Be sure to run "make proto" before "make".
---- orig/generator.c 2005-01-17 23:11:45
-+++ generator.c 2005-01-18 10:55:29
-@@ -44,6 +44,7 @@ extern int size_only;
+--- orig/generator.c 2005-02-03 02:04:20
++++ generator.c 2005-02-03 02:11:10
+@@ -47,6 +47,7 @@ extern int size_only;
extern OFF_T max_size;
extern int io_timeout;
extern int protocol_version;
extern int always_checksum;
extern char *partial_dir;
extern char *basis_dir[];
-@@ -242,6 +243,81 @@ static void generate_and_send_sums(int f
+@@ -227,6 +228,88 @@ static void generate_and_send_sums(int f
+ unmap_file(mapbuf);
}
-
++/* Try to find a filename in the same dir as "fname" with a similar name.
++ *
++ * TODO:
++ * - We should be using a cache of names for the current dir, not
++ * re-reading the destination directory for every file.
++ * - We must not return an rsync tempfile from the current transfer.
++ * - If the highest-rated name is not a normal file, we should fall-
++ * back to the next highest-rated file.
++ * - We must not return a destination file that is being updated
++ * during the current transfer, even if we already processed it
++ * (since the receiver may not be done with it yet).
++ * - We must weed out any names that a daemon's config has excluded.
++ */
+static int find_fuzzy(const char *fname, char *buf, STRUCT_STAT *st_ptr)
+{
+ DIR *d;
+ struct dirent *di;
+ char *basename, *dirname, *slash;
+ char bestname[MAXPATHLEN];
-+ int extlen, basename_len;
++ int suf_len, basename_len;
+ uint32 lowest_dist = 0x7FFFFFFF;
-+ const char *ext;
++ const char *suf;
+
+ strlcpy(buf, fname, MAXPATHLEN);
+ if ((slash = strrchr(buf, '/')) != NULL) {
+ if (slash)
+ *slash = '/';
+
-+ /* Get final extension, eg. .gz; never full basename though. */
-+ for (ext = basename; *ext == '.'; ext++) {}
-+ if (!(ext = strrchr(ext, '.')))
-+ ext = basename + basename_len; /* ext = "" */
-+ extlen = strlen(ext);
++ suf = find_filename_suffix(basename, basename_len, &suf_len);
+
+ bestname[0] = '\0';
+ while ((di = readdir(d)) != NULL) {
-+ const char *dname = d_name(di);
++ const char *dname_suf, *dname = d_name(di);
+ uint32 dist;
-+ int dname_len;
++ int dname_len, dname_suf_len;
+
+ if (dname[0] == '.' && (dname[1] == '\0'
+ || (dname[1] == '.' && dname[2] == '\0')))
+ continue;
+
+ dname_len = strlen(dname);
-+
-+ /* Extensions must match */
-+ if (dname_len <= extlen
-+ || strcmp(dname + dname_len - extlen, ext) != 0)
-+ continue;
++ dname_suf = find_filename_suffix(dname, dname_len, &dname_suf_len);
+
+ dist = fuzzy_distance(dname, dname_len, basename, basename_len);
-+ if (verbose > 1) {
-+ rprintf(FINFO, "fuzzy distance for %s = %lx\n",
-+ dname, (unsigned long)dist);
++ /* Add some extra weight to how well the suffixes match. */
++ dist += fuzzy_distance(dname_suf, dname_suf_len, suf, suf_len) * 10;
++ if (verbose > 4) {
++ rprintf(FINFO, "fuzzy distance for %s = %d (%d)\n",
++ dname, (int)(dist>>16), (int)(dist&0xFFFF));
+ }
-+ if (dist < lowest_dist) {
++ if (dist <= lowest_dist) {
+ strlcpy(bestname, dname, sizeof bestname);
+ lowest_dist = dist;
+ }
+}
+
- /*
- * Acts on file number @p i from @p flist, whose name is @p fname.
