Re: [PATCH] Fix an off by one bug in pretty.c
- From
- Marco Costalba <mcostalba@gmail.com>
- Date
- Jan 5, 2008, 23:18 UTC
- Message-ID
- <e5bfff550801051518u49e9c21ake36584457f4c9488@mail.gmail.com>
- In-Reply-To
- <e5bfff550801051455i3d589c4q9375aeab0cb6aa47@mail.gmail.com>
On Jan 5, 2008 11:55 PM, Marco Costalba <mcostalba@gmail.com> wrote:
> In case author email is <> then we have the following > bug in function 'format_person_part' >
WARNING: Please discard previous patch!
Due to an error I sent the wrong one, the following is the correct one:
------------------- CUT ----------------------------------------------------
In case author email is <> then we have the following bug in function 'format_person_part'
/* parse name */
for (end = 0; end < len && msg[end] != '<'; end++)
; /* do nothing */start = end + 1; /* now start points to '>' */
-- cut ---
/* parse email */ for (end = start + 1; end < len && msg[end] != '>'; end++)
And here 'end' is initialized with 'start + 1' instead of 'start'. This turns out in empty commit date when git log is used with --pretty=format option.
Signed-off-by: Marco Costalba <mcostalba@gmail.com> ---
feba083ccd11374560e34a755ca03f07e7d45533 pretty.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/pretty.c b/pretty.c index 5b1078b..3ce5e6f 100644 --- a/pretty.c +++ b/pretty.c @@ -304,7 +304,7 @@ static void format_person_part return; /* parse email */ - for (end = start + 1; end < len && msg[end] != '>'; end++) + for (end = start; end < len && msg[end] != '>'; end++) ; /* do nothing */ if (end >= len)