Skip to contentSkip to conversation

mergetool: remove unconflicted lines

PATCH In review Felipe Contreras 6 participants 80 messages Dec 23, 2020 – Mar 13, 2021 Documentation

lore
Share & tools
lore
#0 lore raw

v5 cover mergetool: remove unconflicted lines

There's not much to say other that what the commit message of the patch says.

Note: no feedback has been ignored; I replied to all the feedback, I didn't hear anything back.

Changes since v4:

* Improved commit message with suggestions from Phillip Wood.

Felipe Contreras (1):
mergetool: add automerge configuration

 Documentation/config/mergetool.txt |  3 +++
 git-mergetool.sh                   | 17 +++++++++++++++++
 t/t7610-mergetool.sh               | 18 ++++++++++++++++++
 3 files changed, 38 insertions(+)

Range-diff:
1: 776c1fbb97 ! 1: 2dc53f4dda mergetool: add automerge configuration
@@ Metadata
## Commit message ##
mergetool: add automerge configuration

- It doesn't make sense to display lines without conflicts in the
- different views of all mergetools.
+ The purpose of mergetools is to resolve conflicts when git cannot
+ automatically do so.

- Only the lines that warrant conflict markers should be displayed.
+ In order to do that git has added markers in the specific areas that
+ need resolving, which the user must manually fix. The tool is supposed
+ to help with that.

- Most people would want this behavior on, but in case some don't; add a
- new configuration: mergetool.autoMerge.
+ However, by passing the original BASE, LOCAL, and REMOTE files, many
+ changes without conflict are presented to the user when in fact nothing
+ needs to be done for those.
+
+ We can fix that by propagating the final version of the file with the
+ automatic merge to all the panes of the mergetool (BASE, LOCAL, and
+ REMOTE), and only make them differ on the places where there are actual
+ conflicts.
+
+ As most people will want the new behavior, we enable it by default.
+ Users that do not want the new behavior can set the new configuration
+ mergetool.autoMerge to false.

See Seth House's blog post [1] for the idea, and the rationale.

Signature · 1 line
2.30.0.rc1
#1 lore raw

PATCH v5 1/1 mergetool: add automerge configuration

The purpose of mergetools is to resolve conflicts when git cannot automatically do so.

In order to do that git has added markers in the specific areas that need resolving, which the user must manually fix. The tool is supposed to help with that.

However, by passing the original BASE, LOCAL, and REMOTE files, many changes without conflict are presented to the user when in fact nothing needs to be done for those.

We can fix that by propagating the final version of the file with the automatic merge to all the panes of the mergetool (BASE, LOCAL, and REMOTE), and only make them differ on the places where there are actual conflicts.

As most people will want the new behavior, we enable it by default. Users that do not want the new behavior can set the new configuration mergetool.autoMerge to false.

See Seth House's blog post [1] for the idea, and the rationale.

[1] https://www.eseth.org/2020/mergetools.html

---
 Documentation/config/mergetool.txt |  3 +++
 git-mergetool.sh                   | 17 +++++++++++++++++
 t/t7610-mergetool.sh               | 18 ++++++++++++++++++
 3 files changed, 38 insertions(+)
Documentation/config/mergetool.txt +3 -0
index 16a27443a3..7ce6d0d3ac 100644
@@ -61,3 +61,6 @@ mergetool.writeToTemp::
 
 mergetool.prompt::
 	Prompt before each invocation of the merge resolution program.
+
+mergetool.autoMerge::
+	Remove lines without conflicts from all the files. Defaults to `true`.
git-mergetool.sh +17 -0
index e3f6d543fb..f4db0cac8d 100755
@@ -239,6 +239,17 @@ checkout_staged_file () {
 	fi
 }
 
