git-blame --porcelain

The new option makes the command's native output format to emit
output that is easier to handle by Porcelain.

Each line is output after a header.  The header at the minimum
has the first line which has:

 - 40-byte SHA-1 of the commit the line is attributed to;

 - the line number of the line in the original file;

 - the line number of the line in the final file;

 - on a line that starts a group of line from a different commit
   than the previous one, the number of lines in this group.  On
   subsequent lines this field is absent.

This header line is followed by the following information once
for each commit:

 - author name ("author"), email ("author-mail"), time
   ("author-time"), and timezone ("author-tz"); similarly for
   committer.

 - filename in the commit the line is attributed to.

 - the first line of the commit log message ("summary").

The contents of the actual line is output after the above
header, prefixed by a TAB. This is to allow adding more header
elements later.

Signed-off-by: Junio C Hamano <junkio@cox.net>

blame.c

@@ -17,6 +17,7 @@
 #include "diffcore.h"
 #include "revision.h"
 #include "xdiff-interface.h"
+#include "quote.h"
 
 #define DEBUG 0
 
@@ -40,6 +41,7 @@ struct util_info {
 	unsigned long size;
 	int num_lines;
 	const char *pathname;
+	unsigned meta_given:1;
 
 	void *topo_data;
 };
@@ -332,12 +334,8 @@ static struct util_info *get_util(struct commit *commit)
 	if (util)
 		return util;
 
-	util = xmalloc(sizeof(struct util_info));
+	util = xcalloc(1, sizeof(struct util_info));
-	util->buf = NULL;
-	util->size = 0;
-	util->line_map = NULL;
 	util->num_lines = -1;
-	util->pathname = NULL;
 	commit->util = util;
 	return util;
 }
@@ -642,39 +640,99 @@ struct commit_info
 	char *author_mail;
 	unsigned long author_time;
 	char *author_tz;
+
+	/* filled only when asked for details */
+	char *committer;
+	char *committer_mail;
+	unsigned long committer_time;
+	char *committer_tz;
+
+	char *summary;
 };
 
-static void get_commit_info(struct commit *commit, struct commit_info *ret)
+static void get_ac_line(const char *inbuf, const char *what,
+			int bufsz, char *person, char **mail,
+			unsigned long *time, char **tz)
 {
 	int len;
-	char *tmp;
+	char *tmp, *endp;
-	static char author_buf[1024];
+
-
+	tmp = strstr(inbuf, what);
-	tmp = strstr(commit->buffer, "\nauthor ") + 8;
+	if (!tmp)
-	len = strchr(tmp, '\n') - tmp;
+		goto error_out;
-	ret->author = author_buf;
+	tmp += strlen(what);
-	memcpy(ret->author, tmp, len);
+	endp = strchr(tmp, '\n');
+	if (!endp)
+		len = strlen(tmp);
+	else
+		len = endp - tmp;
+	if (bufsz <= len) {
+	error_out:
+		/* Ugh */
+		person = *mail = *tz = "(unknown)";
+		*time = 0;
+		return;
+	}
+	memcpy(person, tmp, len);
 
-	tmp = ret->author;
+	tmp = person;
 	tmp += len;
 	*tmp = 0;
 	while (*tmp != ' ')
 		tmp--;
-	ret->author_tz = tmp+1;
+	*tz = tmp+1;
 
 	*tmp = 0;
 	while (*tmp != ' ')
 		tmp--;
-	ret->author_time = strtoul(tmp, NULL, 10);
+	*time = strtoul(tmp, NULL, 10);
 
 	*tmp = 0;
 	while (*tmp != ' ')
 		tmp--;
-	ret->author_mail = tmp + 1;
+	*mail = tmp + 1;
-
 	*tmp = 0;
 }
 
+static void get_commit_info(struct commit *commit, struct commit_info *ret, int detailed)
+{
+	int len;
+	char *tmp, *endp;
+	static char author_buf[1024];
+	static char committer_buf[1024];
+	static char summary_buf[1024];
+
+	ret->author = author_buf;
+	get_ac_line(commit->buffer, "\nauthor ",
+		    sizeof(author_buf), author_buf, &ret->author_mail,
+		    &ret->author_time, &ret->author_tz);
+
+	if (!detailed)
+		return;
+
+	ret->committer = committer_buf;
+	get_ac_line(commit->buffer, "\ncommitter ",
+		    sizeof(committer_buf), committer_buf, &ret->committer_mail,
+		    &ret->committer_time, &ret->committer_tz);
+
+	ret->summary = summary_buf;
+	tmp = strstr(commit->buffer, "\n\n");
+	if (!tmp) {
+	error_out:
+		sprintf(summary_buf, "(%s)", sha1_to_hex(commit->object.sha1));
+		return;
+	}
+	tmp += 2;
+	endp = strchr(tmp, '\n');
+	if (!endp)
+		goto error_out;
+	len = endp - tmp;
+	if (len >= sizeof(summary_buf))
+		goto error_out;
+	memcpy(summary_buf, tmp, len);
+	summary_buf[len] = 0;
+}
+
 static const char *format_time(unsigned long time, const char *tz_str,
 			       int show_raw_time)
 {
@@ -751,7 +809,7 @@ static int find_orig_linenum(struct util_info *u, int lineno)
 }
 
