apply: handle traditional patches with space in filename
authorJonathan Nieder <jrnieder@gmail.com>
Thu, 19 Aug 2010 01:50:14 +0000 (20:50 -0500)
committerJunio C Hamano <gitster@pobox.com>
Sun, 22 Aug 2010 06:04:29 +0000 (23:04 -0700)
To discover filenames from the --- and +++ lines in a traditional
unified diff, currently "git apply" scans forward for a whitespace
character on each line and stops there. It can't use the whole line
because "diff -u" likes to include timestamps, like so:

--- foo 2000-07-12 16:56:50.020000414 -0500
+++ bar 2010-07-12 16:56:50.020000414 -0500

The whitespace-seeking heuristic works great, even when the tab
has been converted to spaces by some email + copy-and-paste
related corruption.

Except for one problem: if the filename itself contains whitespace,
the inferred filename will be too short.

When Giuseppe ran into this problem, it was for a file creation
patch (for debian/licenses/LICENSE.global BSD-style Chromium).
So one can't use the list of files present in the index to deduce an
appropriate filename (not to mention that way lies madness; see
v0.99~402, 2005-05-31).

Instead, look for a timestamp and use that if present to mark the end
of the filename. If no timestamp is present, the old heuristic is
used, with one exception: the space character \040 is not considered
terminating whitespace any more unless it is followed by a timestamp.

Reported-by: Giuseppe Iuculano <iuculano@debian.org>
Acked-by: Guido Günther <agx@sigxcpu.org>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
builtin/apply.c
t/t4135-apply-weird-filenames.sh
index efc109e5d0a990013fe0212cffd1a0faad2ad123..bd2fcb39f0ea42963df8d18f9910f19960e82f8d 100644 (file)
@@ -449,23 +449,157 @@ static char *find_name_gnu(const char *line, char *def, int p_value)
        return squash_slash(strbuf_detach(&name, NULL));
 }
 
-static char *find_name(const char *line, char *def, int p_value, int terminate)
+static size_t tz_len(const char *line, size_t len)
 {
-       int len;
-       const char *start = NULL;
+       const char *tz, *p;
 
-       if (*line == '"') {
-               char *name = find_name_gnu(line, def, p_value);
-               if (name)
-                       return name;
+       if (len < strlen(" +0500") || line[len-strlen(" +0500")] != ' ')
+               return 0;
+       tz = line + len - strlen(" +0500");
+
+       if (tz[1] != '+' && tz[1] != '-')
+               return 0;
+
+       for (p = tz + 2; p != line + len; p++)
+               if (!isdigit(*p))
+                       return 0;
+
+       return line + len - tz;
+}
+
+static size_t date_len(const char *line, size_t len)
+{
+       const char *date, *p;
+
+       if (len < strlen("72-02-05") || line[len-strlen("-05")] != '-')
+               return 0;
+       p = date = line + len - strlen("72-02-05");
+
+       if (!isdigit(*p++) || !isdigit(*p++) || *p++ != '-' ||
+           !isdigit(*p++) || !isdigit(*p++) || *p++ != '-' ||
+           !isdigit(*p++) || !isdigit(*p++))   /* Not a date. */
+               return 0;
+
+       if (date - line >= strlen("19") &&
+           isdigit(date[-1]) && isdigit(date[-2]))     /* 4-digit year */
+               date -= strlen("19");
+
+       return line + len - date;
+}
+
+static size_t short_time_len(const char *line, size_t len)
+{
+       const char *time, *p;
+
+       if (len < strlen(" 07:01:32") || line[len-strlen(":32")] != ':')
+               return 0;
+       p = time = line + len - strlen(" 07:01:32");
+
+       /* Permit 1-digit hours? */
+       if (*p++ != ' ' ||
+           !isdigit(*p++) || !isdigit(*p++) || *p++ != ':' ||
+           !isdigit(*p++) || !isdigit(*p++) || *p++ != ':' ||
+           !isdigit(*p++) || !isdigit(*p++))   /* Not a time. */
+               return 0;
+
+       return line + len - time;
+}
+
+static size_t fractional_time_len(const char *line, size_t len)
+{
+       const char *p;
+       size_t n;
+
+       /* Expected format: 19:41:17.620000023 */
+       if (!len || !isdigit(line[len - 1]))
+               return 0;
+       p = line + len - 1;
+
+       /* Fractional seconds. */
+       while (p > line && isdigit(*p))
+               p--;
+       if (*p != '.')
+               return 0;
+
+       /* Hours, minutes, and whole seconds. */
+       n = short_time_len(line, p - line);
+       if (!n)
+               return 0;
+
+       return line + len - p + n;
+}
+
+static size_t trailing_spaces_len(const char *line, size_t len)
+{
+       const char *p;
+
+       /* Expected format: ' ' x (1 or more)  */
+       if (!len || line[len - 1] != ' ')
+               return 0;
+
+       p = line + len;
+       while (p != line) {
+               p--;
+               if (*p != ' ')
+                       return line + len - (p + 1);
        }
 