+auto_merge () {
+	git merge-file --diff3 --marker-size=7 -q -p "$LOCAL" "$BASE" "$REMOTE" >"$DIFF3"
+	if test -s "$DIFF3"
+	then
+		sed -e '/^<<<<<<< /,/^||||||| /d' -e '/^=======\r\?$/,/^>>>>>>> /d' "$DIFF3" >"$BASE"
+		sed -e '/^||||||| /,/^>>>>>>> /d' -e '/^<<<<<<< /d' "$DIFF3" >"$LOCAL"
+		sed -e '/^<<<<<<< /,/^=======\r\?$/d' -e '/^>>>>>>> /d' "$DIFF3" >"$REMOTE"
+	fi
+	rm -- "$DIFF3"
+}
+
 merge_file () {
 	MERGED="$1"
 
@@ -274,6 +285,7 @@ merge_file () {
 		BASE=${BASE##*/}
 	fi
 
+	DIFF3="$MERGETOOL_TMPDIR/${BASE}_DIFF3_$$$ext"
 	BACKUP="$MERGETOOL_TMPDIR/${BASE}_BACKUP_$$$ext"
 	LOCAL="$MERGETOOL_TMPDIR/${BASE}_LOCAL_$$$ext"
 	REMOTE="$MERGETOOL_TMPDIR/${BASE}_REMOTE_$$$ext"
@@ -322,6 +334,11 @@ merge_file () {
 	checkout_staged_file 2 "$MERGED" "$LOCAL"
 	checkout_staged_file 3 "$MERGED" "$REMOTE"
 
+	if test "$(git config --bool mergetool.autoMerge)" != "false"
+	then
+		auto_merge
+	fi
+
 	if test -z "$local_mode" || test -z "$remote_mode"
 	then
 		echo "Deleted merge conflict for '$MERGED':"
t/t7610-mergetool.sh +18 -0
index 70afdd06fa..ccabd04823 100755
@@ -828,4 +828,22 @@ test_expect_success 'mergetool -Oorder-file is honored' '
 	test_cmp expect actual
 '
 
+test_expect_success 'mergetool automerge' '
+	test_config mergetool.automerge true &&
+	test_when_finished "git reset --hard" &&
+	git checkout -b test${test_count}_b master &&
+	test_write_lines >file1 base "" a &&
+	git commit -a -m "base" &&
+	test_write_lines >file1 base "" c &&
+	git commit -a -m "remote update" &&
+	git checkout -b test${test_count}_a HEAD~ &&
+	test_write_lines >file1 local "" b &&
+	git commit -a -m "local update" &&
+	test_must_fail git merge test${test_count}_b &&
+	yes "" | git mergetool file1 &&
+	test_write_lines >expect local "" c &&
+	test_cmp expect file1 &&
+	git commit -m "test resolved with mergetool"
+'
+
 test_done
Signature · 1 line
2.30.0.rc1
#2 lore raw

Re: [PATCH v5 1/1] mergetool: add automerge configuration

git-mergetool.sh checkout_staged_file () { · line 245

Quoted 4 lines · Felipe Contreras
git-mergetool.sh +4 -0
+auto_merge () {
+	git merge-file --diff3 --marker-size=7 -q -p "$LOCAL" "$BASE" "$REMOTE" >"$DIFF3"
+	if test -s "$DIFF3"
+	then

We do not want to ignore the exit status from the command. IOW, I think the above wants to be rather

	if git merge-file ... >"$DIFF3" &&
	   test -s "$DIFF3"
	then
		...

to catch a merge-file that writes halfway and then crashes (doing the same check in different ways are probably possible, of course)

git-mergetool.sh checkout_staged_file () { · line 246

Quoted 1 line
git-mergetool.sh +1 -0
+		sed -e '/^<<<<<<< /,/^||||||| /d' -e '/^=======\r\?$/,/^>>>>>>> /d' "$DIFF3" >"$BASE"

Does everybody's sed take "\?" and interprets it as zero-or-one? POSIX uses BRE and it doesn't like \? as far as I recall, and "-E" to force ERE is a GNUism.

git-mergetool.sh checkout_staged_file () { · line 248

Quoted 2 lines
git-mergetool.sh +2 -0
+		sed -e '/^||||||| /,/^>>>>>>> /d' -e '/^<<<<<<< /d' "$DIFF3" >"$LOCAL"
+		sed -e '/^<<<<<<< /,/^=======\r\?$/d' -e '/^>>>>>>> /d' "$DIFF3" >"$REMOTE"

I'd feel safer if these resulting $BASE, $LOCAL and $REMOTE are validated to be conflict-marker free (i.e. '^\([<|=>]\)\1\1\1\1\1\1' does not appear) to make sure there was no funny virtual ancestor that records a conflicted recursive merge result confused our logic.

