fast-export: do not modify memory from get_commit_buffer
fast-export's helper function find_encoding() takes a `const char *`, but modifies that memory despite the `const`. Ultimately, this memory came from get_commit_buffer(), and you're not supposed to modify the memory that you get from get_commit_buffer(). So, get rid of find_encoding() in favor of commit.h:find_commit_header(), which gives back a string length, rather than mutating the memory to insert a '\0' terminator. Because find_commit_header() detects the "\n\n" string that separates the headers and the commit message, move the call to be above the `message = strstr(..., "\n\n")` call. This helps readability, and allows for the value of `encoding` to be used for a better value of "..." so that the same memory doesn't need to be checked twice. Introduce a `commit_buffer_cursor` variable to avoid writing an awkward `encoding ? encoding + encoding_len : committer_end` expression. Signed-off-by: Luke Shumaker <lukeshu@datawire.io> Signed-off-by: Christian Couder <chriscool@tuxfamily.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Luke Shumaker committed
Mar 10, 2025 at 16:57 UTC
dda9bff3c55e45ee8dbfb49fd972c4b35fd4ba7e
1 file changed
+33
-28
builtin/fast-export.c
+33
-28
@@ -510,21 +510,6 @@ static void show_filemodify(struct diff_queue_struct *q,
510
}
511
}
512
513
-static const char *find_encoding(const char *begin, const char *end)
514
-{
515
- const char *needle = "\nencoding ";
516
- char *bol, *eol;
517
-
518
- bol = memmem(begin, end ? end - begin : strlen(begin),
519
- needle, strlen(needle));
520
- if (!bol)
521
- return NULL;
522
- bol += strlen(needle);
523
- eol = strchrnul(bol, '\n');
524
- *eol = '\0';
525
- return bol;
526
-}
527
-
513
static char *anonymize_ref_component(void)
514
{
515
static int counter;
@@ -630,9 +615,11 @@ static void handle_commit(struct commit *commit, struct rev_info *rev,
615
struct string_list *paths_of_changed_objects)
616
{
617
int saved_output_format = rev->diffopt.output_format;
633
- const char *commit_buffer;
618
+ const char *commit_buffer, *commit_buffer_cursor;
619
const char *author, *author_end, *committer, *committer_end;
635
- const char *encoding, *message;
620
+ const char *encoding = NULL;
621
+ size_t encoding_len;
622
+ const char *message;
623
char *reencoded = NULL;
624
struct commit_list *p;
625
const char *refname;
@@ -641,21 +628,35 @@ static void handle_commit(struct commit *commit, struct rev_info *rev,
628
rev->diffopt.output_format = DIFF_FORMAT_CALLBACK;
629
630
parse_commit_or_die(commit);
644
- commit_buffer = repo_get_commit_buffer(the_repository, commit, NULL);
645
- author = strstr(commit_buffer, "\nauthor ");
631
+ commit_buffer_cursor = commit_buffer = repo_get_commit_buffer(the_repository, commit, NULL);
632
+
633
+ author = strstr(commit_buffer_cursor, "\nauthor ");
634
if (!author)
635
die("could not find author in commit %s",
636
oid_to_hex(&commit->object.oid));
637
author++;
650
- author_end = strchrnul(author, '\n');
651
- committer = strstr(author_end, "\ncommitter ");
638
+ commit_buffer_cursor = author_end = strchrnul(author, '\n');
639
+
640
+ committer = strstr(commit_buffer_cursor, "\ncommitter ");
641
if (!committer)
642
die("could not find committer in commit %s",
643
oid_to_hex(&commit->object.oid));
644
committer++;
656
- committer_end = strchrnul(committer, '\n');
657
- message = strstr(committer_end, "\n\n");
658
- encoding = find_encoding(committer_end, message);
645
+ commit_buffer_cursor = committer_end = strchrnul(committer, '\n');
646
+
647
+ /*
648
+ * find_commit_header() gets a `+ 1` because
649
+ * commit_buffer_cursor points at the trailing "\n" at the end
650
+ * of the previous line, but find_commit_header() wants a
651
+ * pointer to the beginning of the next line.
652
+ */
653
+ if (*commit_buffer_cursor == '\n') {
654
+ encoding = find_commit_header(commit_buffer_cursor + 1, "encoding", &encoding_len);
655
+ if (encoding)
656
+ commit_buffer_cursor = encoding + encoding_len;
657
+ }
658
+
659
+ message = strstr(commit_buffer_cursor, "\n\n");
660
if (message)
661
message += 2;
662
@@ -694,16 +695,20 @@ static void handle_commit(struct commit *commit, struct rev_info *rev,
695
if (anonymize) {
696
reencoded = anonymize_commit_message();
697
} else if (encoding) {
698
+ char *buf;
699
switch (reencode_mode) {
700
case REENCODE_YES:
699
- reencoded = reencode_string(message, "UTF-8", encoding);
701
+ buf = xstrfmt("%.*s", (int)encoding_len, encoding);
702
+ reencoded = reencode_string(message, "UTF-8", buf);
703
+ free(buf);
704
break;
705
case REENCODE_NO:
706
break;
707
case REENCODE_ABORT:
704
- die("Encountered commit-specific encoding %s in commit "
708
+ die("Encountered commit-specific encoding %.*s in commit "
709
"%s; use --reencode=[yes|no] to handle it",
706
- encoding, oid_to_hex(&commit->object.oid));
710
+ (int)encoding_len, encoding,
711
+ oid_to_hex(&commit->object.oid));
712
}
713
}
714
if (!commit->parents)
@@ -715,7 +720,7 @@ static void handle_commit(struct commit *commit, struct rev_info *rev,
720
(int)(author_end - author), author,
721
(int)(committer_end - committer), committer);
722
if (!reencoded && encoding)
718
- printf("encoding %s\n", encoding);
723
+ printf("encoding %.*s\n", (int)encoding_len, encoding);
724
printf("data %u\n%s",
725
(unsigned)(reencoded
726
? strlen(reencoded) : message