Add support for gzip --rsyncable output produced by gzip 1.4.

This differs from prior --rsyncable output, which was made by a debian
specific patch and used a different rsync window size, and reset huffman
encodings less often.

debian/changelog

@@ -1,5 +1,9 @@
 pristine-tar (1.14) UNRELEASED; urgency=low
 
+  * Add support for gzip --rsyncable output produced by gzip 1.4.
+    This differs from prior --rsyncable output, which was made by a debian
+    specific patch and used a different rsync window size, and reset
+    huffman encodings less often.
   * Clarify man page.
 
  -- Joey Hess <joeyh@debian.org>  Thu, 07 Jul 2011 12:17:00 -0400

pristine-gz

@@ -192,6 +192,7 @@ sub reproducegz {
 		# and it was produced by gnu gzip.
 		push @try, ['--gnu', @args];
 		push @try, ['--gnu', @args, '--rsyncable'];
+		push @try, ['--gnu', @args, '--new-rsyncable'];
 	}
 
 	if ($name =~ /\//) {
@@ -285,7 +286,7 @@ sub gengz {
 	my @params=split(' ', $delta->{params});
 	while (@params) {
 		$_=shift @params;
-		next if /^(--gnu|--rsyncable|-[nmM1-9])$/;
+		next if /^(--gnu|--rsyncable|--new-rsyncable|-[nmM1-9])$/;
 		if (/^(--original-name|--quirk|--osflag)$/) {
 			shift @params;
 			next;

pristine-tar.spec

@@ -1,5 +1,5 @@
 Name: pristine-tar
-Version: 1.13
+Version: 1.14
 Release: 2%{?dist}
 Summary: regenerate pristine tarballs
 

zgz/gzip/deflate.c

@@ -101,12 +101,19 @@
 #endif
 /* Matches of length 3 are discarded if their distance exceeds TOO_FAR */
 
-#ifndef RSYNC_WIN
-#  define RSYNC_WIN 4096
-#endif
+int RSYNC_WIN = 4096;
 /* Size of rsync window, must be < MAX_DIST */
 
-#define RSYNC_SUM_MATCH(sum) ((sum) % RSYNC_WIN == 0)
+/* Whether to enable compatability with Debian's old rsyncable patch. */
+int debian_rsyncable = 1;
+
+void disable_debian_rsyncable () {
+	debian_rsyncable = 0;
+	RSYNC_WIN = 8192;
+}
+
+#define RSYNC_SUM_MATCH_DEBIAN(sum) ((sum) % RSYNC_WIN == 0)
+#define RSYNC_SUM_MATCH(sum) (((sum) & (RSYNC_WIN - 1)) == 0)
 /* Whether window sum matches magic value */
 
 /* ===========================================================================
@@ -525,7 +532,9 @@ static void rsync_roll(unsigned start, unsigned num)
 	rsync_sum += (ulg)window[i];
 	/* Old character out */
 	rsync_sum -= (ulg)window[i - RSYNC_WIN];
-	if (rsync_chunk_end == 0xFFFFFFFFUL && RSYNC_SUM_MATCH(rsync_sum))
+	if (debian_rsyncable && rsync_chunk_end == 0xFFFFFFFFUL && RSYNC_SUM_MATCH_DEBIAN(rsync_sum))
+	    rsync_chunk_end = i;
+	if (! debian_rsyncable && rsync_chunk_end == 0xFFFFFFFFUL && RSYNC_SUM_MATCH(rsync_sum))
 	    rsync_chunk_end = i;
     }
 }
@@ -614,10 +623,15 @@ static void deflate_fast(int pack_level, int rsync)
             lookahead--;
 	    strstart++;
         }
-	if (rsync && strstart > rsync_chunk_end) {
+	if (rsync && debian_rsyncable && strstart > rsync_chunk_end) {
 	    rsync_chunk_end = 0xFFFFFFFFUL;
 	    flush = 2;
 	}
+	if (rsync && ! debian_rsyncable && strstart > rsync_chunk_end) {
+	    flush = 1;
+	    ct_init();
+	    rsync_chunk_end = 0xFFFFFFFFUL;
+	}
         if (flush) FLUSH_BLOCK(0), block_start = strstart;
 