When we see an unfortunate sign that it happened, we can revert the automerge and let the tool handle the original input.

git-mergetool.sh checkout_staged_file () { · line 252

Quoted 4 lines
git-mergetool.sh +4 -0
+	fi
+	rm -- "$DIFF3"
+}
+

"$DIFF3" is always created (unless shell redirection into it fails), so "rm --" would be fine in practice, I guess, but "rm -f --" would not hurt.

git-mergetool.sh merge_file () { · line 288

Quoted 1 line
git-mergetool.sh +1 -0
+	DIFF3="$MERGETOOL_TMPDIR/${BASE}_DIFF3_$$$ext"

$MERGETOOL_TMPDIR is either "mktemp -d -t "git-mergetool-XXXXXX" or ".". Also, we liberally pass "$DIFF3" to "sed" as an argument and assume that the command would take it as a filename and not an option.

For the above reason, "rm --", while it is not wrong per-se, can be just a simple "rm", as there is no funny leading letters in "$DIFF3" that requires disambiguation.

#3 lore raw

Re: [PATCH v5 1/1] mergetool: add automerge configuration

Quoted 10 lines · Junio C Hamano

git-mergetool.sh checkout_staged_file () { · line 245

Quoted 4 lines · Felipe Contreras
git-mergetool.sh +4 -0
+auto_merge () {
+	git merge-file --diff3 --marker-size=7 -q -p "$LOCAL" "$BASE" "$REMOTE" >"$DIFF3"
+	if test -s "$DIFF3"
+	then

We do not want to ignore the exit status from the command. IOW, I think the above wants to be rather

Quoted 4 lines
	if git merge-file ... >"$DIFF3" &&
	   test -s "$DIFF3"
	then
		...

That doesn't work.

"git merge-file" always returns non-zero status when it succeeds (it's the number of conflicts generated).

Quoted 3 lines

git-mergetool.sh checkout_staged_file () { · line 246

Quoted 1 line
git-mergetool.sh +1 -0
+		sed -e '/^<<<<<<< /,/^||||||| /d' -e '/^=======\r\?$/,/^>>>>>>> /d' "$DIFF3" >"$BASE"

Does everybody's sed take "\?" and interprets it as zero-or-one?

I don't know.

Quoted 2 lines

POSIX uses BRE and it doesn't like \? as far as I recall, and "-E" to force ERE is a GNUism.

Another possibility is \s\*. It's less specific though.

Quoted 10 lines

git-mergetool.sh checkout_staged_file () { · line 248

Quoted 2 lines
git-mergetool.sh +2 -0
+		sed -e '/^||||||| /,/^>>>>>>> /d' -e '/^<<<<<<< /d' "$DIFF3" >"$LOCAL"
+		sed -e '/^<<<<<<< /,/^=======\r\?$/d' -e '/^>>>>>>> /d' "$DIFF3" >"$REMOTE"

I'd feel safer if these resulting $BASE, $LOCAL and $REMOTE are validated to be conflict-marker free (i.e. '^\([<|=>]\)\1\1\1\1\1\1' does not appear) to make sure there was no funny virtual ancestor that records a conflicted recursive merge result confused our logic.

When we see an unfortunate sign that it happened, we can revert the automerge and let the tool handle the original input.

What if the original file does have these markers?

Which is probably something we should be checking beforehand and not attempt an automerge in those cases.

Or we could add the --base option to "git merge-file" so we don't have to do that work by hand.

