[PATCH] diff-raw format update take #2.

This changes the diff-raw format again, following the mailing
list discussion.  The new format explicitly expresses which one
is a rename and which one is a copy.

The documentation and tests are updated to match this change.

Signed-off-by: Junio C Hamano <junkio@cox.net>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

Documentation/diff-format.txt

@@ -19,28 +19,34 @@ git-diff-files [<pattern>...]::
 
 An output line is formatted this way:
 
-  ':' <mode> ' ' <mode> ' ' <sha1> ' ' <sha1> I <path> I <path> L
-
-By default, I and L are '\t' and '\n' respectively.  When '-z'
-flag is in effect, both I and L are '\0'.
-
-In each <mode>, <sha1> and <path> pair, left hand side describes
-the left hand side of what is being compared (<tree-ish> in
-git-diff-cache, <tree-ish-1> in git-diff-tree, cache contents in
-git-diff-files).  Non-existence is shown by having 000000 in the
-<mode> column.  That is, 000000 appears as the first <mode> for
-newly created files, and as the second <mode> for deleted files.
-
-Usually two <path> are the same.  When rename/copy detection is
-used, however, an "create" and another "delete" records can be
-merged into a single record that has two <path>, old name and
-new name.
+in-place edit  :100644 100644 bcd1234... 0123456... M file0
+copy-edit      :100644 100644 abcd123... 1234567... C68 file1 file2
+rename-edit    :100644 100644 abcd123... 1234567... R86 file1 file3
+create         :000000 100644 0000000... 1234567... N file4
+delete         :100644 000000 1234567... 0000000... D file5
+unmerged       :000000 000000 0000000... 0000000... U file6
+
+That is, from the left to the right:
+
+  (1) a colon.
+  (2) mode for "src"; 000000 if creation or unmerged.
+  (3) a space.
+  (4) mode for "dst"; 000000 if deletion or unmerged.
+  (5) a space.
+  (6) sha1 for "src"; 0{40} if creation or unmerged.
+  (7) a space.
+  (8) sha1 for "dst"; 0{40} if creation, unmerged or "look at work tree".
+  (9) status, followed by similarlity index number only for C and R.
+ (10) a tab or a NUL when '-z' option is used.
+ (11) path for "src"
+ (12) a tab or a NUL when '-z' option is used; only exists for C or R.
+ (13) path for "dst"; only exists for C or R.
+ (14) an LF or a NUL when '-z' option is used, to terminate the record.
 
 <sha1> is shown as all 0's if new is a file on the filesystem
 and it is out of sync with the cache.  Example:
 
-  :100644 100644 5be4a4...... 000000......    file.c    file.c
-
+  :100644 100644 5be4a4...... 000000...... M file.c
 
 Generating patches with -p
 --------------------------

Documentation/git-diff-helper.txt

@@ -9,7 +9,7 @@ git-diff-helper - Generates patch format output for git-diff-*
 
 SYNOPSIS
 --------
-'git-diff-helper' [-z] [-R] [-M] [-C] [-S<string>]
+'git-diff-helper' [-z] [-S<string>]
 
 DESCRIPTION
 -----------
@@ -21,22 +21,6 @@ OPTIONS
 -z::
 	\0 line termination on input
 
--R::
-	Output diff in reverse.  This is useful for displaying output from
-	"git-diff-cache" which always compares tree with cache or working
-	file.  E.g.
-
-		git-diff-cache <tree> | git-diff-helper -R file.c
-
-	would show a diff to bring the working file back to what
-	is in the <tree>.
-
--M::
-	Detect renames.
-
--C::
-	Detect copies as well as renames.
-
 -S<string>::
 	Look for differences that contains the change in <string>.
 

diff-cache.c

@@ -234,6 +234,6 @@ int main(int argc, const char **argv)
 		diffcore_pickaxe(pickaxe);
 	if (2 <= argc)
 		diffcore_pathspec(argv + 1);
-	diff_flush(diff_output_format);
+	diff_flush(diff_output_format, 1);
 	return ret;
 }

diff-files.c

@@ -120,6 +120,6 @@ int main(int argc, const char **argv)
 		diffcore_pickaxe(pickaxe);
 	if (1 < argc)
 		diffcore_pathspec(argv + 1);
-	diff_flush(diff_output_format);
+	diff_flush(diff_output_format, 1);
 	return 0;
 }

diff-helper.c

@@ -5,84 +5,21 @@
 #include "strbuf.h"
 #include "diff.h"
 
