Transformed shell script into perl script and improved it to allow

[rsync/rsync-patches.git] / fuzzy.diff

Wayne Davison wrote:

I greatly simplified the changes to generator.c, making the patch
easier to maintain and fixing the failing test in the testsuite.
Very lightly tested.

--- generator.c	30 Jun 2004 07:27:30 -0000	1.93
+++ generator.c	30 Jun 2004 07:43:46 -0000
@@ -41,6 +41,7 @@ extern int ignore_times;
 extern int size_only;
 extern int io_timeout;
 extern int protocol_version;
+extern int fuzzy;
 extern int always_checksum;
 extern char *compare_dest;
 extern int link_dest;
@@ -257,6 +258,94 @@ static void generate_and_send_sums(struc
 }
 
 
+static void split_names(char *fname, char **dirname, char **basename)
+{
+	char *slash = strrchr(fname, '/');
+	if (slash) {
+		*dirname = fname;
+		*slash = '\0';
+		*basename = slash+1;
+	} else {
+		*basename = fname;
+		*dirname = ".";
+	}
+}
+
+
+static unsigned int measure_name(const char *name, const char *basename,
+				 const char *ext)
+{
+	int namelen = strlen(name);
+	int extlen = strlen(ext);
+	unsigned int score = 0;
+
+	/* Extensions must match */
+	if (namelen <= extlen || strcmp(name + namelen - extlen, ext) != 0)
+		return 0;
+
+	/* Now score depends on similarity of prefix */
+	for (; *name == *basename && *name; name++, basename++)
+		score++;
+	return score;
+}
+
+
+static int find_fuzzy(char **fname_ptr, char *buf, STRUCT_STAT *st_ptr)
+{
+	DIR *d;
+	struct dirent *di;
+	char *basename, *dirname;
+	char mangled_name[MAXPATHLEN];
+	char bestname[MAXPATHLEN];
+	unsigned int bestscore = 0;
+	const char *ext;
+
+	strlcpy(mangled_name, *fname_ptr, sizeof mangled_name);
+
+	split_names(mangled_name, &dirname, &basename);
+	if (!(d = opendir(dirname))) {
+		rsyserr(FERROR, errno, "recv_generator opendir(%s)", dirname);
+		return -1;
+	}
+
+	/* Get final extension, eg. .gz; never full basename though. */
+	ext = strrchr(basename + 1, '.');
+	if (!ext)
+		ext = basename + strlen(basename); /* ext = "" */
+
+	while ((di = readdir(d)) != NULL) {
+		const char *dname = d_name(di);
+		unsigned int score;
+
+		if (dname[0] == '.' && (dname[1] == '\0'
+		    || (dname[1] == '.' && dname[2] == '\0')))
+			continue;
+
+		score = measure_name(dname, basename, ext);
+		if (verbose > 4) {
+			rprintf(FINFO, "[%s] fuzzy score for %s = %u\n",
+				who_am_i(), dname, score);
+		}
+		if (score > bestscore) {
+			strlcpy(bestname, dname, sizeof bestname);
+			bestscore = score;
+		}
+	}
+	closedir(d);
+
+	/* Found a candidate. */
+	if (bestscore != 0) {
+		pathjoin(buf, MAXPATHLEN, dirname, bestname);
+		if (verbose > 2) {
+			rprintf(FINFO, "[%s] fuzzy match %s->%s\n",
+				who_am_i(), *fname_ptr, buf);
+		}
+		*fname_ptr = buf;
+		return link_stat(buf, st_ptr, 0);
+	}
+	return -1;
+}
+
 
 /*
  * Acts on file number @p i from @p flist, whose name is @p fname.
@@ -267,12 +356,12 @@ static void generate_and_send_sums(struc
  * out.  It might be wrong.
  */
 static void recv_generator(char *fname, struct file_struct *file, int i,
-			   int f_out)
+			   int f_out, int f_nameout)
 {
 	int fd;
 	STRUCT_STAT st;
 	struct map_struct *mapbuf;
-	int statret;
+	int statret, fuzzy_file = 0;
 	char *fnamecmp;
 	char fnamecmpbuf[MAXPATHLEN];
 