Quoted 8 lines

git-mergetool.sh checkout_staged_file () { · line 252

Quoted 4 lines
git-mergetool.sh +4 -0
+	fi
+	rm -- "$DIFF3"
+}
+

"$DIFF3" is always created (unless shell redirection into it fails), so "rm --" would be fine in practice, I guess, but "rm -f --" would not hurt.

I just did the same as below:

rm -- "$BACKUP"

Quoted 10 lines

git-mergetool.sh merge_file () { · line 288

Quoted 1 line
git-mergetool.sh +1 -0
+	DIFF3="$MERGETOOL_TMPDIR/${BASE}_DIFF3_$$$ext"

$MERGETOOL_TMPDIR is either "mktemp -d -t "git-mergetool-XXXXXX" or ".". Also, we liberally pass "$DIFF3" to "sed" as an argument and assume that the command would take it as a filename and not an option.

For the above reason, "rm --", while it is not wrong per-se, can be just a simple "rm", as there is no funny leading letters in "$DIFF3" that requires disambiguation.

Other parts of the file do this:

rm -f -- "$LOCAL" "$REMOTE" "$BASE" "$BACKUP"

I'm just following what the script already does.

Cheers.

Signature · 1 line
Felipe Contreras
#4 lore raw

Re: [PATCH v5 1/1] mergetool: add automerge configuration

Quoted 16 lines · Felipe Contreras
Quoted 10 lines · Junio C Hamano

git-mergetool.sh checkout_staged_file () { · line 245

Quoted 4 lines · Felipe Contreras
git-mergetool.sh +4 -0
+auto_merge () {
+	git merge-file --diff3 --marker-size=7 -q -p "$LOCAL" "$BASE" "$REMOTE" >"$DIFF3"
+	if test -s "$DIFF3"
+	then

We do not want to ignore the exit status from the command. IOW, I think the above wants to be rather

Quoted 4 lines
	if git merge-file ... >"$DIFF3" &&
	   test -s "$DIFF3"
	then
		...
Quoted 4 lines

That doesn't work.

"git merge-file" always returns non-zero status when it succeeds (it's the number of conflicts generated).

Ah, I forgot about that one. I think "the number of conflicts" was a UI mistake (the original that it mimics is "merge" from RCS suite, which uses 1 and 2 for "conflicts" and "trouble") but we know we will get conflicts, so it is wrong to expect success from the command. Deliberately ignoring the return status is the right thing to do.

Quoted 4 lines

What if the original file does have these markers?

Which is probably something we should be checking beforehand and not attempt an automerge in those cases.

Yes, that is a much better approach to avoid unnecessary work.

When we made the conflict marker length configurable, we were hoping that we no longer have to worry about the cases where payload files (original or ours or theirs) have lines that are confusingly similar to the conflict markers, but because we are interfacing external tools that are unaware of the facility, it probably would not help us in this case all that much.

FWIW, we use a fiarly large size for our own files in t/ and Documentation/ directories ourselves, and it does help topic branch merges somewhat frequently.

#5 lore raw

Re: [PATCH v5 1/1] mergetool: add automerge configuration

Quoted 23 lines · Junio C Hamano
Quoted 16 lines · Felipe Contreras
Quoted 10 lines · Junio C Hamano

git-mergetool.sh checkout_staged_file () { · line 245

Quoted 4 lines · Felipe Contreras
git-mergetool.sh +4 -0
+auto_merge () {
+	git merge-file --diff3 --marker-size=7 -q -p "$LOCAL" "$BASE" "$REMOTE" >"$DIFF3"
+	if test -s "$DIFF3"
+	then

We do not want to ignore the exit status from the command. IOW, I think the above wants to be rather

Quoted 4 lines
	if git merge-file ... >"$DIFF3" &&
	   test -s "$DIFF3"
	then
		...
Quoted 4 lines

That doesn't work.

"git merge-file" always returns non-zero status when it succeeds (it's the number of conflicts generated).

Quoted 6 lines

Ah, I forgot about that one. I think "the number of conflicts" was a UI mistake (the original that it mimics is "merge" from RCS suite, which uses 1 and 2 for "conflicts" and "trouble") but we know we will get conflicts, so it is wrong to expect success from the command. Deliberately ignoring the return status is the right thing to do.

