Florian Achleitner <florian.achleitner.2.6.31@xxxxxxxxx> writes:
> fast-import mark files are stored outside the object database and are
> therefore not fetched and can be lost somehow else. marks provide a
> svn revision --> git sha1 mapping, while the notes that are attached
> to each commit when it is imported provide a git sha1 --> svn revision
> mapping.
>
> If the marks file is not available or not plausible, regenerate it by
> walking through the notes tree. , i.e. The plausibility check tests
> if the highest revision in the marks file matches the revision of the
> top ref. It doesn't ensure that the mark file is completely correct.
> This could only be done with an effort equal to unconditional
> regeneration.
>
> Signed-off-by: Florian Achleitner <florian.achleitner.2.6.31@xxxxxxxxx>
> Signed-off-by: Junio C Hamano <gitster@xxxxxxxxx>
> ---
> remote-testsvn.c | 63 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
> 1 file changed, 63 insertions(+)
>
> diff --git a/remote-testsvn.c b/remote-testsvn.c
> index e90d221..2c0dc99 100644
> --- a/remote-testsvn.c
> +++ b/remote-testsvn.c
> @@ -86,6 +86,68 @@ static int parse_rev_note(const char *msg, struct rev_note *res) {
> return 0;
> }
>
> +static int note2mark_cb(const unsigned char *object_sha1,
> + const unsigned char *note_sha1, char *note_path,
> + void *cb_data) {
> + FILE *file = (FILE *)cb_data;
> + char *msg;
> + unsigned long msglen;
> + enum object_type type;
> + struct rev_note note;
> + if (!(msg = read_sha1_file(note_sha1, &type, &msglen)) ||
> + !msglen || type != OBJ_BLOB) {
> + free(msg);
> + return 1;
> + }
> + if (parse_rev_note(msg, ¬e))
> + return 2;
> + if (fprintf(file, ":%d %s\n", note.rev_nr, sha1_to_hex(object_sha1)) < 1)
> + return 3;
> + return 0;
> +}
> +
> +static void regenerate_marks() {
> + int ret;
> + FILE *marksfile;
> + marksfile = fopen(marksfilename, "w+");
Where does marksfilename come from? Should this be
static void regenerate_marks(const char *marksfilename)
{
> + if (!marksfile)
> + die_errno("Couldn't create mark file %s.", marksfilename);
> + ret = for_each_note(NULL, 0, note2mark_cb, marksfile);
> + if (ret)
> + die("Regeneration of marks failed, returned %d.", ret);
> + fclose(marksfile);
> +}
> +
> +static void check_or_regenerate_marks(int latestrev) {
> + FILE *marksfile;
> + char *line = NULL;
> + size_t linelen = 0;
> + struct strbuf sb = STRBUF_INIT;
> + int found = 0;
> +
> + if (latestrev < 1)
> + return;
> +
> + init_notes(NULL, notes_ref, NULL, 0);
> + marksfile = fopen(marksfilename, "r");
> + if (!marksfile)
> + regenerate_marks(marksfile);
Huh? regenerate_marks() take a NULL pointer of type "FILE *"?
I think you meant something like:
init_notes(NULL, notes_ref, NULL, 0);
marksfile = fopen(marksfilename, "r");
if (!marksfile) {
regenerate_marks(marksfilename);
marksfile = fopen(marksfilename, "r");
if (!marksfile)
die("cannot read marks file!");
} else {
...
Also there is another call to regenerate_marks() without any
argument. Has this even been compile-tested?
--
To unsubscribe from this list: send the line "unsubscribe git" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html