@@ -431,8 +520,10 @@ static void recv_generator(char *fname, 
 		statret = link_stat(fnamecmpbuf, &st, 0);
 		if (!S_ISREG(st.st_mode))
 			statret = -1;
-		if (statret == -1)
+		if (statret < 0) {
 			errno = saveerrno;
+			*fnamecmpbuf = '\0';
+		}
 #if HAVE_LINK
 		else if (link_dest && !dry_run) {
 			if (do_link(fnamecmpbuf, fname) != 0) {
@@ -440,18 +531,30 @@ static void recv_generator(char *fname, 
 					rsyserr(FINFO, errno, "link %s => %s",
 						fnamecmpbuf, fname);
 				}
-			}
-			fnamecmp = fnamecmpbuf;
+				fnamecmp = fnamecmpbuf;
+			} else
+				*fnamecmpbuf = '\0';
 		}
 #endif
 		else
 			fnamecmp = fnamecmpbuf;
+	} else
+		*fnamecmpbuf = '\0';
+
+	if (statret == -1 && fuzzy) {
+		statret = find_fuzzy(&fnamecmp, fnamecmpbuf, &st);
+		if (!S_ISREG(st.st_mode))
+			statret = -1;
+		else
+			fuzzy_file = 1;
 	}
 
 	if (statret == -1) {
 		if (preserve_hard_links && hard_link_check(file, HL_SKIP))
 			return;
 		if (errno == ENOENT) {
+			if (f_nameout >= 0)
+				write(f_nameout, "", 1);
 			write_int(f_out,i);
 			if (!dry_run)
 				write_sum_head(f_out, NULL);
@@ -471,37 +574,43 @@ static void recv_generator(char *fname, 
 		/* now pretend the file didn't exist */
 		if (preserve_hard_links && hard_link_check(file, HL_SKIP))
 			return;
+		if (f_nameout >= 0)
+			write(f_nameout, "", 1);
 		write_int(f_out,i);
 		if (!dry_run)
 			write_sum_head(f_out, NULL);
 		return;
 	}
 
-	if (opt_ignore_existing && fnamecmp == fname) {
+	if (opt_ignore_existing && !*fnamecmpbuf) {
 		if (verbose > 1)
 			rprintf(FINFO,"%s exists\n",fname);
 		return;
 	}
 
-	if (update_only && fnamecmp == fname
+	if (update_only && !*fnamecmpbuf
 	    && cmp_modtime(st.st_mtime, file->modtime) > 0) {
 		if (verbose > 1)
 			rprintf(FINFO,"%s is newer\n",fname);
 		return;
 	}
 
-	if (skip_file(fname, file, &st)) {
-		if (fnamecmp == fname)
+	if (!fuzzy_file && skip_file(fname, file, &st)) {
+		if (!*fnamecmpbuf)
 			set_perms(fname, file, &st, PERMS_REPORT);
 		return;
 	}
 
 	if (dry_run) {
+		if (f_nameout >= 0)
+			write(f_nameout, "", 1);
 		write_int(f_out,i);
 		return;
 	}
 
 	if (disable_deltas_p()) {
+		if (f_nameout >= 0)
+			write(f_nameout, "", 1);
 		write_int(f_out,i);
 		write_sum_head(f_out, NULL);
 		return;
@@ -516,6 +625,8 @@ static void recv_generator(char *fname, 
 		/* pretend the file didn't exist */
 		if (preserve_hard_links && hard_link_check(file, HL_SKIP))
 			return;
+		if (f_nameout >= 0)
+			write(f_nameout, "", 1);
 		write_int(f_out,i);
 		write_sum_head(f_out, NULL);
 		return;
@@ -534,6 +645,8 @@ static void recv_generator(char *fname, 
 	if (verbose > 2)
 		rprintf(FINFO, "generating and sending sums for %d\n", i);
 
+	if (f_nameout >= 0)
+		write(f_nameout, fnamecmpbuf, strlen(fnamecmpbuf) + 1);
 	write_int(f_out,i);
 	generate_and_send_sums(mapbuf, st.st_size, f_out);
 