+       /* All spaces! */
+       return len;
+}
+
+static size_t diff_timestamp_len(const char *line, size_t len)
+{
+       const char *end = line + len;
+       size_t n;
+
+       /*
+        * Posix: 2010-07-05 19:41:17
+        * GNU: 2010-07-05 19:41:17.620000023 -0500
+        */
+
+       if (!isdigit(end[-1]))
+               return 0;
+
+       n = tz_len(line, end - line);
+       end -= n;
+
+       n = short_time_len(line, end - line);
+       if (!n)
+               n = fractional_time_len(line, end - line);
+       end -= n;
+
+       n = date_len(line, end - line);
+       if (!n) /* No date.  Too bad. */
+               return 0;
+       end -= n;
+
+       if (end == line)        /* No space before date. */
+               return 0;
+       if (end[-1] == '\t') {  /* Success! */
+               end--;
+               return line + len - end;
+       }
+       if (end[-1] != ' ')     /* No space before date. */
+               return 0;
+
+       /* Whitespace damage. */
+       end -= trailing_spaces_len(line, end - line);
+       return line + len - end;
+}
+
+static char *find_name_common(const char *line, char *def, int p_value,
+                               const char *end, int terminate)
+{
+       int len;
+       const char *start = NULL;
+
        if (p_value == 0)
                start = line;
-       for (;;) {
+       while (line != end) {
                char c = *line;
 
-               if (isspace(c)) {
+               if (!end && isspace(c)) {
                        if (c == '\n')
                                break;
                        if (name_terminate(start, line-start, c, terminate))
@@ -505,6 +639,37 @@ static char *find_name(const char *line, char *def, int p_value, int terminate)
        return squash_slash(xmemdupz(start, len));
 }
 
+static char *find_name(const char *line, char *def, int p_value, int terminate)
+{
+       if (*line == '"') {
+               char *name = find_name_gnu(line, def, p_value);
+               if (name)
+                       return name;
+       }
+
+       return find_name_common(line, def, p_value, NULL, terminate);
+}
+
+static char *find_name_traditional(const char *line, char *def, int p_value)
+{
+       size_t len = strlen(line);
+       size_t date_len;
+
+       if (*line == '"') {
+               char *name = find_name_gnu(line, def, p_value);
+               if (name)
+                       return name;
+       }
+
+       len = strchrnul(line, '\n') - line;
+       date_len = diff_timestamp_len(line, len);
+       if (!date_len)
+               return find_name_common(line, def, p_value, NULL, TERM_TAB);
+       len -= date_len;
+
+       return find_name_common(line, def, p_value, line + len, 0);
+}
+
 static int count_slashes(const char *cp)
 {
        int cnt = 0;
@@ -527,7 +692,7 @@ static int guess_p_value(const char *nameline)
 
        if (is_dev_null(nameline))
                return -1;
-       name = find_name(nameline, NULL, 0, TERM_SPACE | TERM_TAB);
+       name = find_name_traditional(nameline, NULL, 0);
        if (!name)
                return -1;
        cp = strchr(name, '/');
@@ -646,16 +811,16 @@ static void parse_traditional_patch(const char *first, const char *second, struc
        if (is_dev_null(first)) {
                patch->is_new = 1;
                patch->is_delete = 0;
-               name = find_name(second, NULL, p_value, TERM_SPACE | TERM_TAB);
+               name = find_name_traditional(second, NULL, p_value);
                patch->new_name = name;
        } else if (is_dev_null(second)) {
                patch->is_new = 0;
                patch->is_delete = 1;
-               name = find_name(first, NULL, p_value, TERM_SPACE | TERM_TAB);
+               name = find_name_traditional(first, NULL, p_value);
                patch->old_name = name;
        } else {
-               name = find_name(first, NULL, p_value, TERM_SPACE | TERM_TAB);
-               name = find_name(second, name, p_value, TERM_SPACE | TERM_TAB);
+               name = find_name_traditional(first, NULL, p_value);
+               name = find_name_traditional(second, name, p_value);
                if (has_epoch_timestamp(first)) {
                        patch->is_new = 1;
                        patch->is_delete = 0;
index 9373f644a8654e02f62f56a2ebe3f74bceb714af..1e5aad57ab799819c6f319eaf813149d8bdda136 100755 (executable)
@@ -59,8 +59,8 @@ try_filename() {
 }
 
 try_filename 'plain'            'postimage.txt'
-try_filename 'with spaces'      'post image.txt' '' success failure failure
-try_filename 'with tab'         'post  image.txt' FUNNYNAMES success failure failure
+try_filename 'with spaces'      'post image.txt'
+try_filename 'with tab'         'post  image.txt' FUNNYNAMES
 try_filename 'with backslash'   'post\image.txt' BSLASHPSPEC
 try_filename 'with quote'       '"postimage".txt' FUNNYNAMES success failure success