I agree. My bet is that nobody is checking the return status of "git merge-file" to find out the number of conflicts. Plus, how can you check the difference between 255 conflicts and error -1?

But that's the situation we are in now.

Quoted 14 lines
Quoted 4 lines

What if the original file does have these markers?

Which is probably something we should be checking beforehand and not attempt an automerge in those cases.

Yes, that is a much better approach to avoid unnecessary work.

When we made the conflict marker length configurable, we were hoping that we no longer have to worry about the cases where payload files (original or ours or theirs) have lines that are confusingly similar to the conflict markers, but because we are interfacing external tools that are unaware of the facility, it probably would not help us in this case all that much.

Quoted 3 lines

FWIW, we use a fiarly large size for our own files in t/ and Documentation/ directories ourselves, and it does help topic branch merges somewhat frequently.

We could do something like --marker-size=13 to minimize the chances of that happening.

In that case I would prefer '/^<\{13\} /' (to avoid too many characters). I see those regexes used elsewhere in git, but I don't know how portable that is.

If we wanted to make sure none of those markers remain it's not enough to check for '^[<|=>]{13}', what follows up should be a space, or some delimiter, not another < for example. So maybe '^[<|=>]{13}[^<|=>]'?

So, do we want those three things?

1. A non-standard marker-size
2. Check beforehand the existence of those markers and disable automerge
3. Check afterwards the existence of those markers and disable automerge

Cheers.

Signature · 1 line
Felipe Contreras
#6 lore raw

Re: [PATCH v5 1/1] mergetool: add automerge configuration

Quoted 10 lines · Felipe Contreras
Quoted 6 lines

Ah, I forgot about that one. I think "the number of conflicts" was a UI mistake (the original that it mimics is "merge" from RCS suite, which uses 1 and 2 for "conflicts" and "trouble") but we know we will get conflicts, so it is wrong to expect success from the command. Deliberately ignoring the return status is the right thing to do.

I agree. My bet is that nobody is checking the return status of "git merge-file" to find out the number of conflicts. Plus, how can you check the difference between 255 conflicts and error -1?

Yup, I already mentioned UI mistake so you do not have to repeat it to consume more bandwidth. We're in agreement already.

Quoted 6 lines

We could do something like --marker-size=13 to minimize the chances of that happening.

In that case I would prefer '/^<\{13\} /' (to avoid too many characters). I see those regexes used elsewhere in git, but I don't know how portable that is.

If it is used elsewhere with "sed", then that would be OK, but if it is not with "sed" but with "grep", that's quite a different story.

Quoted 7 lines

So, do we want those three things?

1. A non-standard marker-size
2. Check beforehand the existence of those markers and disable automerge
3. Check afterwards the existence of those markers and disable automerge

I do not think 3 is needed if we do 2 and I do not think 1 would particularly be useful *UNLESS* the code consults with the attribute system to see what marker size the path uses to avoid crashing with the non-standard marker-size the path already uses.

So the easiest would be not to do anything for now, with a note about known limitations in the doc. The second easiest would be to do 2. alone. We could do 1. to be more complete but I tend to think that it is better to leave it as #leftoverbits.

#7 lore raw

Re: [PATCH v5 1/1] mergetool: add automerge configuration

Quoted 9 lines · Junio C Hamano
Quoted 7 lines · Felipe Contreras
Quoted 6 lines

Ah, I forgot about that one. I think "the number of conflicts" was a UI mistake (the original that it mimics is "merge" from RCS suite, which uses 1 and 2 for "conflicts" and "trouble") but we know we will get conflicts, so it is wrong to expect success from the command. Deliberately ignoring the return status is the right thing to do.

Quoted 5 lines
Quoted 3 lines

I agree. My bet is that nobody is checking the return status of "git merge-file" to find out the number of conflicts. Plus, how can you check the difference between 255 conflicts and error -1?

