes + * or stdout with -c option or if stdin used as input. + * If the output file name had to be truncated, the original name is kept + * in the compressed file. + * On MSDOS, file.tmp -> file.tmz. On VMS, file.tmp -> file.tmp-gz. + * + * Using gz on MSDOS would create too many file name conflicts. For + * example, foo.txt -> foo.tgz (.tgz must be reserved as shorthand for + * tar.gz). Similarly, foo.dir and foo.doc would both be mapped to foo.dgz. + * I also considered 12345678.txt -> 12345txt.gz but this truncates the name + * too heavily. There is no ideal solution given the MSDOS 8+3 limitation. + * + * For the meaning of all compilation flags, see comments in Makefile.in. + */ + +#ifdef RCSID +static char rcsid[] = "$Id: gzip.c,v 1.1 1999/10/05 16:24:56 andersen Exp $"; +#endif + +#include <ctype.h> +#include <sys/types.h> +#include <signal.h> +#include <sys/stat.h> +#include <errno.h> + + /* configuration */ + +#ifdef NO_TIME_H +# include <sys/time.h> +#else +# include <time.h> +#endif + +#ifndef NO_FCNTL_H +# include <fcntl.h> +#endif + +#ifdef HAVE_UNISTD_H +# include <unistd.h> +#endif + +#if defined(STDC_HEADERS) || !defined(NO_STDLIB_H) +# include <stdlib.h> +#else + extern int errno; +#endif + +#if defined(DIRENT) +# include <dirent.h> + typedef struct dirent dir_type; +# define NLENGTH(dirent) ((int)strlen((dirent)->d_name)) +# define DIR_OPT "DIRENT" +#else +# define NLENGTH(dirent) ((dirent)->d_namlen) +# ifdef SYSDIR +# include <sys/dir.h> + typedef struct direct dir_type; +# define DIR_OPT "SYSDIR" +# else +# ifdef SYSNDIR +# include <sys/ndir.h> + typedef struct direct dir_type; +# define DIR_OPT "SYSNDIR" +# else +# ifdef NDIR +# include <ndir.h> + typedef struct direct dir_type; +# define DIR_OPT "NDIR" +# else +# define NO_DIR +# define DIR_OPT "NO_DIR" +# endif +# endif +# endif +#endif + +#ifndef NO_UTIME +# ifndef NO_UTIME_H +# include <utime.h> +# define TIME_OPT "UTIME" +# else +# ifdef HAVE_SYS_UTIME_H +# include <sys/utime.h> +# define TIME_OPT "SYS_UTIME" +# else + struct utimbuf { + time_t actime; + time_t modtime; + }; +# define TIME_OPT "" +# endif +# endif +#else +# define TIME_OPT "NO_UTIME" +#endif + +#if !defined(S_ISDIR) && defined(S_IFDIR) +# define S_ISDIR(m) (((m) & S_IFMT) == S_IFDIR) +#endif +#if !defined(S_ISREG) && defined(S_IFREG) +# define S_ISREG(m) (((m) & S_IFMT) == S_IFREG) +#endif + +typedef RETSIGTYPE (*sig_type) OF((int)); + +#ifndef O_BINARY +# define O_BINARY 0 /* creation mode for open() */ +#endif + +#ifndef O_CREAT + /* Pure BSD system? */ +# include <sys/file.h> +# ifndef O_CREAT +# define O_CREAT FCREAT +# endif +# ifndef O_EXCL +# define O_EXCL FEXCL +# endif +#endif + +#ifndef S_IRUSR +# define S_IRUSR 0400 +#endif +#ifndef S_IWUSR +# define S_IWUSR 0200 +#endif +#define RW_USER (S_IRUSR | S_IWUSR) /* creation mode for open() */ + +#ifndef MAX_PATH_LEN +# define MAX_PATH_LEN 1024 /* max pathname length */ +#endif + +#ifndef SEEK_END +# define SEEK_END 2 +#endif + +#ifdef NO_OFF_T + typedef long off_t; + off_t lseek OF((int fd, off_t offset, int whence)); +#endif + +/* Separator for file name parts (see shorten_name()) */ +#ifdef NO_MULTIPLE_DOTS +# define PART_SEP "-" +#else +# define PART_SEP "." +#endif + + /* global buffers */ + +DECLARE(uch, inbuf, INBUFSIZ +INBUF_EXTRA); +DECLARE(uch, outbuf, OUTBUFSIZ+OUTBUF_EXTRA); +DECLARE(ush, d_buf, DIST_BUFSIZE); +DECLARE(uch, window, 2L*WSIZE); +#ifndef MAXSEG_64K + DECLARE(ush, tab_prefix, 1L<<BITS); +#else + DECLARE(ush, tab_prefix0, 1L<<(BITS-1)); + DECLARE(ush, tab_prefix1, 1L<<(BITS-1)); +#endif + + /* local variables */ + +int ascii = 0; /* convert end-of-lines to local OS conventions */ +int to_stdout = 0; /* output to stdout (-c) */ +int decompress = 0; /* decompress (-d) */ +int no_name = -1; /* don't save or restore the original file name */ +int no_time = -1; /* don't save or restore the original file time */ +int foreground; /* set if program run in foreground */ +char *progname; /* program name */ +static int method = DEFLATED;/* compression method */ +static int exit_code = OK; /* program exit code */ +int save_orig_name; /* set if original name must be saved */ +int last_member; /* set for .zip and .Z files */ +int part_nb; /* number of parts in .gz file */ +long time_stamp; /* original time stamp (modification time) */ +long ifile_size; /* input file size, -1 for devices (debug only) */ +char *env; /* contents of GZIP env variable */ +char **args = NULL; /* argv pointer if GZIP env variable defined */ +char z_suffix[MAX_SUFFIX+1]; /* default suffix (can be set with --suffix) */ +int z_len; /* strlen(z_suffix) */ + +long bytes_in; /* number of input bytes */ +long bytes_out; /* number of output bytes */ +char ifname[MAX_PATH_LEN]; /* input file name */ +char ofname[MAX_PATH_LEN]; /* output file name */ +int remove_ofname = 0; /* remove output file on error */ +struct stat istat; /* status for input file */ +int ifd; /* input file descriptor */ +int ofd; /* output file descriptor */ +unsigned insize; /* valid bytes in inbuf */ +unsigned inptr; /* index of next byte to be processed in inbuf */ +unsigned outcnt; /* bytes in output buffer */ + +/* local functions */ + +local void treat_stdin OF((void)); +static int (*work) OF((int infile, int outfile)) = zip; /* function to call */ + +#define strequ(s1, s2) (strcmp((s1),(s2)) == 0) + +/* ======================================================================== */ +// int main (argc, argv) +// int argc; +// char **argv; +int gzip_main(struct FileInfo * i, int argc, char * * argv) +{ + foreground = signal(SIGINT, SIG_IGN) != SIG_IGN; + if (foreground) { + (void) signal (SIGINT, (sig_type)abort_gzip); + } +#ifdef SIGTERM + if (signal(SIGTERM, SIG_IGN) != SIG_IGN) { + (void) signal(SIGTERM, (sig_type)abort_gzip); + } +#endif +#ifdef SIGHUP + if (signal(SIGHUP, SIG_IGN) != SIG_IGN) { + (void) signal(SIGHUP, (sig_type)abort_gzip); + } +#endif + + strncpy(z_suffix, Z_SUFFIX, sizeof(z_suffix)-1); + z_len = strlen(z_suffix); + + to_stdout = 1; + + /* Allocate all global buffers (for DYN_ALLOC option) */ + ALLOC(uch, inbuf, INBUFSIZ +INBUF_EXTRA); + ALLOC(uch, outbuf, OUTBUFSIZ+OUTBUF_EXTRA); + ALLOC(ush, d_buf, DIST_BUFSIZE); + ALLOC(uch, window, 2L*WSIZE); +#ifndef MAXSEG_64K + ALLOC(ush, tab_prefix, 1L<<BITS); +#else + ALLOC(ush, tab_prefix0, 1L<<(BITS-1)); + ALLOC(ush, tab_prefix1, 1L<<(BITS-1)); +#endif + + /* And get to work */ + treat_stdin(); + do_exit(exit_code); + return exit_code; /* just to avoid lint warning */ +} + +/* ======================================================================== + * Compress or decompress stdin + */ +local void treat_stdin() +{ + SET_BINARY_MODE(fileno(stdout)); + strcpy(ifname, "stdin"); + strcpy(ofname, "stdout"); + + /* Get the time stamp on the input file. */ + time_stamp = 0; /* time unknown by default */ + + ifile_size = -1L; /* convention for unknown size */ + + clear_bufs(); /* clear input and output buffers */ + to_stdout = 1; + part_nb = 0; + + /* Actually do the compression/decompression. Loop over zipped members. + */ + if ((*work)(fileno(stdin), fileno(stdout)) != OK) return; +} + +/* ======================================================================== + * Free all dynamically allocated variables and exit with the given code. + */ +local void do_exit(int exitcode) +{ + static int in_exit = 0; + + if (in_exit) exit(exitcode); + in_exit = 1; + if (env != NULL) free(env), env = NULL; + if (args != NULL) free((char*)args), args = NULL; + FREE(inbuf); + FREE(outbuf); + FREE(d_buf); + FREE(window); +#ifndef MAXSEG_64K + FREE(tab_prefix); +#else + FREE(tab_prefix0); + FREE(tab_prefix1); +#endif + exit(exitcode); +} +/* trees.c -- output deflated data using Huffman coding + * Copyright (C) 1992-1993 Jean-loup Gailly + * This is free software; you can redistribute it and/or modify it under the + * terms of the GNU General Public License, see the file COPYING. + */ + +/* + * PURPOSE + * + * Encode various sets of source values using variable-length + * binary code trees. + * + * DISCUSSION + * + * The PKZIP "deflation" process uses several Huffman trees. The more + * common source values are represented by shorter bit sequences. + * + * Each code tree is stored in the ZIP file in a compressed form + * which is itself a Huffman encoding of the lengths of + * all the code strings (in ascending order by source values). + * The actual code strings are reconstructed from the lengths in + * the UNZIP process, as described in the "application note" + * (APPNOTE.TXT) distributed as part of PKWARE's PKZIP program. + * + * REFERENCES + * + * Lynch, Thomas J. + * Data Compression: Techniques and Applications, pp. 53-55. + * Lifetime Learning Publications, 1985. ISBN 0-534-03418-7. + * + * Storer, James A. + * Data Compression: Methods and Theory, pp. 49-50. + * Computer Science Press, 1988. ISBN 0-7167-8156-5. + * + * Sedgewick, R. + * Algorithms, p290. + * Addison-Wesley, 1983. ISBN 0-201-06672-6. + * + * INTERFACE + * + * void ct_init (ush *attr, int *methodp) + * Allocate the match buffer, initialize the various tables and save + * the location of the internal file attribute (ascii/binary) and + * method (DEFLATE/STORE) + * + * void ct_tally (int dist, int lc); + * Save the match info and tally the frequency counts. + * + * long flush_block (char *buf, ulg stored_len, int eof) + * Determine the best encoding for the current block: dynamic trees, + * static trees or store, and output the encoded block to the zip + * file. Returns the total compressed length for the file so far. + * + */ + +#include <ctype.h> + +#ifdef RCSID +static char rcsid[] = "$Id: gzip.c,v 1.1 1999/10/05 16:24:56 andersen Exp $"; +#endif + +/* =========================================================================== + * Constants + */ + +#define MAX_BITS 15 +/* All codes must not exceed MAX_BITS bits */ + +#define MAX_BL_BITS 7 +/* Bit length codes must not exceed MAX_BL_BITS bits */ + +#define