         /* Make sure that we always have enough lookahead, except
@@ -636,14 +650,21 @@ static void deflate_fast(int pack_level, int rsync)
  * evaluation for matches: a match is finally adopted only if there is
  * no better match at the next window position.
  */
-void gnu_deflate(int pack_level, int rsync)
+void gnu_deflate(int pack_level, int rsync, int newrsync)
 {
     IPos hash_head;          /* head of hash chain */
     IPos prev_match;         /* previous match */
     int flush = 0;           /* 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 */
 
+    /* The newrsync mode superscedes Debian's old, somewhat broken
+     * rsync patch. */
+    if (newrsync) {
+	    rsync=1;
+	    disable_debian_rsyncable();
+    }
+
     if (pack_level <= 3) {
         deflate_fast(pack_level, rsync); /* optimized for speed */
         return;
@@ -705,10 +726,15 @@ void gnu_deflate(int pack_level, int rsync)
             match_length = MIN_MATCH-1;
             strstart++;
 
-	    if (rsync && strstart > rsync_chunk_end) {
+	    if (rsync && debian_rsyncable && strstart > rsync_chunk_end) {
 		rsync_chunk_end = 0xFFFFFFFFUL;
 		flush = 2;
 	    }
+	    if (rsync && ! debian_rsyncable && strstart > rsync_chunk_end) {
+		ct_init();
+		rsync_chunk_end = 0xFFFFFFFFUL;
+		flush = 1;
+	    }
             if (flush) FLUSH_BLOCK(0), block_start = strstart;
         } else if (match_available) {
             /* If there was no match at the previous position, output a
@@ -717,10 +743,15 @@ void gnu_deflate(int pack_level, int rsync)
              */
             Tracevv((stderr,"%c",window[strstart-1]));
 	    flush = ct_tally (pack_level, 0, window[strstart-1]);
-	    if (rsync && strstart > rsync_chunk_end) {
+	    if (rsync && debian_rsyncable && strstart > rsync_chunk_end) {
 		rsync_chunk_end = 0xFFFFFFFFUL;
 		flush = 2;
 	    }
+	    if (rsync && ! debian_rsyncable && strstart > rsync_chunk_end) {
+		ct_init();
+		rsync_chunk_end = 0xFFFFFFFFUL;
+		flush = 1;
+	    }
             if (flush) FLUSH_BLOCK(0), block_start = strstart;
 	    RSYNC_ROLL(strstart, 1);
             strstart++;
@@ -729,12 +760,18 @@ void gnu_deflate(int pack_level, int rsync)
             /* There is no previous match to compare with, wait for
              * the next step to decide.
              */
-	    if (rsync && strstart > rsync_chunk_end) {
+	    if (rsync && debian_rsyncable && strstart > rsync_chunk_end) {
 		/* Reset huffman tree */
 		rsync_chunk_end = 0xFFFFFFFFUL;
 		flush = 2;
 		FLUSH_BLOCK(0), block_start = strstart;
 	    }
+	    if (rsync && ! debian_rsyncable && strstart > rsync_chunk_end) {
+		ct_init();
+		/* Reset huffman tree */
+		rsync_chunk_end = 0xFFFFFFFFUL;
+		FLUSH_BLOCK(0), block_start = strstart;
+	    }
             match_available = 1;
 	    RSYNC_ROLL(strstart, 1);
             strstart++;

zgz/gzip/gzip.c

@@ -38,7 +38,7 @@ static off_t bytes_in;  /* number of input bytes */
  * Deflate in to out.
  * IN assertions: the input and output buffers are cleared.
  */
-void gnuzip(int in, int out, char *origname, ulg timestamp, int level, int osflag, int rsync)
+void gnuzip(int in, int out, char *origname, ulg timestamp, int level, int osflag, int rsync, int newrsync)
 {
     uch  flags = 0;         /* general purpose bit flags */
     ush  deflate_flags = 0; /* pkzip -es, -en or -ex equivalent */
@@ -76,7 +76,7 @@ void gnuzip(int in, int out, char *origname, ulg timestamp, int level, int osfla
 	} while (*p++);
     }
 