-static int detect_rename = 0;
-static int diff_score_opt = 0;
 static const char *pickaxe = NULL;
-static int diff_output_style = DIFF_FORMAT_PATCH;
 static int line_termination = '\n';
 static int inter_name_termination = '\t';
 
-static int parse_diff_raw(char *buf1, char *buf2, char *buf3)
-{
-	char old_path[PATH_MAX];
-	unsigned char old_sha1[20], new_sha1[20];
-	char *ep;
-	char *cp = buf1;
-	int ch, old_mode, new_mode;
-
-	old_mode = new_mode = 0;
-	while ((ch = *cp) && ('0' <= ch && ch <= '7')) {
-		old_mode = (old_mode << 3) | (ch - '0');
-		cp++;
-	}
-	if (*cp++ != ' ')
-		return -1;
-	while ((ch = *cp) && ('0' <= ch && ch <= '7')) {
-		new_mode = (new_mode << 3) | (ch - '0');
-		cp++;
-	}
-	if (*cp++ != ' ')
-		return -1;
-	if (get_sha1_hex(cp, old_sha1))
-		return -1;
-	cp += 40;
-	if (*cp++ != ' ')
-		return -1;
-	if (get_sha1_hex(cp, new_sha1))
-		return -1;
-	cp += 40;
-	if (*cp++ != inter_name_termination)
-		return -1;
-	if (buf2)
-		cp = buf2;
-	ep = strchr(cp, inter_name_termination);
-	if (!ep)
-		return -1;
-	*ep++ = 0;
-	strcpy(old_path, cp);
-	diff_guif(old_mode, new_mode, old_sha1, new_sha1,
-		  old_path, buf3 ? buf3 : ep);
-	return 0;
-}
-
 static const char *diff_helper_usage =
-	"git-diff-helper [-z] [-R] [-M] [-C] [-S<string>] paths...";
+	"git-diff-helper [-z] [-S<string>] paths...";
 
 int main(int ac, const char **av) {
-	struct strbuf sb1, sb2, sb3;
-	int reverse_diff = 0;
+	struct strbuf sb;
 
-	strbuf_init(&sb1);
-	strbuf_init(&sb2);
-	strbuf_init(&sb3);
+	strbuf_init(&sb);
 
 	while (1 < ac && av[1][0] == '-') {
-		if (av[1][1] == 'R')
-			reverse_diff = 1;
-		else if (av[1][1] == 'z')
+		if (av[1][1] == 'z')
 			line_termination = inter_name_termination = 0;
-		else if (av[1][1] == 'p') /* hidden from the help */
-			diff_output_style = DIFF_FORMAT_HUMAN;
-		else if (av[1][1] == 'P') /* hidden from the help */
-			diff_output_style = DIFF_FORMAT_MACHINE;
-		else if (av[1][1] == 'M') {
-			detect_rename = DIFF_DETECT_RENAME;
-			diff_score_opt = diff_scoreopt_parse(av[1]);
-		}
-		else if (av[1][1] == 'C') {
-			detect_rename = DIFF_DETECT_COPY;
-			diff_score_opt = diff_scoreopt_parse(av[1]);
-		}
 		else if (av[1][1] == 'S') {
 			pickaxe = av[1] + 2;
 		}
@@ -92,45 +29,114 @@ int main(int ac, const char **av) {
 	}
 	/* the remaining parameters are paths patterns */
 