@@ -543,7 +656,8 @@ static void recv_generator(char *fname, 
 }
 
 
-void generate_files(int f, struct file_list *flist, char *local_name)
+void generate_files(int f, struct file_list *flist, char *local_name,
+		    int f_nameout)
 {
 	int i;
 	int phase = 0;
@@ -584,7 +698,7 @@ void generate_files(int f, struct file_l
 		}
 
 		recv_generator(local_name ? local_name : f_name_to(file, fbuf),
-			       file, i, f);
+			       file, i, f, f_nameout);
 	}
 
 	phase++;
@@ -601,7 +715,7 @@ void generate_files(int f, struct file_l
 	while ((i = get_redo_num()) != -1) {
 		struct file_struct *file = flist->files[i];
 		recv_generator(local_name ? local_name : f_name_to(file, fbuf),
-			       file, i, f);
+			       file, i, f, f_nameout);
 	}
 
 	phase++;
@@ -620,7 +734,7 @@ void generate_files(int f, struct file_l
 		if (!file->basename || !S_ISDIR(file->mode))
 			continue;
 		recv_generator(local_name ? local_name : f_name(file),
-			       file, i, -1);
+			       file, i, -1, -1);
 	}
 
 	if (verbose > 2)
--- main.c	30 Jun 2004 07:27:30 -0000	1.202
+++ main.c	30 Jun 2004 07:43:47 -0000
@@ -429,7 +429,7 @@ static int do_recv(int f_in,int f_out,st
 {
 	int pid;
 	int status = 0;
-	int error_pipe[2];
+	int error_pipe[2], name_pipe[2];
 
 	if (preserve_hard_links)
 		init_hard_links(flist);
@@ -441,8 +441,8 @@ static int do_recv(int f_in,int f_out,st
 		}
 	}
 
-	if (fd_pair(error_pipe) < 0) {
-		rprintf(FERROR,"error pipe failed in do_recv\n");
+	if (fd_pair(error_pipe) < 0 || fd_pair(name_pipe) < 0) {
+		rprintf(FERROR, "fd_pair() failed in do_recv\n");
 		exit_cleanup(RERR_SOCKETIO);
 	}
 
@@ -450,8 +450,10 @@ static int do_recv(int f_in,int f_out,st
 
 	if ((pid = do_fork()) == 0) {
 		close(error_pipe[0]);
+		close(name_pipe[1]);
 		if (f_in != f_out)
 			close(f_out);
+		set_blocking(name_pipe[0]);
 
 		/* we can't let two processes write to the socket at one time */
 		io_multiplexing_close();
@@ -459,7 +461,7 @@ static int do_recv(int f_in,int f_out,st
 		/* set place to send errors */
 		set_msg_fd_out(error_pipe[1]);
 
-		recv_files(f_in,flist,local_name);
+		recv_files(f_in, flist, local_name, name_pipe[0]);
 		io_flush(FULL_FLUSH);
 		report(f_in);
 
@@ -475,14 +477,16 @@ static int do_recv(int f_in,int f_out,st
 	am_generator = 1;
 
 	close(error_pipe[1]);
+	close(name_pipe[0]);
 	if (f_in != f_out)
 		close(f_in);
+	set_blocking(name_pipe[1]);
 
 	io_start_buffering_out(f_out);
 
 	set_msg_fd_in(error_pipe[0]);
 
-	generate_files(f_out, flist, local_name);
+	generate_files(f_out, flist, local_name, name_pipe[1]);
 
 	get_redo_num(); /* Read final MSG_DONE and any prior messages. */
 	report(-1);
--- options.c	20 Jun 2004 19:47:05 -0000	1.157
+++ options.c	30 Jun 2004 07:43:47 -0000
@@ -94,6 +94,7 @@ int ignore_errors = 0;
 int modify_window = 0;
 int blocking_io = -1;
 int checksum_seed = 0;
+int fuzzy = 0;
 unsigned int block_size = 0;
 
 
@@ -270,6 +271,7 @@ void usage(enum logcode F)
   rprintf(F," -T  --temp-dir=DIR          create temporary files in directory DIR\n");
   rprintf(F,"     --compare-dest=DIR      also compare destination files relative to DIR\n");
   rprintf(F,"     --link-dest=DIR         create hardlinks to DIR for unchanged files\n");
+  rprintf(F,"     --fuzzy                 use similar file as basis if basis doesn't exist\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");
@@ -368,6 +370,7 @@ static struct poptOption long_options[] 
   {"temp-dir",        'T', POPT_ARG_STRING, &tmpdir, 0, 0, 0 },
   {"compare-dest",     0,  POPT_ARG_STRING, &compare_dest, 0, 0, 0 },
   {"link-dest",        0,  POPT_ARG_STRING, &compare_dest,  OPT_LINK_DEST, 0, 0 },
+  {"fuzzy",            0,  POPT_ARG_NONE,   &fuzzy, 0, 0, 0 },
   /* TODO: Should this take an optional int giving the compression level? */
   {"compress",        'z', POPT_ARG_NONE,   &do_compression, 0, 0, 0 },
   {"daemon",           0,  POPT_ARG_NONE,   &daemon_opt, 0, 0, 0 },
@@ -989,6 +992,9 @@ void server_options(char **args,int *arg
 		}
 	}
 
