[v2,0/5] fix interactions with "-w" and "--exit-code"

Message

Junio C Hamano Aug. 17, 2023, 10:29 p.m. UTC

So here is the reworked series.  The previous attempt mostly
correspond to [4/5] of this series, but the approach has been quite
updated, so here is marked as "v2" but I omitted the interdiff as it
would be more or less useless.

The first patch is not part of the main theme, but because I'll be
adding the first use of "--dirstat" to t4015 that has been leak
sanitizer clean, and the "--dirstat" codepath has known leaks, I am
plugging the leaks as a preliminary step to avoid having to mark
t4015 as leak sanitizer unclean again.

The second patch flips around order of processing without changing
the meaning.  After this change, blocks that compute output that
also take contents into account are grouped together, before the
fallback code for output formats that do not look at the contents to
compute their result.  It is purely done as a clean-up.

The earlier series claimed that "--exit-code -w" is reliable as long
as "--patch" output is used, but it turns out that there are corner
case holes in the "--patch" codepath.  The third patch fixes them.

The fourth patch teachs "--stat" codepath to help "--exit-code -w";
as it looks at contents to produce its output, we note the fact that
we found (or did not find) differences, and use that for the exit
code computation.

The fifth patch deals with other output modes that do not look at
contents for their output to reuse the fallback code.

Junio C Hamano (5):
  diff: --dirstat leakfix
  diff: move the fallback "--exit-code" code down
  diff: mode-only change should be noticed by "--patch -w --exit-code"
  diff: teach "--stat -w --exit-code" to notice differences
  diff: teach "--name-status" and friends to honor "--exit-code -w"

 diff.c                     | 71 +++++++++++++++++++++++++++-----------
 t/t4015-diff-whitespace.sh | 36 ++++++++++++++++++-
 t/t4047-diff-dirstat.sh    |  2 ++
 3 files changed, 87 insertions(+), 22 deletions(-)

Comments

Junio C Hamano Aug. 17, 2023, 10:37 p.m. UTC | #1

Junio C Hamano <[e-mail]> writes:

> The first patch is not part of the main theme, but because I'll be
> adding the first use of "--dirstat" to t4015 that has been leak
> sanitizer clean, and the "--dirstat" codepath has known leaks, I am
> plugging the leaks as a preliminary step to avoid having to mark
> t4015 as leak sanitizer unclean again.

I completely forgot that I already have done 1e1dcb2a (Merge branch
'jc/dirstat-plug-leaks', 2023-05-15) in Git 2.41 timeframe.  I was
basing my work on Git 2.38 maintenance track and needed [1/5] but in
newer codebase it is not needed.