This is the mail archive of the
mailing list for the GDB project.
Re: (patch) hpjyg15: hppah-nat.c & related
- To: Stan Shebs <shebs at cygnus dot com>
- Subject: Re: (patch) hpjyg15: hppah-nat.c & related
- From: Kevin Buettner <kevinb at cygnus dot com>
- Date: Thu, 11 Nov 1999 18:12:45 -0700
- Cc: gdb-patches at sourceware dot cygnus dot com
- References: <199911120046.QAA29304@andros.cygnus.com>
On Nov 11, 4:46pm, Stan Shebs wrote:
> I've seen some inconsistent behavior from indent, and was planning to
> investigate further one of these days. Nothing to get too excited
> about, the oddities amounting to about ~100 in 400K lines of code,
> but good to be aware of. diff -w is useful for separating out the
> "real" diffs from the uninteresting ones.
The output of indent should be idempotent wrt indent; i.e, if you run
indent on the output of a previous indent run, you should get the same
output from the second run of indent. E.g,
cp foo.c foo1.c
diff foo.c foo1.c
The diff at the end should show no changes.
Unfortunately, this is not the case. In particular, indent seems to
want to move comments around and add/delete whitespace from the ends
I used to have some patches (for a much older version of indent) that
fixed many of these problems at least for the code that I was
indenting at the time. If there is sufficient interest, I can rumage
around on my backup tapes to see if I can find these patches.