LENGTH_CODES 29 +/* number of length codes, not counting the special END_BLOCK code */ + +#define LITERALS 256 +/* number of literal bytes 0..255 */ + +#define END_BLOCK 256 +/* end of block literal code */ + +#define L_CODES (LITERALS+1+LENGTH_CODES) +/* number of Literal or Length codes, including the END_BLOCK code */ + +#define D_CODES 30 +/* number of distance codes */ + +#define BL_CODES 19 +/* number of codes used to transfer the bit lengths */ + + +local int near extra_lbits[LENGTH_CODES] /* extra bits for each length code */ + = {0,0,0,0,0,0,0,0,1,1,1,1,2,2,2,2,3,3,3,3,4,4,4,4,5,5,5,5,0}; + +local int near extra_dbits[D_CODES] /* extra bits for each distance code */ + = {0,0,0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7,8,8,9,9,10,10,11,11,12,12,13,13}; + +local int near extra_blbits[BL_CODES]/* extra bits for each bit length code */ + = {0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,2,3,7}; + +#define STORED_BLOCK 0 +#define STATIC_TREES 1 +#define DYN_TREES 2 +/* The three kinds of block type */ + +#ifndef LIT_BUFSIZE +# ifdef SMALL_MEM +# define LIT_BUFSIZE 0x2000 +# else +# ifdef MEDIUM_MEM +# define LIT_BUFSIZE 0x4000 +# else +# define LIT_BUFSIZE 0x8000 +# endif +# endif +#endif +#ifndef DIST_BUFSIZE +# define DIST_BUFSIZE LIT_BUFSIZE +#endif +/* Sizes of match buffers for literals/lengths and distances. There are + * 4 reasons for limiting LIT_BUFSIZE to 64K: + * - frequencies can be kept in 16 bit counters + * - if compression is not successful for the first block, all input data is + * still in the window so we can still emit a stored block even when input + * comes from standard input. (This can also be done for all blocks if + * LIT_BUFSIZE is not greater than 32K.) + * - if compression is not successful for a file smaller than 64K, we can + * even emit a stored file instead of a stored block (saving 5 bytes). + * - creating new Huffman trees less frequently may not provide fast + * adaptation to changes in the input data statistics. (Take for + * example a binary file with poorly compressible code followed by + * a highly compressible string table.) Smaller buffer sizes give + * fast adaptation but have of course the overhead of transmitting trees + * more frequently. + * - I can't count above 4 + * The current code is general and allows DIST_BUFSIZE < LIT_BUFSIZE (to save + * memory at the expense of compression). Some optimizations would be possible + * if we rely on DIST_BUFSIZE == LIT_BUFSIZE. + */ +#if LIT_BUFSIZE > INBUFSIZ + error cannot overlay l_buf and inbuf +#endif + +#define REP_3_6 16 +/* repeat previous bit length 3-6 times (2 bits of repeat count) */ + +#define REPZ_3_10 17 +/* repeat a zero length 3-10 times (3 bits of repeat count) */ + +#define REPZ_11_138 18 +/* repeat a zero length 11-138 times (7 bits of repeat count) */ + +/* =========================================================================== + * Local data + */ + +/* Data structure describing a single value and its code string. */ +typedef struct ct_data { + union { + ush freq; /* frequency count */ + ush code; /* bit string */ + } fc; + union { + ush dad; /* father node in Huffman tree */ + ush len; /* length of bit string */ + } dl; +} ct_data; + +#define Freq fc.freq +#define Code fc.code +#define Dad dl.dad +#define Len dl.len + +#define HEAP_SIZE (2*L_CODES+1) +/* maximum heap size */ + +local ct_data near dyn_ltree[HEAP_SIZE]; /* literal and length tree */ +local ct_data near dyn_dtree[2*D_CODES+1]; /* distance tree */ + +local ct_data near static_ltree[L_CODES+2]; +/* The static literal tree. Since the bit lengths are imposed, there is no + * need for the L_CODES extra codes used during heap construction. However + * The codes 286 and 287 are needed to build a canonical tree (see ct_init + * below). + */ + +local ct_data near static_dtree[D_CODES]; +/* The static distance tree. (Actually a trivial tree since all codes use + * 5 bits.) + */ + +local ct_data near bl_tree[2*BL_CODES+1]; +/* Huffman tree for the bit lengths */ + +typedef struct tree_desc { + ct_data near *dyn_tree; /* the dynamic tree */ + ct_data near *static_tree; /* corresponding static tree or NULL */ + int near *extra_bits; /* extra bits for each code or NULL */ + int extra_base; /* base index for extra_bits */ + int elems; /* max number of elements in the tree */ + int max_length; /* max bit length for the codes */ + int max_code; /* largest code with non zero frequency */ +} tree_desc; + +local tree_desc near l_desc = +{dyn_ltree, static_ltree, extra_lbits, LITERALS+1, L_CODES, MAX_BITS, 0}; + +local tree_desc near d_desc = +{dyn_dtree, static_dtree, extra_dbits, 0, D_CODES, MAX_BITS, 0}; + +local tree_desc near bl_desc = +{bl_tree, (ct_data near *)0, extra_blbits, 0, BL_CODES, MAX_BL_BITS, 0}; + + +local ush near bl_count[MAX_BITS+1]; +/* number of codes at each bit length for an optimal tree */ + +local uch near bl_order[BL_CODES] + = {16,17,18,0,8,7,9,6,10,5,11,4,12,3,13,2,14,1,15}; +/* The lengths of the bit length codes are sent in order of decreasing + * probability, to avoid transmitting the lengths for unused bit length codes. + */ + +local int near heap[2*L_CODES+1]; /* heap used to build the Huffman trees */ +local int heap_len; /* number of elements in the heap */ +local int heap_max; /* element of largest frequency */ +/* The sons of heap[n] are heap[2*n] and heap[2*n+1]. heap[0] is not used. + * The same heap array is used to build all trees. + */ + +local uch near depth[2*L_CODES+1]; +/* Depth of each subtree used as tie breaker for trees of equal frequency */ + +local uch length_code[MAX_MATCH-MIN_MATCH+1]; +/* length code for each normalized match length (0 == MIN_MATCH) */ + +local uch dist_code[512]; +/* distance codes. The first 256 values correspond to the distances + * 3 .. 258, the last 256 values correspond to the top 8 bits of + * the 15 bit distances. + */ + +local int near base_length[LENGTH_CODES]; +/* First normalized length for each code (0 = MIN_MATCH) */ + +local int near base_dist[D_CODES]; +/* First normalized distance for each code (0 = distance of 1) */ + +#define l_buf inbuf +/* DECLARE(uch, l_buf, LIT_BUFSIZE); buffer for literals or lengths */ + +/* DECLARE(ush, d_buf, DIST_BUFSIZE); buffer for distances */ + +local uch near flag_buf[(LIT_BUFSIZE/8)]; +/* flag_buf is a bit array distinguishing literals from lengths in + * l_buf, thus indicating the presence or absence of a distance. + */ + +local unsigned last_lit; /* running index in l_buf */ +local unsigned last_dist; /* running index in d_buf */ +local unsigned last_flags; /* running index in flag_buf */ +local uch flags; /* current flags not yet saved in flag_buf */ +local uch flag_bit; /* current bit used in flags */ +/* bits are filled in flags starting at bit 0 (least significant). + * Note: these flags are overkill in the current code since we don't + * take advantage of DIST_BUFSIZE == LIT_BUFSIZE. + */ + +local ulg opt_len; /* bit length of current block with optimal trees */ +local ulg static_len; /* bit length of current block with static trees */ + +local ulg compressed_len; /* total bit length of compressed file */ + +local ulg input_len; /* total byte length of input file */ +/* input_len is for debugging only since we can get it by other means. */ + +ush *file_type; /* pointer to UNKNOWN, BINARY or ASCII */ +int *file_method; /* pointer to DEFLATE or STORE */ + +#ifdef DEBUG +extern ulg bits_sent; /* bit length of the compressed data */ +extern long isize; /* byte length of input file */ +#endif + +extern long block_start; /* window offset of current block */ +extern unsigned near strstart; /* window offset of current string */ + +/* =========================================================================== + * Local (static) routines in this file. + */ + +local void init_block OF((void)); +local void pqdownheap OF((ct_data near *tree, int k)); +local void gen_bitlen OF((tree_desc near *desc)); +local void gen_codes OF((ct_data near *tree, int max_code)); +local void build_tree OF((tree_desc near *desc)); +local void scan_tree OF((ct_data near *tree, int max_code)); +local void send_tree OF((ct_data near *tree, int max_code)); +local int build_bl_tree OF((void)); +local void send_all_trees OF((int lcodes, int dcodes, int blcodes)); +local void compress_block OF((ct_data near *ltree, ct_data near *dtree)); +local void set_file_type OF((void)); + + +#ifndef DEBUG +# define send_code(c, tree) send_bits(tree[c].Code, tree[c].Len) + /* Send a code of the given tree. c and tree must not have side effects */ + +#else /* DEBUG */ +# define send_code(c, tree) \ + { if (verbose>1) fprintf(stderr,"\ncd %3d ",(c)); \ + send_bits(tree[c].Code, tree[c].Len); } +#endif + +#define d_code(dist) \ + ((dist) < 256 ? dist_code[dist] : dist_code[256+((dist)>>7)]) +/* Mapping from a distance to a distance code. dist is the distance - 1 and + * must not have side effects. dist_code[256] and dist_code[257] are never + * used. + */ + +#define MAX(a,b) (a >= b ? a : b) +/* the arguments must not have side effects */ + +/* =========================================================================== + * Allocate the match buffer, initialize the various tables and save the + * location of the internal file attribute (ascii/binary) and method + * (DEFLATE/STORE). + */ +void ct_init(attr, methodp) + ush *attr; /* pointer to internal file attribute */ + int *methodp; /* pointer to compression method */ +{ + int n; /* iterates over tree elements */ + int bits; /* bit counter */ + int length; /* length value */ + int code; /* code value */ + int dist; /* distance index */ + + file_type = attr; + file_method = methodp; + compressed_len = input_len = 0L; + + if (static_dtree[0].Len != 0) return; /* ct_init already called */ + + /* Initialize the mapping length (0..255) -> length code (0..28) */ + length = 0; + for (code = 0; code < LENGTH_CODES-1; code++) { + base_length[code] = length; + for (n = 0; n < (1<<extra_lbits[code]); n++) { + length_code[length++] = (uch)code; + } + } + Assert (length == 256, "ct_init: length != 256"); + /* Note that the length 255 (match length 258) can be represented + * in two different ways: code 284 + 5 bits or code 285, so we + * overwrite length_code[255] to use the best encoding: + */ + length_code[length-1] = (uch)code; + + /* Initialize the mapping dist (0..32K) -> dist code (0..29) */ + dist = 0; + for (code = 0 ; code < 16; code++) { + base_dist[code] = dist; + for (n = 0; n < (1<<extra_dbits[code]); n++) { + dist_code[dist++] = (uch)code; + } + } + Assert (dist == 256, "ct_init: dist != 256"); + dist >>= 7; /* from now on, all distances are divided by 128 */ + for ( ; code < D_CODES; code++) { + base_dist[code] = dist << 7; + for (n = 0; n < (1<<(extra_dbits[code]-7)); n++) { + dist_code[256 + dist++] = (uch)code; + } + } + Assert (dist == 256, "ct_init: 256+dist != 512"); + + /* Construct the codes of the static literal tree */ + for (bits = 0; bits <= MAX_BITS; bits++) bl_count[bits] = 0; + n = 0; + while (n <= 143) static_ltree[n++].Len = 8, bl_count[8]++; + while (n <= 255) static_ltree[n++].Len = 9, bl_count[9]++; + while (n <= 279) static_ltree[n++].Len = 7, bl_count[7]++; + while (n <= 287) static_ltree[n++].Len = 8, bl_count[8]++; + /* Codes 286 and 287 do not exist, but we must include them in the + * tree construction to get a canonical Huffman tree (longest code + * all ones) + */ + gen_codes((ct_data near *)static_ltree, L_CODES+1); + + /* The static distance tree is trivial: */ + for (n = 0; n < D_CODES; n++) { + static_dtree[n].Len = 5; + static_dtree[n].Code = bi_reverse(n, 5); + } + + /* Initialize the first block of the first file: */ + init_block(); +} + +/* =========================================================================== + * Initialize a new block. + */ +local void init_block() +{ + int n; /* iterates over tree elements */ + + /* Initialize the trees. */ + for (n = 0; n < L_CODES; n++) dyn_ltree[n].Freq = 0; + for (n = 0; n < D_CODES; n++) dyn_dtree[n].Freq = 0; + for (n = 0; n < BL_CODES; n++) bl_tree[n].Freq = 0; + + dyn_ltree[END_BLOCK].Freq = 1; + opt_len = static_len = 0L; + last_lit = last_dist = last_flags = 0; + flags = 0; flag_bit = 1; +} + +#define SMALLEST 1 +/* Index within the heap array of least frequent node in the Huffman tree */ + + +/* =========================================================================== + * Remove the smallest element from the heap and recreate the heap with + * one less element. Updates heap and heap_len. + */ +#define pqremove(tree, top) \ +{\ + top = heap[SMALLEST]; \ + heap[SMALLEST] = heap[heap_len--]; \ + pqdownheap(tree, SMALLEST); \ +} + +/* =========================================================================== + * Compares to subtrees, using the tree depth as tie breaker when + * the subtrees have equal frequency. This minimizes the worst case length. + */ +#define smaller(tree, n, m) \ + (tree[n].Freq < tree[m].Freq || \ + (tree[n].Freq == tree[m].Freq && depth[n] <= depth[m])) + +/* =========================================================================== + * Restore the heap property by moving down the tree starting at node k, + * exchanging a node with the smallest of its two sons if necessary, stopping + * when the heap property is re-established (each father smaller than its + * two sons). + */ +local void pqdownheap(tree, k) + ct_data near *tree; /* the tree to restore */ + int k; /* node to move down */ +{ + int v = heap[k]; + int j = k << 1; /* left son of k */ + while (j <= heap_len) { + /* Set j to the smallest of the two sons: */ + if (j < heap_len && smaller(tree, heap[j+1], heap[j])) j++; + + /* Exit if v is smaller than both sons */ + if (smaller(tree, v, heap[j])) break; + + /* Exchange v with the smallest son */ + heap[k] = heap[j]; k = j; + + /* And continue down the tree, setting j to the left son of k */ + j <<= 1; + } + heap[k] = v; +} + +/* =========================================================================== + * Compute the optimal bit lengths for a tree and update the total bit length + * for the current block. + * IN assertion: the fields freq and dad are set, heap[heap_max] and + * above are the tree nodes sorted by increasing frequency. + * OUT assertions: the field len is set to the optimal bit length, the + * array bl_count contains the frequencies for each bit length. + * The length opt_len is updated; static_len is also updated if stree is + * not null. + */ +local void gen_bitlen(desc) + tree_desc near *desc; /* the tree descriptor */ +{ + ct_data near *tree = desc->dyn_tree; + int near *extra = desc->extra_bits; + int base = desc->extra_base; + int max_code = desc->max_code; + int max_length = desc->max_length; + ct_data near *stree = desc->static_tree; + int h; /* heap index */ + int n, m; /* iterate over the tree elements */ + int bits; /* bit length */ + int xbits; /* extra bits */ + ush f; /* frequency */ + int overflow = 0; /* number of elements with bit length too large */ + + for (bits = 0; bits <= MAX_BITS; bits++) bl_count[bits] = 0; + + /* In a first pass, compute the optimal bit lengths (which may + * overflow in the case of the bit length tree). + */ + tree[heap[heap_max]].Len = 0; /* root of the heap */ + + for (h = heap_max+1; h < HEAP_SIZE; h++) { + n = heap[h]; + bits = tree[tree[n].Dad].Len + 1; + if (bits > max_length) bits = max_length, overflow++; + tree[n].Len = (ush)bits; + /* We overwrite tree[n].Dad which is no longer needed */ + + if (n > max_code) continue; /* not a leaf node */ + + bl_count[bits]++; + xbits = 0; + if (n >= base) xbits = extra[n-base]; + f = tree[n].Freq; + opt_len += (ulg)f * (bits + xbits); + if (stree) static_len += (ulg)f * (stree[n].Len + xbits); + } + if (overflow == 0) return; + + Trace((stderr,"\nbit length overflow\n")); + /* This happens for example on obj2 and pic of the Calgary corpus */ + + /* Find the first bit length which could increase: */ + do { + bits = max_length-1; + while (bl_count[bits] == 0) bits--; + bl_count[bits]--; /* move one leaf down the tree */ + bl_count[bits+1] += 2; /* move one overflow item as its brother */ + bl_count[max_length]--; + /* The brother of the overflow item also moves one step up, + * but this does not affect bl_count[max_length] + */ + overflow -= 2; + } while (overflow > 0); + + /* Now recompute all bit lengths, scanning in increasing frequency. + * h is still equal to HEAP_SIZE. (It is simpler to reconstruct all + * lengths instead of fixing only the wrong ones. This idea is taken + * from 'ar' written by Haruhiko Okumura.) + */ + for (bits = max_length; bits != 0; bits--) { + n = bl_count[bits]; + while (n != 0) { + m = heap[--h]; + if (m > max_code) continue; + if (tree[m].Len != (unsigned) bits) { + Trace((stderr,"code %d bits %d->%d\n", m, tree[m].Len, bits)); + opt_len += ((long)bits-(long)tree[m].Len)*(long)tree[m].Freq; + tree[m].Len = (ush)bits; + } + n--; + } + } +} + +/* =========================================================================== + * Generate the codes for a given tree and bit counts (which need not be + * optimal). + * IN assertion: the array bl_count contains the bit length statistics for + * the given tree and the field len is set for all tree elements. + * OUT assertion: the field code is set for all tree elements of non + * zero code length. + */ +local void gen_codes (tree, max_code) + ct_data near *tree; /* the tree to decorate */ + int max_code; /* largest code with non zero frequency */ +{ + ush next_code[MAX_BITS+1]; /* next code value for each bit length */ + ush code = 0; /* running code value */ + int bits; /* bit index */ + int n; /* code index */ + + /* The distribution counts are first used to generate the code values + * without bit reversal. + */ + for (bits = 1; bits <= MAX_BITS; bits++) { + next_code[bits] = code = (code + bl_count[bits-1]) << 1; + } + /* Check that the bit counts in bl_count are consistent. The last code + * must be all ones. + */ + Assert (code + bl_count[MAX_BITS]-1 == (1<<MAX_BITS)-1, + "inconsistent bit counts"); + Tracev((stderr,"\ngen_codes: max_code %d ", max_code)); + + for (n = 0; n <= max_code; n++) { + int len = tree[n].Len; + if (len == 0) continue; + /* Now reverse the bits */ + tree[n].Code = bi_reverse(next_code[len]++, len); + + Tracec(tree != static_ltree, (stderr,"\nn %3d %c l %2d c %4x (%x) ", + n, (isgraph(n) ? n : ' '), len, tree[n].Code, next_code[len]-1)); + } +} + +/* =========================================================================== + * Construct one Huffman tree and assigns the code bit strings and lengths. + * Update the total bit length for the current block. + * IN assertion: the field freq is set for all tree elements. + * OUT assertions: the fields len and code are set to the optimal bit length + * and corresponding code. The length opt_len is updated; static_len is + * also updated if stree is not null. The field max_code is set. + */ +local void build_tree(desc) + tree_desc near *desc; /* the tree descriptor */ +{ + ct_data near *tree = desc->dyn_tree; + ct_data near *stree = desc->static_tree; + int elems = desc->elems; + int n, m; /* iterate over heap elements */ + int max_code = -1; /* largest code with non zero frequency */ + int node = elems; /* next internal node of the tree */ + + /* Construct the initial heap, with least frequent element in + * heap[SMALLEST]. The sons of heap[n] are heap[2*n] and heap[2*n+1]. + * heap[0] is not used. + */ + heap_len = 0, heap_max = HEAP_SIZE; + + for (n = 0; n < elems; n++) { + if (tree[n].Freq != 0) { + heap[++heap_len] = max_code = n; + depth[n] = 0; + } else { + tree[n].Len = 0; + } + } + + /* The pkzip format requires that at least one distance code exists, + * and that at least one bit should be sent even if there is only one + * possible code. So to avoid special checks later on we force at least + * two codes of non zero frequency. + */ + while (heap_len < 2) { + int new = heap[++heap_len] = (max_code < 2 ? ++max_code : 0); + tree[new].Freq = 1; + depth[new] = 0; + opt_len--; if (stree) static_len -= stree[new].Len; + /* new is 0 or 1 so it does not have extra bits */ + } + desc->max_code = max_code; + + /* The elements heap[heap_len/2+1 .. heap_len] are leaves of the tree, + * establish sub-heaps of increasing lengths: + */ + for (n = heap_len/2; n >= 1; n--) pqdownheap(tree, n); + + /* Construct the Huffman tree by repeatedly combining the least two + * frequent nodes. + */ + do { + pqremove(tree, n); /* n = node of least frequency */ + m = heap[SMALLEST]; /* m = node of next least frequency */ + + heap[--heap_max] = n; /* keep the nodes sorted by frequency */ + heap[--heap_max] = m; + + /* Create a new node father of n and m */ + tree[node].Freq = tree[n].Freq + tree[m].Freq; + depth[node] = (uch) (MAX(depth[n], depth[m]) + 1); + tree[n].Dad = tree[m].Dad = (ush)node; +#ifdef DUMP_BL_TREE + if (tree == bl_tree) { + fprintf(stderr,"\nnode %d(%d), sons %d(%d) %d(%d)", + node, tree[node].Freq, n, tree[n].Freq, m, tree[m].Freq); + } +#endif + /* and insert the new node in the heap */ + heap[SMALLEST] = node++; + pqdownheap(tree, SMALLEST); + + } while (heap_len >= 2); + + heap[--heap_max] = heap[SMALLEST]; + + /* At this point, the fields freq and dad are set. We can now + * generate the bit lengths. + */ + gen_bitlen((tree_desc near *)desc); + + /* The field len is now set, we can generate the bit codes */ + gen_codes ((ct_data near *)tree, max_code); +} + +/* =========================================================================== + * Scan a literal or distance tree to determine the frequencies of the codes + * in the bit length tree. Updates opt_len to take into account the repeat + * counts. (The contribution of the bit length codes will be added later + * during the construction of bl_tree.) + */ +local void scan_tree (tree, max_code) + ct_data near *tree; /* the tree to be scanned */ + int max_code; /* and its largest code of non zero frequency */ +{ + int n; /* iterates over all tree elements */ + int prevlen = -1; /* last emitted length */ + int curlen; /* length of current code */ + int nextlen = tree[0].Len; /* length of next code */ + int count = 0; /* repeat count of the current code */ + int max_count = 7; /* max repeat count */ + int min_count = 4; /* min repeat count */ + + if (nextlen == 0) max_count = 138, min_count = 3; + tree[max_code+1].Len = (ush)0xffff; /* guard */ + + for (n = 0; n <= max_code; n++) { + curlen = nextlen; nextlen = tree[n+1].Len; + if (++count < max_count && curlen == nextlen) { + continue; + } else if (count < min_count) { + bl_tree[curlen].Freq += count; + } else if (curlen != 0) { + if (curlen != prevlen) bl_tree[curlen].Freq++; + bl_tree[REP_3_6].Freq++; + } else if (count <= 10) { + bl_tree[REPZ_3_10].Freq++; + } else { + bl_tree[REPZ_11_138].Freq++; + } + count = 0; prevlen = curlen; + if (nextlen == 0) { + max_count = 138, min_count = 3; + } else if (curlen == nextlen) { + max_count = 6, min_count = 3; + } else { + max_count = 7, min_count = 4; + } + } +} + +/* =========================================================================== + * Send a literal or distance tree in compressed form, using the codes in + * bl_tree. + */ +local void send_tree (tree, max_code) + ct_data near *tree; /* the tree to be scanned */ + int max_code; /* and its largest code of non zero frequency */ +{ + int n; /* iterates over all tree elements */ + int prevlen = -1; /* last emitted length */ + int curlen; /* length of current code */ + int nextlen = tree[0].Len; /* length of next code */ + int count = 0; /* repeat count of the current code */ + int max_count = 7; /* max repeat count */ + int min_count = 4; /* min repeat count */ + + /* tree[max_code+1].Len = -1; */ /* guard already set */ + if (nextlen == 0) max_count = 138, min_count = 3; + + for (n = 0; n <= max_code; n++) { + curlen = nextlen; nextlen = tree[n+1].Len; + if (++count < max_count && curlen == nextlen) { + continue; + } else if (count < min_count) { + do { send_code(curlen, bl_tree); } while (--count != 0); + + } else if (curlen != 0) { + if (curlen != prevlen) { + send_code(curlen, bl_tree); count--; + } + Assert(count >= 3 && count <= 6, " 3_6?"); + send_code(REP_3_6, bl_tree); send_bits(count-3, 2); + + } else if (count <= 10) { + send_code(REPZ_3_10, bl_tree); send_bits(count-3, 3); + + } else { + send_code(REPZ_11_138, bl_tree); send_bits(count-11, 7); + } + count = 0; prevlen = curlen; + if (nextlen == 0) { + max_count = 138, min_count = 3; + } else if (curlen == nextlen) { + max_count = 6, min_count = 3; + } else { + max_count = 7, min_count = 4; + } + } +} + +/* =========================================================================== + * Construct the Huffman tree for the bit lengths and return the index in + * bl_order of the last bit length code to send. + */ +local int build_bl_tree() +{ + int max_blindex; /* index of last bit length code of non zero freq */ + + /* Determine the bit length frequencies for literal and distance trees */ + scan_tree((ct_data near *)dyn_ltree, l_desc.max_code); + scan_tree((ct_data near *)dyn_dtree, d_desc.max_code); + + /* Build the bit length tree: */ + build_tree((tree_desc near *)(&bl_desc)); + /* opt_len now includes the length of the tree representations, except + * the lengths of the bit lengths codes and the 5+5+4 bits for the counts. + */ + + /* Determine the number of bit length codes to send. The pkzip format + * requires that at least 4 bit length codes be sent. (appnote.txt says + * 3 but the actual value used is 4.) + */ + for (max_blindex = BL_CODES-1; max_blindex >= 3; max_blindex--) { + if (bl_tree[bl_order[max_blindex]].Len != 0) break; + } + /* Update opt_len to include the bit length tree and counts */ + opt_len += 3*(max_blindex+1) + 5+5+4; + Tracev((stderr, "\ndyn trees: dyn %ld, stat %ld", opt_len, static_len)); + + return max_blindex; +} + +/* =========================================================================== + * Send the header for a block using dynamic Huffman trees: the counts, the + * lengths of the bit length codes, the literal tree and the distance tree. + * IN assertion: lcodes >= 257, dcodes >= 1, blcodes >= 4. + */ +local void send_all_trees(lcodes, dcodes, blcodes) + int lcodes, dcodes, blcodes; /* number of codes for each tree */ +{ + int rank; /* index in bl_order */ + + Assert (lcodes >= 257 && dcodes >= 1 && blcodes >= 4, "not enough codes"); + Assert (lcodes <= L_CODES && dcodes <= D_CODES && blcodes <= BL_CODES, + "too many codes"); + Tracev((stderr, "\nbl counts: ")); + send_bits(lcodes-257, 5); /* not +255 as stated in appnote.txt */ + send_bits(dcodes-1, 5); + send_bits(blcodes-4, 4); /* not -3 as stated in appnote.txt */ + for (rank = 0; rank < blcodes; rank++) { + Tracev((stderr, "\nbl code %2d ", bl_order[rank])); + send_bits(bl_tree[bl_order[rank]].Len, 3); + } + Tracev((stderr, "\nbl tree: sent %ld", bits_sent)); + + send_tree((ct_data near *)dyn_ltree, lcodes-1); /* send the literal tree */ + Tracev((stderr, "\nlit tree: sent %ld", bits_sent)); + + send_tree((ct_data near *)dyn_dtree, dcodes-1); /* send the distance tree */ + Tracev((stderr, "\ndist tree: sent %ld", bits_sent)); +} + +/* =========================================================================== + * Determine the best encoding for the current block: dynamic trees, static + * trees or store, and output the encoded block to the zip file. This function + * returns the total compressed length for the file so far. + */ +ulg flush_block(buf, stored_len, eof) + char *buf; /* input block, or NULL if too old */ + ulg stored_len; /* length of input block */ + int eof; /* true if this is the last block for a file */ +{ + ulg opt_lenb, static_lenb; /* opt_len and static_len in bytes */ + int max_blindex; /* index of last bit length code of non zero freq */ + + flag_buf[last_flags] = flags; /* Save the flags for the last 8 items */ + + /* Check if the file is ascii or binary */ + if (*file_type == (ush)UNKNOWN) set_file_type(); + + /* Construct the literal and distance trees */ + build_tree((tree_desc near *)(&l_desc)); + Tracev((stderr, "\nlit data: dyn %ld, stat %ld", opt_len, static_len)); + + build_tree((tree_desc near *)(&d_desc)); + Tracev((stderr, "\ndist data: dyn %ld, stat %ld", opt_len, static_len)); + /* At this point, opt_len and static_len are the total bit lengths of + * the compressed block data, excluding the tree representations. + */ + + /* Build the bit length tree for the above two trees, and get the index + * in bl_order of the last bit length code to send. + */ + max_blindex = build_bl_tree(); + + /* Determine the best encoding. Compute first the block length in bytes */ + opt_lenb = (opt_len+3+7)>>3; + static_lenb = (static_len+3+7)>>3; + input_len += stored_len; /* for debugging only */ + + Trace((stderr, "\nopt %lu(%lu) stat %lu(%lu) stored %lu lit %u dist %u ", + opt_lenb, opt_len, static_lenb, static_len, stored_len, + last_lit, last_dist)); + + if (static_lenb <= opt_lenb) opt_lenb = static_lenb; + + /* If compression failed and this is the first and last block, + * and if the zip file can be seeked (to rewrite the local header), + * the whole file is transformed into a stored file: + */ +#ifdef FORCE_METHOD +#else + if (stored_len <= opt_lenb && eof && compressed_len == 0L && seekable()) { +#endif + /* Since LIT_BUFSIZE <= 2*WSIZE, the input data must be there: */ + if (buf == (char*)0) error ("block vanished"); + + copy_block(buf, (unsigned)stored_len, 0); /* without header */ + compressed_len = stored_len << 3; + *file_method = STORED; + +#ifdef FORCE_METHOD +#else + } else if (stored_len+4 <= opt_lenb && buf != (char*)0) { + /* 4: two words for the lengths */ +#endif + /* The test buf != NULL is only necessary if LIT_BUFSIZE > WSIZE. + * Otherwise we can't have processed more than WSIZE input bytes since + * the last block flush, because