Yup, I already mentioned UI mistake so you do not have to repeat

You said it was a UI mistake, not me. I am a different mind than yours.

This [1] is the first time *you* communicated it was a UI mistake.

This [2] is the first time *I* communicated it was a UI mistake.

I communicated that fact after you, so I did not repeat anything, because I hadn't said that before. *You* did, not *me*.

Quoted 1 line

it to consume more bandwidth.

This is what is consuming bandwidth.

Not me stating *for the first time* that I agree what you just stated.

You could have skipped what I said *for the first time*, if you didn't find it particularly interesting, and that would have saved bandwidth.

Quoted 9 lines
Quoted 6 lines

We could do something like --marker-size=13 to minimize the chances of that happening.

In that case I would prefer '/^<\{13\} /' (to avoid too many characters). I see those regexes used elsewhere in git, but I don't know how portable that is.

If it is used elsewhere with "sed", then that would be OK, but if it is not with "sed" but with "grep", that's quite a different story.

In t/t3427-rebase-subtree.sh there is:

sed -e "s%\([0-9a-f]\{40\} \)files_subtree/%\1%"

Not sure if that counts. There's other places in the tests.

However, I don't see the point if the marker-size is a low enough number, like 7.

Quoted 8 lines
Quoted 7 lines

So, do we want those three things?

1. A non-standard marker-size
2. Check beforehand the existence of those markers and disable automerge
3. Check afterwards the existence of those markers and disable automerge

Quoted 4 lines

I do not think 3 is needed if we do 2 and I do not think 1 would particularly be useful *UNLESS* the code consults with the attribute system to see what marker size the path uses to avoid crashing with the non-standard marker-size the path already uses.

But what is more likely? a) That the marker-size is 7 (the default), or
b) that the marker-size is not the default, but that there's a marker-size attribute *and* the value is precisely 13?

I think a) is way more likely than b).

Quoted 4 lines

So the easiest would be not to do anything for now, with a note about known limitations in the doc. The second easiest would be to do 2. alone. We could do 1. to be more complete but I tend to think that it is better to leave it as #leftoverbits.

OK. I think 1. is low-hanging fruit, but I'm fine with not doing anything, or trying 2.

I don't think 2. would be that hard, so I will try that before re-rolling the series.

(unless somebody replies to my other pending arguments)

Cheers.

[1] https://lore.kernel.org/git/[email protected]/
[2] https://lore.kernel.org/git/[email protected]/

Signature · 1 line
Felipe Contreras
#8 lore raw

Re: [PATCH v5 1/1] mergetool: add automerge configuration

Quoted 3 lines · Felipe Contreras
Quoted 1 line

Yup, I already mentioned UI mistake so you do not have to repeat

You said it was a UI mistake, not me. I am a different mind than yours.

Yes, but the point is that I do not need to nor particularly want to hear your opinion on the behaviour of "git merge-file". I know (and others reading the thread on the list also know) that the exit code of the command is misdesigned already.

Quoted 2 lines

I communicated that fact after you, so I did not repeat anything, because I hadn't said that before. *You* did, not *me*.

Again, please realize that on list discussion is a team effort to come up together a better design of a shared solution. And if you already know that (I don't read your mind ;-), please act like you do, too.

Thanks.

#9 lore raw

Re: [PATCH v5 1/1] mergetool: add automerge configuration

Quoted 8 lines · Junio C Hamano
Quoted 3 lines · Felipe Contreras
Quoted 1 line

Yup, I already mentioned UI mistake so you do not have to repeat

You said it was a UI mistake, not me. I am a different mind than yours.

Yes, but the point is that I do not need to nor particularly want to hear your opinion on the behaviour of "git merge-file".

Then disregard the comment.

Quoted 2 lines

I know (and others reading the thread on the list also know) that the exit code of the command is misdesigned already.

Unless you can read minds, you don't know that.

And even if you do, I don't know what you know. I can't read your mind.

Plus, they can disregard the comment as well.

Quoted 5 lines
Quoted 2 lines

I communicated that fact after you, so I did not repeat anything, because I hadn't said that before. *You* did, not *me*.