-@@ -496,6 +572,15 @@ static void recv_generator(char *fname,
+ /* Acts on flist->file's ndx'th item, whose name is fname. If a directory,
+ * make sure it exists, and has the right permissions/timestamp info. For
+@@ -477,6 +560,15 @@ static void recv_generator(char *fname,
} else
partialptr = NULL;
if (statret == -1) {
if (preserve_hard_links && hard_link_check(file, HL_SKIP))
return;
-@@ -524,6 +609,8 @@ static void recv_generator(char *fname,
+@@ -505,6 +597,8 @@ static void recv_generator(char *fname,
if (!compare_dest && fnamecmp_type <= FNAMECMP_BASIS_DIR_HIGH)
;
else if (unchanged_file(fnamecmp, file, &st)) {
if (fnamecmp_type == FNAMECMP_FNAME)
set_perms(fname, file, &st, PERMS_REPORT);
-@@ -598,8 +685,24 @@ notify_others:
- write_int(f_out, i);
+@@ -579,8 +673,24 @@ notify_others:
+ write_int(f_out, ndx);
if (protocol_version >= 29 && inplace && !read_batch)
write_byte(f_out, fnamecmp_type);
- if (f_out_name >= 0)
if (dry_run || read_batch)
return;
---- orig/main.c 2005-01-17 23:11:45
+--- orig/main.c 2005-01-30 10:07:21
+++ main.c 2005-01-14 18:33:15
@@ -48,6 +48,7 @@ extern int keep_dirlinks;
extern int preserve_hard_links;
extern int relative_paths;
extern int rsync_port;
extern int whole_file;
-@@ -464,7 +465,8 @@ static int do_recv(int f_in,int f_out,st
+@@ -491,7 +492,8 @@ static int do_recv(int f_in,int f_out,st
int pid;
int status = 0;
int error_pipe[2], name_pipe[2];
/* The receiving side mustn't obey this, or an existing symlink that
* points to an identical file won't be replaced by the referent. */
---- orig/options.c 2005-01-17 23:11:45
-+++ options.c 2005-01-15 21:08:13
-@@ -86,6 +86,7 @@ int copy_unsafe_links = 0;
+--- orig/options.c 2005-02-01 10:39:22
++++ options.c 2005-01-28 19:31:20
+@@ -90,6 +90,7 @@ int copy_unsafe_links = 0;
int size_only = 0;
int daemon_bwlimit = 0;
int bwlimit = 0;
+int fuzzy_basis = 0;
size_t bwlimit_writemax = 0;
- int delete_after = 0;
int only_existing = 0;
-@@ -288,6 +289,7 @@ void usage(enum logcode F)
+ int opt_ignore_existing = 0;
+@@ -303,6 +304,7 @@ void usage(enum logcode F)
rprintf(F," --compare-dest=DIR also compare destination files relative to DIR\n");
rprintf(F," --copy-dest=DIR ... and include copies of unchanged files\n");
rprintf(F," --link-dest=DIR hardlink to files in DIR when unchanged\n");
+ rprintf(F," --fuzzy find similar file for basis when no dest file\n");
- rprintf(F," -P equivalent to --partial --progress\n");
rprintf(F," -z, --compress compress file data\n");
- rprintf(F," -C, --cvs-exclude auto ignore files in the same way CVS does\n");
-@@ -384,6 +386,7 @@ static struct poptOption long_options[]
+ rprintf(F," -C, --cvs-exclude auto-ignore files the same way CVS does\n");
+ rprintf(F," -f, --filter=RULE add a file-filtering RULE\n");
+@@ -408,6 +410,7 @@ static struct poptOption long_options[]
{"compare-dest", 0, POPT_ARG_STRING, 0, OPT_COMPARE_DEST, 0, 0 },
{"copy-dest", 0, POPT_ARG_STRING, 0, OPT_COPY_DEST, 0, 0 },
{"link-dest", 0, POPT_ARG_STRING, 0, OPT_LINK_DEST, 0, 0 },
/* TODO: Should this take an optional int giving the compression level? */
{"compress", 'z', POPT_ARG_NONE, &do_compression, 0, 0, 0 },
{"stats", 0, POPT_ARG_NONE, &do_stats, 0, 0, 0 },
-@@ -1234,6 +1237,9 @@ void server_options(char **args,int *arg
- args[ac++] = "--no-relative";
- }
+@@ -1321,6 +1324,9 @@ void server_options(char **args,int *arg
+ if (!implied_dirs && !am_sender)
+ args[ac++] = "--no-implied-dirs";
+ if (fuzzy_basis && am_sender)
+ args[ac++] = "--fuzzy";
*argc = ac;
return;
---- orig/receiver.c 2005-01-17 23:11:45
+--- orig/receiver.c 2005-02-03 02:04:20
+++ receiver.c 2005-01-15 21:21:02
-@@ -324,6 +324,27 @@ static int receive_data(int f_in, char *
+@@ -256,6 +256,27 @@ static int receive_data(int f_in, char *
}
static void discard_receive_data(int f_in, OFF_T length)
{
receive_data(f_in, NULL, -1, 0, NULL, -1, length);
-@@ -454,6 +475,10 @@ int recv_files(int f_in, struct file_lis
+@@ -395,6 +416,10 @@ int recv_files(int f_in, struct file_lis
case FNAMECMP_BACKUP:
fnamecmp = get_backup_name(fname);
break;
default:
if (j >= basis_dir_cnt) {
rprintf(FERROR,
---- orig/rsync.h 2005-01-17 23:11:45
-+++ rsync.h 2005-01-15 21:24:09
-@@ -128,6 +128,7 @@
+--- orig/rsync.h 2005-02-03 02:04:20
++++ rsync.h 2005-01-19 18:36:47
+@@ -131,6 +131,7 @@
#define FNAMECMP_FNAME 0x80
#define FNAMECMP_PARTIAL_DIR 0x81
#define FNAMECMP_BACKUP 0x82
+#define FNAMECMP_FUZZY 0x83
-
- /* Log-message categories. FLOG is only used on the daemon side to
---- orig/rsync.yo 2005-01-17 23:11:46
-+++ rsync.yo 2005-01-15 21:48:52
-@@ -358,6 +358,7 @@ verb(
+ /* For calling delete_file() */
+ #define DEL_DIR (1<<0)
+--- orig/rsync.yo 2005-02-01 10:39:23
++++ rsync.yo 2005-01-28 19:31:36
+@@ -354,6 +354,7 @@ to the detailed description below for a
--compare-dest=DIR also compare received files relative to DIR
--copy-dest=DIR ... and include copies of unchanged files
--link-dest=DIR hardlink to files in DIR when unchanged
+ --fuzzy find similar file for basis when no dest
- -P equivalent to --partial --progress
-z, --compress compress file data
- -C, --cvs-exclude auto ignore files in the same way CVS does
-@@ -878,6 +879,11 @@ Note that rsync versions prior to 2.6.1
- (or implied by -a). You can work-around this bug by avoiding the -o option
+ -C, --cvs-exclude auto-ignore files in the same way CVS does
+ -f, --filter=RULE add a file-filtering RULE
+@@ -937,6 +938,14 @@ bf(--link-dest) from working properly fo
+ (or implied by bf(-a)). You can work-around this bug by avoiding the bf(-o) option
when sending to an old rsync.