 static void emit_meta(struct commit *c, int lno,
-		      int sha1_len, int compatibility,
+		      int sha1_len, int compatibility, int porcelain,
 		      int show_name, int show_number, int show_raw_time,
 		      int longest_file, int longest_author,
 		      int max_digits, int max_orig_digits)
@@ -763,7 +821,47 @@ static void emit_meta(struct commit *c, int lno,
 	u = c->util;
 	lineno = find_orig_linenum(u, lno);
 
-	get_commit_info(c, &ci);
+	if (porcelain) {
+		int group_size = -1;
+		struct commit *cc = (lno == 0) ? NULL : blame_lines[lno-1];
+		if (cc != c) {
+			/* This is the beginning of this group */
+			int i;
+			for (i = lno + 1; i < num_blame_lines; i++)
+				if (blame_lines[i] != c)
+					break;
+			group_size = i - lno;
+		}
+		if (0 < group_size)
+			printf("%s %d %d %d\n", sha1_to_hex(c->object.sha1),
+			       lineno, lno + 1, group_size);
+		else
+			printf("%s %d %d\n", sha1_to_hex(c->object.sha1),
+			       lineno, lno + 1);
+		if (!u->meta_given) {
+			get_commit_info(c, &ci, 1);
+			printf("author %s\n", ci.author);
+			printf("author-mail %s\n", ci.author_mail);
+			printf("author-time %lu\n", ci.author_time);
+			printf("author-tz %s\n", ci.author_tz);
+			printf("committer %s\n", ci.committer);
+			printf("committer-mail %s\n", ci.committer_mail);
+			printf("committer-time %lu\n", ci.committer_time);
+			printf("committer-tz %s\n", ci.committer_tz);
+			printf("filename ");
+			if (quote_c_style(u->pathname, NULL, NULL, 0))
+				quote_c_style(u->pathname, NULL, stdout, 0);
+			else
+				fputs(u->pathname, stdout);
+			printf("\nsummary %s\n", ci.summary);
+
+			u->meta_given = 1;
+		}
+		putchar('\t');
+		return;
+	}
+
+	get_commit_info(c, &ci, 0);
 	fwrite(sha1_to_hex(c->object.sha1), sha1_len, 1, stdout);
 	if (compatibility) {
 		printf("\t(%10s\t%10s\t%d)", ci.author,
@@ -809,6 +907,7 @@ int main(int argc, const char **argv)
 	int longest_file, longest_author, longest_file_lines;
 	int show_name = 0;
 	int show_number = 0;
+	int porcelain = 0;
 
 	const char *prefix = setup_git_directory();
 	git_config(git_default_config);
@@ -852,6 +951,12 @@ int main(int argc, const char **argv)
 				show_number = 1;
 				continue;
 			}
+			if (!strcmp(argv[i], "--porcelain")) {
+				porcelain = 1;
+				sha1_len = 40;
+				show_raw_time = 1;
+				continue;
+			}
 			if (!strcmp(argv[i], "--")) {
 				options = 0;
 				continue;
@@ -934,7 +1039,7 @@ int main(int argc, const char **argv)
 			longest_file = strlen(u->pathname);
 		if (longest_file_lines < u->num_lines)
 			longest_file_lines = u->num_lines;
-		get_commit_info(c, &ci);
+		get_commit_info(c, &ci, 0);
 		if (longest_author < strlen(ci.author))
 			longest_author = strlen(ci.author);
 	}
@@ -943,7 +1048,7 @@ int main(int argc, const char **argv)
 
 	for (i = 0; i < num_blame_lines; i++) {
 		emit_meta(blame_lines[i], i,
-			  sha1_len, compatibility,
+			  sha1_len, compatibility, porcelain,
 			  show_name, show_number, show_raw_time,
 			  longest_file, longest_author,
 			  max_digits, max_orig_digits);