Hi,
On Sun, 11 Jan 2009, Thomas Rast wrote:
> Johannes Schindelin wrote:
> >
> > But at least _I_ think it is easy to follow, and it actually makes the code
> > more readable/hackable. Correct me if I'm wrong.
>
> It indeed seems a sane approach.
Thanks.
> However, the final result segfaults and/or prints garbage (on
> apparently every commit except very small changes) when using the regex
> '\S+', which IMHO should give exactly the same result as not using a
> regex at all.
No, it should not. The correct regex is '^\S+'.
As it happens, your regex matches _anything_ + non-whitespace.
Unfortunately, this includes a newline which utterly confuses the diff,
and therefore the code that tries to get the true offsets.
Consequently, it crashes.
> Plain --color-words is not affected.
Of course, I did not change anything outside the code path of
--color-words.
Ciao,
Dscho
-- snipsnap --
[PATCH] color-words: \n must not be a part of the word.
Allowing \n as part of a word is a pilot error, but that is not a
reason for the code to crash.
Signed-off-by: Johannes Schindelin <Johannes.Schindelin@xxxxxx>
---
diff.c | 6 ++++--
1 files changed, 4 insertions(+), 2 deletions(-)
diff --git a/diff.c b/diff.c
index d6bba72..676eb79 100644
--- a/diff.c
+++ b/diff.c
@@ -381,8 +381,10 @@ static int find_word_boundary(mmfile_t *buffer, int i, regex_t *word_regex)
if (word_regex) {
regmatch_t match[1];
- if (!regexec(word_regex, buffer->ptr + i, 1, match, 0))
- i += match[0].rm_eo;
+ if (!regexec(word_regex, buffer->ptr + i, 1, match, 0)) {
+ char *p = memchr(buffer->ptr + i, '\n', match[0].rm_eo);
+ i = p ? p - buffer->ptr : match[0].rm_eo + i;
+ }
}
else
while (i < buffer->size && !isspace(buffer->ptr[i]))
--
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