Again, please realize that on list discussion is a team effort to come up together a better design of a shared solution.

Which is why agreement in a team with different minds and different viewpoints is important.

Just to show a few instances of Jeff King telling you he agrees with you:

1. "I agree it's not all that useful in that example" [1]. 16 Dec 2020
2. "I agree with the current definition" [2]. 18 Dec 2020 (same thread)
3. "I agree the two should behave the same" [3]. 18 Dec 2020 (same mail)

The fact that you value Jeff King's agreement and don't care what some other members in the community think, is a personal vale judgement, and doesn't necessarily mean the viewpoints of such community members are objectively worthless.

Cheers.

[1] https://lore.kernel.org/git/[email protected]/
[2] https://lore.kernel.org/git/X9xJ6BHM9VY0%[email protected]/
[3] https://lore.kernel.org/git/X9xJ6BHM9VY0%[email protected]/

Signature · 1 line
Felipe Contreras
#10 lore raw

Re: [PATCH v5 1/1] mergetool: add automerge configuration

Quoted 7 lines · Felipe Contreras
Quoted 6 lines · Junio C Hamano
Quoted 3 lines · Felipe Contreras
Quoted 1 line

Yup, I already mentioned UI mistake so you do not have to repeat

You said it was a UI mistake, not me. I am a different mind than yours.

Quoted 7 lines
Quoted 5 lines

Yes, but the point is that I do not need to nor particularly want to hear your opinion on the behaviour of "git merge-file".

I know (and others reading the thread on the list also know) that the exit code of the command is misdesigned already.

Unless you can read minds, you don't know that.

Actually I do, because they heard from me already ;-). If this were the case where our messages crossed, perhaps, but in this case yours was a response to my message.

Quoted 5 lines
Quoted 2 lines

Again, please realize that on list discussion is a team effort to come up together a better design of a shared solution.

Which is why agreement in a team with different minds and different viewpoints is important.

It is not like opinions on all points are important. Whether the exit code from merge-file is or is not a UI mistake does NOT have any influence on what we were discussing. My opinion is that exit code from merge-file is a UI mistake, but even if you disagree with that, that would not change the conclusion we already reached that the code should ignore its exit status, like you originally wrote.

I am already trying to ignore your opinions on things that do not matter in the context of this project, as you told me earlier ;-) But just like patches, messages are written only once but read by many people, so I'd always aim for reducing noise at the source.

Anyway, happy holidays and pleasant new year to you and to everybody.

#11 lore raw

Re: [PATCH v5 1/1] mergetool: add automerge configuration

Quoted 12 lines · Junio C Hamano
Quoted 10 lines · Felipe Contreras
Quoted 8 lines · Junio C Hamano
Quoted 3 lines · Felipe Contreras
Quoted 1 line

Yup, I already mentioned UI mistake so you do not have to repeat

You said it was a UI mistake, not me. I am a different mind than yours.

Yes, but the point is that I do not need to nor particularly want to hear your opinion on the behaviour of "git merge-file".

Quoted 6 lines
Quoted 4 lines
Quoted 2 lines

I know (and others reading the thread on the list also know) that the exit code of the command is misdesigned already.

Unless you can read minds, you don't know that.

Actually I do, because they heard from me already ;-).

They heard that you *think* it's a UI mistake.

The fact that you think something is a mistake doesn't necessarily mean it's actually a mistake, and other community members might think otherwise.

You do not dictate what others on the list know.

Quoted 9 lines
Quoted 5 lines
Quoted 2 lines

Again, please realize that on list discussion is a team effort to come up together a better design of a shared solution.

Which is why agreement in a team with different minds and different viewpoints is important.

It is not like opinions on all points are important. Whether the exit code from merge-file is or is not a UI mistake does NOT have any influence on what we were discussing.

Which is why I initially did not express such an opinion.

But you did, presumably you had some reason to do so, so I simply did the same and expressed mine.

