Re: BUG: `git commit` adding unstaged chunks of files to commits on 2.33.0

Andrew Thorp [off-list ref] writes:

quoted

git commit" without pathspec would have made a partial commit that

records only the changes that were added with "git add -p" to the
index, though.

Yeah, that was the expected behavior, but not what I experienced.

Sorry but it does not reproduce for me.

Here is my "reproduction recipe", designed to be followable by
anybody who has a copy of git and this piece of e-mail without
referencing anything else.

$ edit Makefile ; git diff Makefile

diff --git i/Makefile w/Makefile
index 9573190f1d..6bf8f3147b 100644
--- i/Makefile
+++ w/Makefile

@@ -1,4 +1,4 @@
-# The default target of this Makefile is...
+# The default target of this Makefile is.....
 all::
 
 # Define V=1 to have a more verbose compile.

@@ -30,7 +30,7 @@ all::
 # Define USE_LIBPCRE if you have and want to use libpcre. Various
 # commands such as log and grep offer runtime options to use
 # Perl-compatible regular expressions instead of standard or extended
-# POSIX regular expressions.
+# POSIX regular expressions...
 #
 # Only libpcre version 2 is supported. USE_LIBPCRE2 is a synonym for
 # USE_LIBPCRE, support for the old USE_LIBPCRE1 has been removed.

... You can see that we have two changes to a single file now ...

$ printf "y\nn\n" | git add -p >/dev/null
$ git status -uno
On branch test
Changes to be committed:
  (use "git restore --staged <file>..." to unstage)
	modified:   Makefile

Changes not staged for commit:
  (use "git add <file>..." to update what will be committed)
  (use "git restore <file>..." to discard changes in working directory)
	modified:   Makefile

Untracked files not listed (use -u option to show untracked files)

... And tell "add -p" to add the first but not the second hunk.

$ git commit -m test

... And commit.

$ git status -uno
On branch test
Changes not staged for commit:
  (use "git add <file>..." to update what will be committed)
  (use "git restore <file>..." to discard changes in working directory)
	modified:   Makefile

no changes added to commit (use "git add" and/or "git commit -a")

... Then the working tree file stays dirty because the second hunk
was not added.

$ git diff

diff --git i/Makefile w/Makefile
index fe6d649c27..6bf8f3147b 100644
--- i/Makefile
+++ w/Makefile

@@ -30,7 +30,7 @@ all::
 # Define USE_LIBPCRE if you have and want to use libpcre. Various
 # commands such as log and grep offer runtime options to use
 # Perl-compatible regular expressions instead of standard or extended
-# POSIX regular expressions.
+# POSIX regular expressions...
 #
 # Only libpcre version 2 is supported. USE_LIBPCRE2 is a synonym for
 # USE_LIBPCRE, support for the old USE_LIBPCRE1 has been removed.

... just as we can see here.  Of course, the first hunk is in the
commit, as can be seen ehre

$ git diff HEAD^ HEAD

diff --git a/Makefile b/Makefile
index 9573190f1d..fe6d649c27 100644
--- a/Makefile
+++ b/Makefile

@@ -1,4 +1,4 @@
-# The default target of this Makefile is...
+# The default target of this Makefile is.....
 all::
 
 # Define V=1 to have a more verbose compile.

Ah, you're right. I have a pre-commit hook which does some formatting and
then re-adds the files. Doing so adds the unstaged chunks silently.

Thanks for taking a look Junio. I'll try to reproduce on a clean repo next time.

Cheers,

Andrew Thorp