+	if (fuzzy && am_sender)
+		args[ac++] = "--fuzzy";
+
 	*argc = ac;
 	return;
 
--- receiver.c	30 Jun 2004 07:27:30 -0000	1.84
+++ receiver.c	30 Jun 2004 07:43:47 -0000
@@ -36,7 +36,6 @@ extern int preserve_perms;
 extern int cvs_exclude;
 extern int io_error;
 extern char *tmpdir;
-extern char *compare_dest;
 extern int make_backups;
 extern int do_progress;
 extern char *backup_dir;
@@ -293,14 +292,15 @@ static int receive_data(int f_in,struct 
  * main routine for receiver process.
  *
  * Receiver process runs on the same host as the generator process. */
-int recv_files(int f_in,struct file_list *flist,char *local_name)
+int recv_files(int f_in, struct file_list *flist, char *local_name,
+	       int f_name)
 {
 	int fd1,fd2;
 	STRUCT_STAT st;
 	char *fname, fbuf[MAXPATHLEN];
 	char template[MAXPATHLEN];
 	char fnametmp[MAXPATHLEN];
-	char *fnamecmp;
+	char *fnamecmp, *cp;
 	char fnamecmpbuf[MAXPATHLEN];
 	struct map_struct *mapbuf;
 	struct file_struct *file;
@@ -364,19 +364,19 @@ int recv_files(int f_in,struct file_list
 		if (verbose > 2)
 			rprintf(FINFO,"recv_files(%s)\n",fname);
 
-		fnamecmp = fname;
+		for (cp = fnamecmpbuf; ; cp++) {
+			if (read(f_name, cp, 1) <= 0) {
+				rsyserr(FERROR, errno, "fname-pipe read failed");
+				exit_cleanup(RERR_PROTOCOL);
+			}
+			if (!*cp)
+				break;
+		}
+		fnamecmp = *fnamecmpbuf ? fnamecmpbuf : fname;
 
 		/* open the file */
 		fd1 = do_open(fnamecmp, O_RDONLY, 0);
 
-		if (fd1 == -1 && compare_dest != NULL) {
-			/* try the file at compare_dest instead */
-			pathjoin(fnamecmpbuf, sizeof fnamecmpbuf,
-				 compare_dest, fname);
-			fnamecmp = fnamecmpbuf;
-			fd1 = do_open(fnamecmp, O_RDONLY, 0);
-		}
-
 		if (fd1 != -1 && do_fstat(fd1,&st) != 0) {
 			rsyserr(FERROR, errno, "fstat %s failed",
 				full_fname(fnamecmp));
@@ -385,7 +385,7 @@ int recv_files(int f_in,struct file_list
 			continue;
 		}
 
-		if (fd1 != -1 && S_ISDIR(st.st_mode) && fnamecmp == fname) {
+		if (fd1 != -1 && S_ISDIR(st.st_mode) && !*fnamecmpbuf) {
 			/* this special handling for directories
 			 * wouldn't be necessary if robust_rename()
 			 * and the underlying robust_unlink could cope
--- rsync.yo	5 Jun 2004 16:16:30 -0000	1.171
+++ rsync.yo	30 Jun 2004 07:43:48 -0000
@@ -325,6 +325,7 @@ verb(
  -T  --temp-dir=DIR          create temporary files in directory DIR
      --compare-dest=DIR      also compare received files relative to DIR
      --link-dest=DIR         create hardlinks to DIR for unchanged files
+     --fuzzy                 use similar file as basis if basis is gone
  -P                          equivalent to --partial --progress
  -z, --compress              compress file data
  -C, --cvs-exclude           auto ignore files in the same way CVS does