compression would have been + * successful. If LIT_BUFSIZE <= WSIZE, it is never too late to + * transform a block into a stored block. + */ + send_bits((STORED_BLOCK<<1)+eof, 3); /* send block type */ + compressed_len = (compressed_len + 3 + 7) & ~7L; + compressed_len += (stored_len + 4) << 3; + + copy_block(buf, (unsigned)stored_len, 1); /* with header */ + +#ifdef FORCE_METHOD +#else + } else if (static_lenb == opt_lenb) { +#endif + send_bits((STATIC_TREES<<1)+eof, 3); + compress_block((ct_data near *)static_ltree, (ct_data near *)static_dtree); + compressed_len += 3 + static_len; + } else { + send_bits((DYN_TREES<<1)+eof, 3); + send_all_trees(l_desc.max_code+1, d_desc.max_code+1, max_blindex+1); + compress_block((ct_data near *)dyn_ltree, (ct_data near *)dyn_dtree); + compressed_len += 3 + opt_len; + } + Assert (compressed_len == bits_sent, "bad compressed size"); + init_block(); + + if (eof) { + Assert (input_len == isize, "bad input size"); + bi_windup(); + compressed_len += 7; /* align on byte boundary */ + } + Tracev((stderr,"\ncomprlen %lu(%lu) ", compressed_len>>3, + compressed_len-7*eof)); + + return compressed_len >> 3; +} + +/* =========================================================================== + * Save the match info and tally the frequency counts. Return true if + * the current block must be flushed. + */ +int ct_tally (dist, lc) + int dist; /* distance of matched string */ + int lc; /* match length-MIN_MATCH or unmatched char (if dist==0) */ +{ + l_buf[last_lit++] = (uch)lc; + if (dist == 0) { + /* lc is the unmatched char */ + dyn_ltree[lc].Freq++; + } else { + /* Here, lc is the match length - MIN_MATCH */ + dist--; /* dist = match distance - 1 */ + Assert((ush)dist < (ush)MAX_DIST && + (ush)lc <= (ush)(MAX_MATCH-MIN_MATCH) && + (ush)d_code(dist) < (ush)D_CODES, "ct_tally: bad match"); + + dyn_ltree[length_code[lc]+LITERALS+1].Freq++; + dyn_dtree[d_code(dist)].Freq++; + + d_buf[last_dist++] = (ush)dist; + flags |= flag_bit; + } + flag_bit <<= 1; + + /* Output the flags if they fill a byte: */ + if ((last_lit & 7) == 0) { + flag_buf[last_flags++] = flags; + flags = 0, flag_bit = 1; + } + /* Try to guess if it is profitable to stop the current block here */ + if ((last_lit & 0xfff) == 0) { + /* Compute an upper bound for the compressed length */ + ulg out_length = (ulg)last_lit*8L; + ulg in_length = (ulg)strstart-block_start; + int dcode; + for (dcode = 0; dcode < D_CODES; dcode++) { + out_length += (ulg)dyn_dtree[dcode].Freq*(5L+extra_dbits[dcode]); + } + out_length >>= 3; + Trace((stderr,"\nlast_lit %u, last_dist %u, in %ld, out ~%ld(%ld%%) ", + last_lit, last_dist, in_length, out_length, + 100L - out_length*100L/in_length)); + if (last_dist < last_lit/2 && out_length < in_length/2) return 1; + } + return (last_lit == LIT_BUFSIZE-1 || last_dist == DIST_BUFSIZE); + /* We avoid equality with LIT_BUFSIZE because of wraparound at 64K + * on 16 bit machines and because stored blocks are restricted to + * 64K-1 bytes. + */ +} + +/* =========================================================================== + * Send the block data compressed using the given Huffman trees + */ +local void compress_block(ltree, dtree) + ct_data near *ltree; /* literal tree */ + ct_data near *dtree; /* distance tree */ +{ + unsigned dist; /* distance of matched string */ + int lc; /* match length or unmatched char (if dist == 0) */ + unsigned lx = 0; /* running index in l_buf */ + unsigned dx = 0; /* running index in d_buf */ + unsigned fx = 0; /* running index in flag_buf */ + uch flag = 0; /* current flags */ + unsigned code; /* the code to send */ + int extra; /* number of extra bits to send */ + + if (last_lit != 0) do { + if ((lx & 7) == 0) flag = flag_buf[fx++]; + lc = l_buf[lx++]; + if ((flag & 1) == 0) { + send_code(lc, ltree); /* send a literal byte */ + Tracecv(isgraph(lc), (stderr," '%c' ", lc)); + } else { + /* Here, lc is the match length - MIN_MATCH */ + code = length_code[lc]; + send_code(code+LITERALS+1, ltree); /* send the length code */ + extra = extra_lbits[code]; + if (extra != 0) { + lc -= base_length[code]; + send_bits(lc, extra); /* send the extra length bits */ + } + dist = d_buf[dx++]; + /* Here, dist is the match distance - 1 */ + code = d_code(dist); + Assert (code < D_CODES, "bad d_code"); + + send_code(code, dtree); /* send the distance code */ + extra = extra_dbits[code]; + if (extra != 0) { + dist -= base_dist[code]; + send_bits(dist, extra); /* send the extra distance bits */ + } + } /* literal or match pair ? */ + flag >>= 1; + } while (lx < last_lit); + + send_code(END_BLOCK, ltree); +} + +/* =========================================================================== + * Set the file type to ASCII or BINARY, using a crude approximation: + * binary if more than 20% of the bytes are <= 6 or >= 128, ascii otherwise. + * IN assertion: the fields freq of dyn_ltree are set and the total of all + * frequencies does not exceed 64K (to fit in an int on 16 bit machines). + */ +local void set_file_type() +{ + int n = 0; + unsigned ascii_freq = 0; + unsigned bin_freq = 0; + while (n < 7) bin_freq += dyn_ltree[n++].Freq; + while (n < 128) ascii_freq += dyn_ltree[n++].Freq; + while (n < LITERALS) bin_freq += dyn_ltree[n++].Freq; + *file_type = bin_freq > (ascii_freq >> 2) ? BINARY : ASCII; + if (*file_type == BINARY && translate_eol) { + warn("-l used on binary file", ""); + } +} +/* util.c -- utility functions for gzip support + * Copyright (C) 1992-1993 Jean-loup Gailly + * This is free software; you can redistribute it and/or modify it under the + * terms of the GNU General Public License, see the file COPYING. + */ + +#ifdef RCSID +static char rcsid[] = "$Id: gzip.c,v 1.1 1999/10/05 16:24:56 andersen Exp $"; +#endif + +#include <ctype.h> +#include <errno.h> +#include <sys/types.h> + +#ifdef HAVE_UNISTD_H +# include <unistd.h> +#endif +#ifndef NO_FCNTL_H +# include <fcntl.h> +#endif + +#if defined(STDC_HEADERS) || !defined(NO_STDLIB_H) +# include <stdlib.h> +#else + extern int errno; +#endif + +extern ulg crc_32_tab[]; /* crc table, defined below */ + +/* =========================================================================== + * Copy input to output unchanged: zcat == cat with --force. + * IN assertion: insize bytes have already been read in inbuf. + */ +int copy(in, out) + int in, out; /* input and output file descriptors */ +{ + errno = 0; + while (insize != 0 && (int)insize != EOF) { + write_buf(out, (char*)inbuf, insize); + bytes_out += insize; + insize = read(in, (char*)inbuf, INBUFSIZ); + } + if ((int)insize == EOF && errno != 0) { + read_error(); + } + bytes_in = bytes_out; + return OK; +} + +/* ======================================================================== + * Put string s in lower case, return s. + */ +char *strlwr(s) + char *s; +{ + char *t; + for (t = s; *t; t++) *t = tolow(*t); + return s; +} + +#if defined(NO_STRING_H) && !defined(STDC_HEADERS) + +/* Provide missing strspn and strcspn functions. */ + +# ifndef __STDC__ +# define const +# endif + +int strspn OF((const char *s, const char *accept)); +int strcspn OF((const char *s, const char *reject)); + +/* ======================================================================== + * Return the length of the maximum initial segment + * of s which contains only characters in accept. + */ +int strspn(s, accept) + const char *s; + const char *accept; +{ + register const char *p; + register const char *a; + register int count = 0; + + for (p = s; *p != '\0'; ++p) { + for (a = accept; *a != '\0'; ++a) { + if (*p == *a) break; + } + if (*a == '\0') return count; + ++count; + } + return count; +} + +/* ======================================================================== + * Return the length of the maximum inital segment of s + * which contains no characters from reject. + */ +int strcspn(s, reject) + const char *s; + const char *reject; +{ + register int count = 0; + + while (*s != '\0') { + if (strchr(reject, *s++) != NULL) return count; + ++count; + } + return count; +} + +#endif /* NO_STRING_H */ + +/* ======================================================================== + * Add an environment variable (if any) before argv, and update argc. + * Return the expanded environment variable to be freed later, or NULL + * if no options were added to argv. + */ +#define SEPARATOR " \t" /* separators in env variable */ + +char *add_envopt(argcp, argvp, env) + int *argcp; /* pointer to argc */ + char ***argvp; /* pointer to argv */ + char *env; /* name of environment variable */ +{ + char *p; /* running pointer through env variable */ + char **oargv; /* runs through old argv array */ + char **nargv; /* runs through new argv array */ + int oargc = *argcp; /* old argc */ + int nargc = 0; /* number of arguments in env variable */ + + env = (char*)getenv(env); + if (env == NULL) return NULL; + + p = (char*)xmalloc(strlen(env)+1); + env = strcpy(p, env); /* keep env variable intact */ + + for (p = env; *p; nargc++ ) { /* move through env */ + p += strspn(p, SEPARATOR); /* skip leading separators */ + if (*p == '\0') break; + + p += strcspn(p, SEPARATOR); /* find end of word */ + if (*p) *p++ = '\0'; /* mark it */ + } + if (nargc == 0) { + free(env); + return NULL; + } + *argcp += nargc; + /* Allocate the new argv array, with an extra element just in case + * the original arg list did not end with a NULL. + */ + nargv = (char**)calloc(*argcp+1, sizeof(char *)); + if (nargv == NULL) error("out of memory"); + oargv = *argvp; + *argvp = nargv; + + /* Copy the program name first */ + if (oargc-- < 0) error("argc<=0"); + *(nargv++) = *(oargv++); + + /* Then copy the environment args */ + for (p = env; nargc > 0; nargc--) { + p += strspn(p, SEPARATOR); /* skip separators */ + *(nargv++) = p; /* store start */ + while (*p++) ; /* skip over word */ + } + + /* Finally copy the old args and add a NULL (usual convention) */ + while (oargc--) *(nargv++) = *(oargv++); + *nargv = NULL; + return env; +} +/* ======================================================================== + * Display compression ratio on the given stream on 6 characters. + */ +void display_ratio(num, den, file) + long num; + long den; + FILE *file; +{ + long ratio; /* 1000 times the compression ratio */ + + if (den == 0) { + ratio = 0; /* no compression */ + } else if (den < 2147483L) { /* (2**31 -1)/1000 */ + ratio = 1000L*num/den; + } else { + ratio = num/(den/1000L); + } + if (ratio < 0) { + putc('-', file); + ratio = -ratio; + } else { + putc(' ', file); + } + fprintf(file, "%2ld.