-	diff_setup(reverse_diff);
+	diff_setup(0);
 	while (1) {
-		int status;
-		read_line(&sb1, stdin, line_termination);
-		if (sb1.eof)
+		unsigned old_mode, new_mode;
+		unsigned char old_sha1[20], new_sha1[20];
+		char old_path[PATH_MAX];
+		int status, score, two_paths;
+		char new_path[PATH_MAX];
+
+		int ch;
+		char *cp, *ep;
+
+		read_line(&sb, stdin, line_termination);
+		if (sb.eof)
 			break;
-		switch (sb1.buf[0]) {
-		case 'U':
-			diff_unmerge(sb1.buf + 2);
-			continue;
+		switch (sb.buf[0]) {
 		case ':':
-			break;
-		default:
-			goto unrecognized;
-		}
-		if (!line_termination) {
-			read_line(&sb2, stdin, line_termination);
-			if (sb2.eof)
+			/* parse the first part up to the status */
+			cp = sb.buf + 1;
+			old_mode = new_mode = 0;
+			while ((ch = *cp) && ('0' <= ch && ch <= '7')) {
+				old_mode = (old_mode << 3) | (ch - '0');
+				cp++;
+			}
+			if (*cp++ != ' ')
 				break;
-			read_line(&sb3, stdin, line_termination);
-			if (sb3.eof)
+			while ((ch = *cp) && ('0' <= ch && ch <= '7')) {
+				new_mode = (new_mode << 3) | (ch - '0');
+				cp++;
+			}
+			if (*cp++ != ' ')
 				break;
-			status = parse_diff_raw(sb1.buf+1, sb2.buf, sb3.buf);
-		}
-		else
-			status = parse_diff_raw(sb1.buf+1, NULL, NULL);
-		if (status) {
-		unrecognized:
-			diff_flush(diff_output_style);
-			printf("%s\n", sb1.buf);
+			if (get_sha1_hex(cp, old_sha1))
+				break;
+			cp += 40;
+			if (*cp++ != ' ')
+				break;
+			if (get_sha1_hex(cp, new_sha1))
+				break;
+			cp += 40;
+			if (*cp++ != ' ')
+				break;
+			status = *cp++;
+			if (!strchr("MCRNDU", status))
+				break;
+			two_paths = score = 0;
+			if (status == 'R' || status == 'C') {
+				two_paths = 1;
+				sscanf(cp, "%d", &score);
+				if (line_termination) {
+					cp = strchr(cp,
+						    inter_name_termination);
+					if (!cp)
+						break;
+				}
+			}
+
+			if (*cp++ != inter_name_termination)
+				break;
+
+			/* first pathname */
+			if (!line_termination) {
+				read_line(&sb, stdin, line_termination);
+				if (sb.eof)
+					break;
+				strcpy(old_path, sb.buf);
+			}
+			else if (!two_paths)
+				strcpy(old_path, cp);
+			else {
+				ep = strchr(cp, inter_name_termination);
+				if (!ep)
+					break;
+				strncpy(old_path, cp, ep-cp);
+				old_path[ep-cp] = 0;
+				cp = ep + 1;
+			}
+
+			/* second pathname */
+			if (!two_paths)
+				strcpy(new_path, old_path);
+			else {
+				if (!line_termination) {
+					read_line(&sb, stdin,
+						  line_termination);
+					if (sb.eof)
+						break;
+					strcpy(new_path, sb.buf);
+				}
+				else
+					strcpy(new_path, cp);
+			}
+			diff_helper_input(old_mode, new_mode,
+					  old_sha1, new_sha1,
+					  old_path, status, score,
+					  new_path);
+			continue;
 		}
+		if (pickaxe)
+			diffcore_pickaxe(pickaxe);
+		if (1 < ac)
+			diffcore_pathspec(av + 1);
+		diff_flush(DIFF_FORMAT_PATCH, 0);
+		printf("%s\n", sb.buf);
 	}
-	if (detect_rename)
-		diffcore_rename(detect_rename, diff_score_opt);
-	diffcore_prune();
 	if (pickaxe)
 		diffcore_pickaxe(pickaxe);
-	if (ac)
+	if (1 < ac)
 		diffcore_pathspec(av + 1);
-	diff_flush(diff_output_style);
+	diff_flush(DIFF_FORMAT_PATCH, 0);
 	return 0;
 }

diff-tree.c

@@ -270,7 +270,7 @@ static int call_diff_flush(void)
 	if (pickaxe) {
 		diffcore_pickaxe(pickaxe);
 		if (diff_queue_is_empty()) {
-			diff_flush(DIFF_FORMAT_NO_OUTPUT);
+			diff_flush(DIFF_FORMAT_NO_OUTPUT, 0);
 			return 0;
 		}
 	}
@@ -291,7 +291,7 @@ static int call_diff_flush(void)
 		}
 		header = NULL;
 	}
-	diff_flush(diff_output_format);
+	diff_flush(diff_output_format, 1);
 	return 1;
 }