+dit(bf(--fuzzy)) This option tells rsync that it should look around for a
+basis file for any destination file that is missing. The current algorithm
+looks for a similarly-named file in the same directory as the destination
-+file, and, if found, uses that to try to speed up the transfer.
++file, and, if found, uses that to try to speed up the transfer. Note that
++the use of the --delete option might get rid of any potential fuzzy-match
++files, so either use --delete-after or filename exclusions if you need to
++prevent this.
+
dit(bf(-z, --compress)) With this option, rsync compresses any data from
the files that it sends to the destination machine. This
option is useful on slow connections. The compression method used is the
---- orig/util.c 2004-09-07 21:45:30
-+++ util.c 2005-01-18 11:18:46
-@@ -1217,3 +1217,50 @@ void *_realloc_array(void *ptr, unsigned
+--- orig/util.c 2005-01-28 19:08:20
++++ util.c 2005-01-19 17:30:51
+@@ -1213,3 +1213,108 @@ void *_realloc_array(void *ptr, unsigned
return malloc(size * num);
return realloc(ptr, size * num);
}
+
++/* Take a filename and filename length and return the most significant
++ * filename suffix we can find. This ignores suffixes such as "~",
++ * ".bak", ".orig", ".~1~", etc. */
++const char *find_filename_suffix(const char *fn, int fn_len, int *len_ptr)
++{
++ const char *suf, *s;
++ BOOL had_tilde;
++ int s_len;
++
++ /* One or more dots at the start aren't a suffix. */
++ while (fn_len && *fn == '.') fn++, fn_len--;
++
++ /* Ignore the ~ in a "foo~" filename. */
++ if (fn_len > 1 && fn[fn_len-1] == '~')
++ fn_len--, had_tilde = True;
++ else
++ had_tilde = False;
++
++ /* Assume we don't find an suffix. */
++ suf = "";
++ *len_ptr = 0;
++
++ /* Find the last significant suffix. */
++ for (s = fn + fn_len - 1; fn_len > 1; ) {
++ while (*s != '.' && s != fn) s--;
++ if (s == fn)
++ break;
++ s_len = fn_len - (s - fn);
++ fn_len = s - fn;
++ if (s_len == 3) {
++ if (strcmp(s+1, "bak") == 0
++ || strcmp(s+1, "old") == 0)
++ continue;
++ } else if (s_len == 4) {
++ if (strcmp(s+1, "orig") == 0)
++ continue;
++ } else if (s_len > 2 && had_tilde
++ && s[1] == '~' && isdigit(s[2]))
++ continue;
++ *len_ptr = s_len;
++ suf = s;
++ /* Determine if the suffix is all digits. */
++ for (s++, s_len--; s_len > 0; s++, s_len--) {
++ if (!isdigit(*s))
++ return suf;
++ }
++ /* An all-digit suffix may not be that signficant. */
++ continue;
++ }
++
++ return suf;
++}
++
+/* This is an implementation of the Levenshtein distance algorithm. It
+ * was implemented to avoid needing a two-dimensional matrix (to save
+ * memory). It was also tweaked to try to factor in the ASCII distance
+ int32 cost;
+ int i1, i2;
+
-+ if (!len1)
-+ return (int32)len2 * UNIT;
-+ if (!len2)
-+ return (int32)len1 * UNIT;
++ if (!len1 || !len2) {
++ if (!len1) {
++ s1 = s2;
++ len1 = len2;
++ }
++ for (i1 = 0, cost = 0; i1 < len1; i1++)
++ cost += s1[i1];
++ return (int32)len1 * UNIT + cost;
++ }
+
+ for (i2 = 0; i2 < len2; i2++)
+ a[i2] = (i2+1) * UNIT;