%1ld%%", ratio / 10L, ratio % 10L); +} + + +/* zip.c -- compress files to the gzip or pkzip format + * Copyright (C) 1992-1993 Jean-loup Gailly + * This is free software; you can redistribute it and/or modify it under the + * terms of the GNU General Public License, see the file COPYING. + */ + +#ifdef RCSID +static char rcsid[] = "$Id: gzip.c,v 1.1 1999/10/05 16:24:56 andersen Exp $"; +#endif + +#include <ctype.h> +#include <sys/types.h> + +#ifdef HAVE_UNISTD_H +# include <unistd.h> +#endif +#ifndef NO_FCNTL_H +# include <fcntl.h> +#endif + +local ulg crc; /* crc on uncompressed file data */ +long header_bytes; /* number of bytes in gzip header */ + +/* =========================================================================== + * Deflate in to out. + * IN assertions: the input and output buffers are cleared. + * The variables time_stamp and save_orig_name are initialized. + */ +int zip(in, out) + int in, out; /* input and output file descriptors */ +{ + uch flags = 0; /* general purpose bit flags */ + ush attr = 0; /* ascii/binary flag */ + ush deflate_flags = 0; /* pkzip -es, -en or -ex equivalent */ + + ifd = in; + ofd = out; + outcnt = 0; + + /* Write the header to the gzip file. See algorithm.doc for the format */ + + method = DEFLATED; + put_byte(GZIP_MAGIC[0]); /* magic header */ + put_byte(GZIP_MAGIC[1]); + put_byte(DEFLATED); /* compression method */ + + put_byte(flags); /* general flags */ + put_long(time_stamp); + + /* Write deflated file to zip file */ + crc = updcrc(0, 0); + + bi_init(out); + ct_init(&attr, &method); + lm_init(&deflate_flags); + + put_byte((uch)deflate_flags); /* extra flags */ + put_byte(OS_CODE); /* OS identifier */ + + header_bytes = (long)outcnt; + + (void)deflate(); + + /* Write the crc and uncompressed size */ + put_long(crc); + put_long(isize); + header_bytes += 2*sizeof(long); + + flush_outbuf(); + return OK; +} + + +/* =========================================================================== + * Read a new buffer from the current input file, perform end-of-line + * translation, and update the crc and input file size. + * IN assertion: size >= 2 (for end-of-line translation) + */ +int file_read(buf, size) + char *buf; + unsigned size; +{ + unsigned len; + + Assert(insize == 0, "inbuf not empty"); + + len = read(ifd, buf, size); + if (len == (unsigned)(-1) || len == 0) return (int)len; + + crc = updcrc((uch*)buf, len); + isize += (ulg)len; + return (int)len; +} +#endif diff --git a/archival/tar.c b/archival/tar.c new file mode 100644 index 0000000..03da967 --- /dev/null +++ b/archival/tar.c @@ -0,0 +1,1425 @@ +/* + * Copyright (c) 1999 by David I. Bell + * Permission is granted to use, distribute, or modify this source, + * provided that this copyright notice remains intact. + * + * The "tar" command, taken from sash. + * This allows creation, extraction, and listing of tar files. + * + * Permission to distribute this code under the GPL has been granted. + * Modified for busybox by Erik Andersen <andersee@debian.org> <andersen@lineo.com> + */ + + +#include "internal.h" + +#ifdef BB_TAR + +const char tar_usage[] = +"Create, extract, or list files from a TAR file\n\n" +"usage: tar -[cxtvOf] [tarFileName] [FILE] ...\n" +"\tc=create, x=extract, t=list contents, v=verbose,\n" +"\tO=extract to stdout, f=tarfile or \"-\" for stdin\n"; + + + +#include <stdio.h> +#include <dirent.h> +#include <errno.h> +#include <fcntl.h> +#include <signal.h> +#include <time.h> + +/* + * Tar file constants. + */ +#define TAR_BLOCK_SIZE 512 +#define TAR_NAME_SIZE 100 + + +/* + * The POSIX (and basic GNU) tar header format. + * This structure is always embedded in a TAR_BLOCK_SIZE sized block + * with zero padding. We only process this information minimally. + */ +typedef struct +{ + char name[TAR_NAME_SIZE]; + char mode[8]; + char uid[8]; + char gid[8]; + char size[12]; + char mtime[12]; + char checkSum[8]; + char typeFlag; + char linkName[TAR_NAME_SIZE]; + char magic[6]; + char version[2]; + char uname[32]; + char gname[32]; + char devMajor[8]; + char devMinor[8]; + char prefix[155]; +} TarHeader; + +#define TAR_MAGIC "ustar" +#define TAR_VERSION "00" + +#define TAR_TYPE_REGULAR '0' +#define TAR_TYPE_HARD_LINK '1' +#define TAR_TYPE_SOFT_LINK '2' + + +/* + * Static data. + */ +static BOOL listFlag; +static BOOL extractFlag; +static BOOL createFlag; +static BOOL verboseFlag; +static BOOL tostdoutFlag; + +static BOOL inHeader; +static BOOL badHeader; +static BOOL errorFlag; +static BOOL skipFileFlag; +static BOOL warnedRoot; +static BOOL eofFlag; +static long dataCc; +static int outFd; +static char outName[TAR_NAME_SIZE]; + + +/* + * Static data associated with the tar file. + */ +static const char * tarName; +static int tarFd; +static dev_t tarDev; +static ino_t tarInode; + + +/* + * Local procedures to restore files from a tar file. + */ +static void readTarFile(int fileCount, char ** fileTable); +static void readData(const char * cp, int count); +static void createPath(const char * name, int mode); +static long getOctal(const char * cp, int len); + +static void readHeader(const TarHeader * hp, + int fileCount, char ** fileTable); + + +/* + * Local procedures to save files into a tar file. + */ +static void saveFile(const char * fileName, BOOL seeLinks); + +static void saveRegularFile(const char * fileName, + const struct stat * statbuf); + +static void saveDirectory(const char * fileName, + const struct stat * statbuf); + +static BOOL wantFileName(const char * fileName, + int fileCount, char ** fileTable); + +static void writeHeader(const char * fileName, + const struct stat * statbuf); + +static void writeTarFile(int fileCount, char ** fileTable); +static void writeTarBlock(const char * buf, int len); +static BOOL putOctal(char * cp, int len, long value); +extern const char * modeString(int mode); +extern const char * timeString(time_t timeVal); +extern int fullWrite(int fd, const char * buf, int len); +extern int fullRead(int fd, char * buf, int len); + + +extern int +tar_main(struct FileInfo *unused, int argc, char ** argv) +{ + const char * options; + + argc--; + argv++; + + if (argc < 1) + { + fprintf(stderr, "%s", tar_usage); + return 1; + } + + + errorFlag = FALSE; + extractFlag = FALSE; + createFlag = FALSE; + listFlag = FALSE; + verboseFlag = FALSE; + tostdoutFlag = FALSE; + tarName = NULL; + tarDev = 0; + tarInode = 0; + tarFd = -1; + + /* + * Parse the options. + */ + options = *argv++; + argc--; + + if (**argv == '-') { + for (; *options; options++) + { + switch (*options) + { + case 'f': + if (tarName != NULL) + { + fprintf(stderr, "Only one 'f' option allowed\n"); + + return 1; + } + + tarName = *argv++; + argc--; + + break; + + case 't': + listFlag = TRUE; + break; + + case 'x': + extractFlag = TRUE; + break; + + case 'c': + createFlag = TRUE; + break; + + case 'v': + verboseFlag = TRUE; + break; + + case 'O': + tostdoutFlag = TRUE; + break; + + case '-': + break; + + default: + fprintf(stderr, "Unknown tar flag '%c'\n", *options); + + return 1; + } + } + } + + /* + * Validate the options. + */ + if (extractFlag + listFlag + createFlag != 1) + { + fprintf(stderr, "Exactly one of 'c', 'x' or 't' must be specified\n"); + + return 1; + } + + /* + * Do the correct type of action supplying the rest of the + * command line arguments as the list of files to process. + */ + if (createFlag) + writeTarFile(argc, argv); + else + readTarFile(argc, argv); + if (errorFlag) + fprintf(stderr, "\n"); + return( errorFlag); +} + + +/* + * Read a tar file and extract or list the specified files within it. + * If the list is empty than all files are extracted or listed. + */ +static void +readTarFile(int fileCount, char ** fileTable) +{ + const char * cp; + int cc; + int inCc; + int blockSize; + char buf[BUF_SIZE]; + + skipFileFlag = FALSE; + badHeader = FALSE; + warnedRoot = FALSE; + eofFlag = FALSE; + inHeader = TRUE; + inCc = 0; + dataCc = 0; + outFd = -1; + blockSize = sizeof(buf); + cp = buf; + + /* + * Open the tar file for reading. + */ + if ( (tarName==NULL) || !strcmp( tarName, "-") ) { + tarFd = STDIN; + } + else + tarFd = open(tarName, O_RDONLY); + + if (tarFd < 0) + { + perror(tarName); + errorFlag = TRUE; + return; + } + + /* + * Read blocks from the file until an end of file header block + * has been seen. (A real end of file from a read is an error.) + */ + while (!eofFlag) + { + /* + * Read the next block of data if necessary. + * This will be a large block if possible, which we will + * then process in the small tar blocks. + */ + if (inCc <= 0) + { + cp = buf; + inCc = fullRead(tarFd, buf, blockSize); + + if (inCc < 0) + { + perror(tarName); + errorFlag=TRUE; + goto done; + } + + if (inCc == 0) + { + fprintf(stderr, + "Unexpected end of file from \"%s\"", + tarName); + errorFlag=TRUE; + goto done; + } + } + + /* + * If we are expecting a header block then examine it. + */ + if (inHeader) + { + readHeader((const TarHeader *) cp, fileCount, fileTable); + + cp += TAR_BLOCK_SIZE; + inCc -= TAR_BLOCK_SIZE; + + continue; + } + + /* + * We are currently handling the data for a file. + * Process the minimum of the amount of data we have available + * and the amount left to be processed for the file. + */ + cc = inCc; + + if (cc > dataCc) + cc = dataCc; + + readData(cp, cc); + + /* + * If the amount left isn't an exact multiple of the tar block + * size then round it up to the