-    gnu_deflate(level, rsync);
+    gnu_deflate(level, rsync, newrsync);
 
     /* Write the crc and uncompressed size */
     put_long(crc);

zgz/gzip/gzip.h

@@ -129,7 +129,7 @@ extern int file_read(char *buf,  unsigned size);
 
         /* in deflate.c */
 void lm_init(int pack_level, ush *flags);
-void gnu_deflate(int pack_level, int rsync);
+void gnu_deflate(int pack_level, int rsync, int newrsync);
 
         /* in trees.c */
 void ct_init(void);

zgz/zgz.c

@@ -72,7 +72,7 @@
 #include <getopt.h>
 #include <time.h>
 
-extern void gnuzip(int in, int out, char *origname, unsigned long timestamp, int level, int osflag, int rsync);
+extern void gnuzip(int in, int out, char *origname, unsigned long timestamp, int level, int osflag, int rsync, int newrsync);
 extern void old_bzip2(int level);
 
 #define BUFLEN		(64 * 1024)
@@ -145,7 +145,6 @@ static const struct option longopts[] = {
 	{ "no-name",		no_argument,		0,	'n' },
 	{ "name",		no_argument,		0,	'N' },
 	{ "quiet",		no_argument,		0,	'q' },
-	{ "recursive",		no_argument,		0,	'r' },
 	{ "fast",		no_argument,		0,	'1' },
 	{ "best",		no_argument,		0,	'9' },
 	{ "ascii",		no_argument,		0,	'a' },
@@ -154,6 +153,7 @@ static const struct option longopts[] = {
 	{ "old-bzip2",          no_argument,            0,      'O' },
 	{ "zlib",               no_argument,            0,      'Z' },
 	{ "rsyncable",          no_argument,            0,      'R' },
+	{ "new-rsyncable",      no_argument,            0,      'r' },
 	{ "no-timestamp",	no_argument,		0,	'm' },
 	{ "force-timestamp",	no_argument,		0,	'M' },
 	{ "timestamp",		required_argument,	0,	'T' },
@@ -185,6 +185,7 @@ main(int argc, char **argv)
 	int level = 6;
 	int osflag = GZIP_OS_UNIX;
 	int rsync = 0;
+	int new_rsync = 0;
 	int ch;
 
 	if (strcmp(progname, "gunzip") == 0 ||
@@ -280,8 +281,7 @@ main(int argc, char **argv)
 			rsync = 1;
 			break;
 		case 'r':
-			fprintf(stderr, "%s: recursive is not supported on this version\n", progname);
-			usage();
+			new_rsync = 1;
 			break;
 		case 'd':
 			fprintf(stderr, "%s: decompression is not supported on this version\n", progname);
@@ -322,15 +322,15 @@ main(int argc, char **argv)
 			fprintf(stderr, "%s: quirks not supported with --gnu\n", progname);
 			return 1;
 		}
-		gnuzip(STDIN_FILENO, STDOUT_FILENO, origname, timestamp, level, osflag, rsync);
+		gnuzip(STDIN_FILENO, STDOUT_FILENO, origname, timestamp, level, osflag, rsync, new_rsync);
 	} else if (bzold) {
 		if (quirks) {
 			fprintf(stderr, "%s: quirks not supported with --old-bzip\n", progname);
 			return 1;
 		}
 		old_bzip2(level);
 	} else {
-		if (rsync) {
+		if (rsync || new_rsync) {
 			fprintf(stderr, "%s: --rsyncable not supported with --zlib\n", progname);
 			return 1;
 		}
@@ -528,6 +528,7 @@ usage(void)
     " -T --timestamp SECONDS   set the timestamp to the specified number of seconds\n"
     " \ngnu-specific options:\n"
     " -R --rsyncable           make rsync-friendly archive\n"
+    " -r --new-rsyncable       make rsync-friendly archive (new version)\n"
     " \nzlib-specific options:\n"
     " -k --quirk QUIRK         enable a format quirk (buggy-bsd, ntfs, perl)\n");
 	exit(0);