next block boundary since there + * is padding at the end of the file. + */ + if (cc % TAR_BLOCK_SIZE) + cc += TAR_BLOCK_SIZE - (cc % TAR_BLOCK_SIZE); + + cp += cc; + inCc -= cc; + } + +done: + /* + * Close the tar file if needed. + */ + if ((tarFd >= 0) && (close(tarFd) < 0)) + perror(tarName); + + /* + * Close the output file if needed. + * This is only done here on a previous error and so no + * message is required on errors. + */ + if (tostdoutFlag==FALSE) { + if (outFd >= 0) + (void) close(outFd); + } +} + + +/* + * Examine the header block that was just read. + * This can specify the information for another file, or it can mark + * the end of the tar file. + */ +static void +readHeader(const TarHeader * hp, int fileCount, char ** fileTable) +{ + int mode; + int uid; + int gid; + int checkSum; + long size; + time_t mtime; + const char * name; + int cc; + BOOL hardLink; + BOOL softLink; + + /* + * If the block is completely empty, then this is the end of the + * archive file. If the name is null, then just skip this header. + */ + name = hp->name; + + if (*name == '\0') + { + for (cc = TAR_BLOCK_SIZE; cc > 0; cc--) + { + if (*name++) + return; + } + + eofFlag = TRUE; + + return; + } + + /* + * There is another file in the archive to examine. + * Extract the encoded information and check it. + */ + mode = getOctal(hp->mode, sizeof(hp->mode)); + uid = getOctal(hp->uid, sizeof(hp->uid)); + gid = getOctal(hp->gid, sizeof(hp->gid)); + size = getOctal(hp->size, sizeof(hp->size)); + mtime = getOctal(hp->mtime, sizeof(hp->mtime)); + checkSum = getOctal(hp->checkSum, sizeof(hp->checkSum)); + + if ((mode < 0) || (uid < 0) || (gid < 0) || (size < 0)) + { + if (!badHeader) + fprintf(stderr, "Bad tar header, skipping\n"); + + badHeader = TRUE; + + return; + } + + badHeader = FALSE; + skipFileFlag = FALSE; + + /* + * Check for the file modes. + */ + hardLink = ((hp->typeFlag == TAR_TYPE_HARD_LINK) || + (hp->typeFlag == TAR_TYPE_HARD_LINK - '0')); + + softLink = ((hp->typeFlag == TAR_TYPE_SOFT_LINK) || + (hp->typeFlag == TAR_TYPE_SOFT_LINK - '0')); + + /* + * Check for a directory or a regular file. + */ + if (name[strlen(name) - 1] == '/') + mode |= S_IFDIR; + else if ((mode & S_IFMT) == 0) + mode |= S_IFREG; + + /* + * Check for absolute paths in the file. + * If we find any, then warn the user and make them relative. + */ + if (*name == '/') + { + while (*name == '/') + name++; + + if (!warnedRoot) + { + fprintf(stderr, + "Absolute path detected, removing leading slashes\n"); + } + + warnedRoot = TRUE; + } + + /* + * See if we want this file to be restored. + * If not, then set up to skip it. + */ + if (!wantFileName(name, fileCount, fileTable)) + { + if (!hardLink && !softLink && S_ISREG(mode)) + { + inHeader = (size == 0); + dataCc = size; + } + + skipFileFlag = TRUE; + + return; + } + + /* + * This file is to be handled. + * If we aren't extracting then just list information about the file. + */ + if (!extractFlag) + { + if (verboseFlag) + { + printf("%s %3d/%-d %9ld %s %s", modeString(mode), + uid, gid, size, timeString(mtime), name); + } + else + printf("%s", name); + + if (hardLink) + printf(" (link to \"%s\")", hp->linkName); + else if (softLink) + printf(" (symlink to \"%s\")", hp->linkName); + else if (S_ISREG(mode)) + { + inHeader = (size == 0); + dataCc = size; + } + + printf("\n"); + + return; + } + + /* + * We really want to extract the file. + */ + if (verboseFlag) + printf("x %s\n", name); + + if (hardLink) + { + if (link(hp->linkName, name) < 0) + perror(name); + + return; + } + + if (softLink) + { +#ifdef S_ISLNK + if (symlink(hp->linkName, name) < 0) + perror(name); +#else + fprintf(stderr, "Cannot create symbolic links\n"); +#endif + return; + } + + /* + * If the file is a directory, then just create the path. + */ + if (S_ISDIR(mode)) + { + createPath(name, mode); + + return; + } + + /* + * There is a file to write. + * First create the path to it if necessary with a default permission. + */ + createPath(name, 0777); + + inHeader = (size == 0); + dataCc = size; + + /* + * Start the output file. + */ + if (tostdoutFlag==TRUE) + outFd = STDOUT; + else + outFd = open(name, O_WRONLY | O_CREAT | O_TRUNC, mode); + + if (outFd < 0) + { + perror(name); + skipFileFlag = TRUE; + return; + } + + /* + * If the file is empty, then that's all we need to do. + */ + if (size == 0 && tostdoutFlag == FALSE) + { + (void) close(outFd); + outFd = -1; + } +} + + +/* + * Handle a data block of some specified size that was read. + */ +static void +readData(const char * cp, int count) +{ + /* + * Reduce the amount of data left in this file. + * If there is no more data left, then we need to read + * the header again. + */ + dataCc -= count; + + if (dataCc <= 0) + inHeader = TRUE; + + /* + * If we aren't extracting files or this file is being + * skipped then do nothing more. + */ + if (!extractFlag || skipFileFlag) + return; + + /* + * Write the data to the output file. + */ + if (fullWrite(outFd, cp, count) < 0) + { + perror(outName); + if (tostdoutFlag==FALSE) { + (void) close(outFd); + outFd = -1; + } + skipFileFlag = TRUE; + return; + } + + /* + * If the write failed, close the file and disable further + * writes to this file. + */ + if (dataCc <= 0 && tostdoutFlag==FALSE) + { + if (close(outFd)) + perror(outName); + + outFd = -1; + } +} + + +/* + * Write a tar file containing the specified files. + */ +static void +writeTarFile(int fileCount, char ** fileTable) +{ + struct stat statbuf; + + /* + * Make sure there is at least one file specified. + */ + if (fileCount <= 0) + { + fprintf(stderr, "No files specified to be saved\n"); + errorFlag=TRUE; + } + + /* + * Create the tar file for writing. + */ + if ( (tarName==NULL) || !strcmp( tarName, "-") ) { + tostdoutFlag = TRUE; + tarFd = STDOUT; + } + else + tarFd = open(tarName, O_WRONLY | O_CREAT | O_TRUNC, 0666); + + if (tarFd < 0) + { + perror(tarName); + errorFlag=TRUE; + return; + } + + /* + * Get the device and inode of the tar file for checking later. + */ + if (fstat(tarFd, &statbuf) < 0) + { + perror(tarName); + errorFlag = TRUE; + goto done; + } + + tarDev = statbuf.st_dev; + tarInode = statbuf.st_ino; + + /* + * Append each file name into the archive file. + * Follow symbolic links for these top level file names. + */ + while (!errorFlag && (fileCount-- > 0)) + { + saveFile(*fileTable++, FALSE); + } + + /* + * Now write an empty block of zeroes to end the archive. + */ + writeTarBlock("", 1); + + +done: + /* + * Close the tar file and check for errors if it was opened. + */ + if ( (tostdoutFlag==FALSE) && (tarFd >= 0) && (close(tarFd) < 0)) + perror(tarName); +} + + +/* + * Save one file into the tar file. + * If the file is a directory, then this will recursively save all of + * the files and directories within the directory. The seeLinks + * flag indicates whether or not we want to see symbolic links as + * they really are, instead of blindly following them. + */ +static void +saveFile(const char * fileName, BOOL seeLinks) +{ + int status; + int mode; + struct stat statbuf; + + if (verboseFlag) + printf("a %s\n", fileName); + + /* + * Check that the file name will fit in the header. + */ + if (strlen(fileName) >= TAR_NAME_SIZE) + { + fprintf(stderr, "%s: File name is too long\n", fileName); + + return; + } + + /* + * Find out about the file. + */ +#ifdef S_ISLNK + if (seeLinks) + status = lstat(fileName, &statbuf); + else +#endif + status = stat(fileName, &statbuf); + + if (status < 0) + { + perror(fileName); + + return; + } + + /* + * Make sure we aren't trying to save our file into itself. + */ + if ((statbuf.st_dev == tarDev) && (statbuf.st_ino == tarInode)) + { + fprintf(stderr, "Skipping saving of archive file itself\n"); + + return; + } + + /* + * Check the type of file. + */ + mode = statbuf.st_mode; + + if (S_ISDIR(mode)) + { + saveDirectory(fileName, &statbuf); + + return; + } + + if (S_ISREG(mode)) + { + saveRegularFile(fileName, &statbuf); + + return; + } + + /* + * The file is a strange type of file, ignore it. + */ + fprintf(stderr, "%s: not a directory or regular file\n", fileName); +} + + +/* + * Save a regular file to the tar file. + */ +static void +saveRegularFile(const char * fileName, const struct stat * statbuf) +{ + BOOL sawEof; + int fileFd; + int cc; + int dataCount; + long fullDataCount; + char data[TAR_BLOCK_SIZE * 16]; + + /* + * Open the file for reading. + */ + fileFd = open(fileName, O_RDONLY); + + if (fileFd < 0) + { + perror(fileName); + + return; + } + + /* + * Write out the header for the file. + */ + writeHeader(fileName, statbuf); + + /* + * Write the data blocks of the file. + * We must be careful to write the amount of data that the stat + * buffer indicated, even if the file has changed size. Otherwise + * the tar file will be incorrect. + */ + fullDataCount = statbuf->st_size; + sawEof = FALSE; + + while (fullDataCount > 0) + { + /* + * Get the amount to write this iteration which is + * the minumum of the amount left to write and the + * buffer size. + */ + dataCount = sizeof(data); + + if (dataCount > fullDataCount) + dataCount = (int) fullDataCount; + + /* + * Read the data from the file if we haven't seen the + * end of file yet. + */ + cc = 0; + + if (!sawEof) + { + cc = fullRead(fileFd, data, dataCount); + + if (cc < 0) + { + perror(fileName); + + (void) close(fileFd); + errorFlag = TRUE; + + return; + } + + /* + * If the file ended too soon, complain and set + * a flag so we will zero fill the rest of it. + */ + if (cc < dataCount) + { + fprintf(stderr, + "%s: Short read - zero filling", + fileName); + + sawEof = TRUE; + } + } + + /* + * Zero fill the rest of the data if necessary. + */ + if (cc < dataCount) + memset(data + cc, 0, dataCount - cc); + + /* + * Write the buffer to the TAR file. + */ + writeTarBlock(data, dataCount); + + fullDataCount -= dataCount; + } + + /* + * Close the file. + */ + if ( (tostdoutFlag==FALSE) && close(fileFd) < 0) + fprintf(stderr, "%s: close: %s\n", fileName, strerror(errno)); +} + + +/* + * Save a directory and all of its files to the tar file. + */ +static void +saveDirectory(const char * dirName, const struct stat * statbuf) +{ + DIR * dir; + struct dirent * entry; + BOOL needSlash; + char fullName[PATH_LEN]; + + /* + * Construct the directory name as used in the tar file by appending + * a slash character to it. + */ + strcpy(fullName, dirName); + strcat(fullName, "/"); + + /* + * Write out the header for the directory entry. + */ + writeHeader(fullName, statbuf); + + /* + * Open the directory. + */ + dir = opendir(dirName); + + if (dir == NULL) + { + fprintf(stderr, "Cannot read directory \"%s\": %s\n", + dirName, strerror(errno)); + + return; + } + + /* + * See if a slash is needed. + */ + needSlash = (*dirName && (dirName[strlen(dirName) - 1] != '/')); + + /* + * Read all of the directory entries and check them, + * except for the current and parent directory entries. + */ + while (!errorFlag && ((entry = readdir(dir)) != NULL)) + { + if ((strcmp(entry->d_name, ".") == 0) || + (strcmp(entry->d_name, "..") == 0)) + { + continue; + } + + /* + * Build the full path name to the file. + */ + strcpy(fullName, dirName); + + if (needSlash) + strcat(fullName, "/"); + + strcat(fullName, entry->d_name); + + /* + * Write this file to the tar file, noticing whether or not + * the file is a symbolic link. + */ + saveFile(fullName, TRUE); + } + + /* + * All done, close the directory. + */ + closedir(dir); +} + + +/* + * Write a tar header for the specified file name and status. + * It is assumed that the file name fits. + */ +static void +writeHeader(const char * fileName, const struct stat * statbuf) +{ + long checkSum; + const unsigned char * cp; + int len; + TarHeader header; + + /* + * Zero the header block in preparation for filling it in. + */ + memset((char *) &header, 0, sizeof(header)); + + /* + * Fill in the header. + */ + strcpy(header.name, fileName); + + strncpy(header.magic, TAR_MAGIC, sizeof(header.magic)); + strncpy(header.version, TAR_VERSION, sizeof(header.version)); + + putOctal(header.mode, sizeof(header.mode), statbuf->st_mode & 0777); + putOctal(header.uid, sizeof(header.uid), statbuf->st_uid); + putOctal(header.gid, sizeof(header.gid), statbuf->st_gid); + putOctal(header.size, sizeof(header.size), statbuf->st_size); + putOctal(header.mtime, sizeof(header.mtime), statbuf->st_mtime); + + header.typeFlag = TAR_TYPE_REGULAR; + + /* + * Calculate and store the checksum. + * This is the sum of all of the bytes of the header, + * with the checksum field itself treated as blanks. + */ + memset(header.checkSum, ' ', sizeof(header.checkSum)); + + cp = (const unsigned char *) &header; + len = sizeof(header); + checkSum = 0; + + while (len-- > 0) + checkSum += *cp++; + + putOctal(header.checkSum, sizeof(header.checkSum), checkSum); + + /* + * Write the tar header. + */ + writeTarBlock((const char *) &header, sizeof(header)); +} + + +/* + * Write data to one or more blocks of the tar file. + * The data is always padded out to a multiple of TAR_BLOCK_SIZE. + * The errorFlag static variable is set on an error. + */ +static void +writeTarBlock(const char * buf, int len) +{ + int partialLength; + int completeLength; + char fullBlock[TAR_BLOCK_SIZE]; + + /* + * If we had a write error before, then do nothing more. + */ + if (errorFlag) + return; + + /* + * Get the amount of complete and partial blocks. + */ + partialLength = len % TAR_BLOCK_SIZE; + completeLength = len - partialLength; + + /* + * Write all of the complete blocks. + */ + if ((completeLength > 0) && !fullWrite(tarFd, buf, completeLength)) + { + perror(tarName); + + errorFlag = TRUE; + + return; + } + + /* + * If there are no partial blocks left, we are done. + */ + if (partialLength == 0) + return; + + /* + * Copy the partial data into a complete block, and pad the rest + * of it with zeroes. + */ + memcpy(fullBlock, buf + completeLength, partialLength); + memset(fullBlock + partialLength, 0, TAR_BLOCK_SIZE - partialLength); + + /* + * Write the last complete block. + */ + if (!fullWrite(tarFd, fullBlock, TAR_BLOCK_SIZE)) + { + perror(tarName); + + errorFlag = TRUE; + } +} + + +/* + * Attempt to create the directories along the specified path, except for + * the final component. The mode is given for the final directory only, + * while all previous ones get default protections. Errors are not reported + * here, as failures to restore files can be reported later. + */ +static void +createPath(const char * name, int mode) +{ + char * cp; + char * cpOld; + char buf[TAR_NAME_SIZE]; + + strcpy(buf, name); + + cp = strchr(buf, '/'); + + while (cp) + { + cpOld = cp; + cp = strchr(cp + 1, '/'); + + *cpOld = '\0'; + + if (mkdir(buf, cp ? 0777 : mode) == 0) + printf("Directory \"%s\" created\n", buf); + + *cpOld = '/'; + } +} + + +/* + * Read an octal value in a field of the specified width, with optional + * spaces on both sides of the number and with an optional null character + * at the end. Returns -1 on an illegal format. + */ +static long +getOctal(const char * cp, int len) +{ + long val; + + while ((len > 0) && (*cp == ' ')) + { + cp++; + len--; + } + + if ((len == 0) || !isOctal(*cp)) + return -1; + + val = 0; + + while ((len > 0) && isOctal(*cp)) + { + val = val * 8 + *cp++ - '0'; + len--; + } + + while ((len > 0) && (*cp == ' ')) + { + cp++; + len--; + } + + if ((len > 0) && *cp) + return -1; + + return val; +} + + +/* + * Put an octal string into the specified buffer. + * The number is zero and space padded and possibly null padded. + * Returns TRUE if successful. + */ +static BOOL +putOctal(char * cp, int len, long value) +{ + int tempLength; + char * tempString; + char tempBuffer[32]; + + /* + * Create a string of the specified length with an initial space, + * leading zeroes and the octal number, and a trailing null. + */ + tempString = tempBuffer; + + sprintf(tempString, " %0*lo", len - 2, value); + + tempLength = strlen(tempString) + 1; + + /* + * If the string is too large, suppress the leading space. + */ + if (tempLength > len) + { + tempLength--; + tempString++; + } + + /* + * If the string is still too large, suppress the trailing null. + */ + if (tempLength > len) + tempLength--; + + /* + * If the string is still too large, fail. + */ + if (tempLength > len) + return FALSE; + + /* + * Copy the string to the field. + */ + memcpy(cp, tempString, len); + + return TRUE; +} + + +/* + * See if the specified file name belongs to one of the specified list + * of path prefixes. An empty list implies that all files are wanted. + * Returns TRUE if the file is selected. + */ +static BOOL +wantFileName(const char * fileName, int fileCount, char ** fileTable) +{ + const char * pathName; + int fileLength; + int pathLength; + + /* + * If there are no files in the list, then the file is wanted. + */ + if (fileCount == 0) + return TRUE; + + fileLength = strlen(fileName); + + /* + * Check each of the test paths. + */ + while (fileCount-- > 0) + { + pathName = *fileTable++; + + pathLength = strlen(pathName); + + if (fileLength < pathLength) + continue; + + if (memcmp(fileName, pathName, pathLength) != 0) + continue; + + if ((fileLength == pathLength) || + (fileName[pathLength] == '/')) + { + return TRUE; + } + } + + return FALSE; +} + + + +/* + * Return the standard ls-like mode string from a file mode. + * This is static and so is overwritten on each call. + */ +const char * +modeString(int mode) +{ + static char buf[12]; + + strcpy(buf, "----------"); + + /* + * Fill in the file type. + */ + if (S_ISDIR(mode)) + buf[0] = 'd'; + if (S_ISCHR(mode)) + buf[0] = 'c'; + if (S_ISBLK(mode)) + buf[0] = 'b'; + if (S_ISFIFO(mode)) + buf[0] = 'p'; +#ifdef S_ISLNK + if (S_ISLNK(mode)) + buf[0] = 'l'; +#endif +#ifdef S_ISSOCK + if (S_ISSOCK(mode)) + buf[0] = 's'; +#endif + + /* + * Now fill in the normal file permissions. + */ + if (mode & S_IRUSR) + buf[1] = 'r'; + if (mode & S_IWUSR) + buf[2] = 'w'; + if (mode & S_IXUSR) + buf[3] = 'x'; + if (mode & S_IRGRP) + buf[4] = 'r'; + if (mode & S_IWGRP) + buf[5] = 'w'; + if (mode & S_IXGRP) + buf[6] = 'x'; + if (mode & S_IROTH) + buf[7] = 'r'; + if (mode & S_IWOTH) + buf[8] = 'w'; + if (mode & S_IXOTH) + buf[9] = 'x'; + + /* + * Finally fill in magic stuff like suid and sticky text. + */ + if (mode & S_ISUID) + buf[3] = ((mode & S_IXUSR) ? 's' : 'S'); + if (mode & S_ISGID) + buf[6] = ((mode & S_IXGRP) ? 's' : 'S'); + if (mode & S_ISVTX) + buf[9] = ((mode & S_IXOTH) ? 't' : 'T'); + + return buf; +} + + +/* + * Get the time string to be used for a file. + * This is down to the minute for new files, but only the date for old files. + * The string is returned from a static buffer, and so is overwritten for + * each call. + */ +const char * +timeString(time_t timeVal) +{ + time_t now; + char * str; + static char buf[26]; + + time(&now); + + str = ctime(&timeVal); + + strcpy(buf, &str[4]); + buf[12] = '\0'; + + if ((timeVal > now) || (timeVal < now - 365*24*60*60L)) + { + strcpy(&buf[7], &str[20]); + buf[11] = '\0'; + } + + return buf; +} + + + +/* + * Write all of the supplied buffer out to a file. + * This does multiple writes as necessary. + * Returns the amount written, or -1 on an error. + */ +int +fullWrite(int fd, const char * buf, int len) +{ + int cc; + int total; + + total = 0; + + while (len > 0) + { + cc = write(fd, buf, len); + + if (cc < 0) + return -1; + + buf += cc; + total+= cc; + len -= cc; + } + + return total; +} + + +/* + * Read all of the supplied buffer from a file. + * This does multiple reads as necessary. + * Returns the amount read, or -1 on an error. + * A short read is returned on an end of file. + */ +int +fullRead(int fd, char * buf, int len) +{ + int cc; + int total; + + total = 0; + + while (len > 0) + { + cc = read(fd, buf, len); + + if (cc < 0) + return -1; + + if (cc == 0) + break; + + buf += cc; + total+= cc; + len -= cc; + } + + return total; +} + + + +#endif +/* END CODE */ + + |