Quoted 4 lines

I am already trying to ignore your opinions on things that do not matter in the context of this project, as you told me earlier ;-) But just like patches, messages are written only once but read by many people, so I'd always aim for reducing noise at the source.

What you consider noise others might not.

Good writers say you should not assume what your readers know.

Yes, some readers might think exactly like you do, and they don't need what you consider obvious information. But for every person that thinks exactly like you, there are dozens that don't, and it's those you should keep in mind.

Most people err on the side of not providing enough information to the minds dissimilar to theirs.

This is called the curse of knowledge [1].

I try not to do that.

Quoted 2 lines

Anyway, happy holidays and pleasant new year to you and to everybody.

Same to you.

Cheers.

[1] https://en.wikipedia.org/wiki/Curse_of_knowledge

Signature · 1 line
Felipe Contreras
#12 lore raw
In reply to Felipe Contreras
Quoted 10 lines · Felipe Contreras

There's not much to say other that what the commit message of the patch says.

Note: no feedback has been ignored; I replied to all the feedback, I didn't hear anything back.

Changes since v4:

* Improved commit message with suggestions from Phillip Wood.

Felipe Contreras (1):
mergetool: add automerge configuration

This breakage is possibly a fallout from either this patch or 1e2ae142 (t7[5-9]*: adjust the references to the default branch name "main", 2020-11-18).

https://github.com/git/git/runs/1602803804#step:7:10358

I cannot quite tell how the two strings compared with 'test' on output line 10355 are different in the output, though.

Thanks.

#13 lore raw
Quoted 10 lines · Junio C Hamano
Quoted 8 lines · Felipe Contreras

There's not much to say other that what the commit message of the patch says.

Note: no feedback has been ignored; I replied to all the feedback, I didn't hear anything back.

Changes since v4:

* Improved commit message with suggestions from Phillip Wood.

Quoted 8 lines
Quoted 2 lines

Felipe Contreras (1):
mergetool: add automerge configuration

This breakage is possibly a fallout from either this patch or 1e2ae142 (t7[5-9]*: adjust the references to the default branch name "main", 2020-11-18).

https://github.com/git/git/runs/1602803804#step:7:10358

It seems likely it's the mergetool patch.

This regex '/^=======\r\?$/' is supposed to handle the crlf situation.

I can't imagine what would be different in Windows regarding that situation.

Signature · 1 line
Felipe Contreras
#14 lore raw
In reply to Junio C Hamano

Hi Junio,

Quoted 13 lines · Junio C Hamano
Quoted 10 lines · Felipe Contreras

There's not much to say other that what the commit message of the patch says.

Note: no feedback has been ignored; I replied to all the feedback, I didn't hear anything back.

Changes since v4:

* Improved commit message with suggestions from Phillip Wood.

Felipe Contreras (1):
mergetool: add automerge configuration

Quoted 8 lines

This breakage is possibly a fallout from either this patch or 1e2ae142 (t7[5-9]*: adjust the references to the default branch name "main", 2020-11-18).

https://github.com/git/git/runs/1602803804#step:7:10358

I cannot quite tell how the two strings compared with 'test' on output line 10355 are different in the output, though.

I spent more time than I cared to spend on this, and still have not quite figured out what is the fault, but I can state with conviction that the problem is not even introduced by any merge into seen. The fc/mergetool-automerge branch itself is already broken:
https://github.com/gitgitgadget/git/actions/runs/441233234

Ciao,
Dscho

#15 lore raw
Quoted 7 lines · Johannes Schindelin
Quoted 6 lines · Junio C Hamano

This breakage is possibly a fallout from either this patch or 1e2ae142 (t7[5-9]*: adjust the references to the default branch name "main", 2020-11-18).

https://github.com/git/git/runs/1602803804#step:7:10358

Quoted 8 lines
Quoted 2 lines

I cannot quite tell how the two strings compared with 'test' on output line 10355 are different in the output, though.

I spent more time than I cared to spend on this, and still have not quite figured out what is the fault, but I can state with conviction that the problem is not even introduced by any merge into seen. The fc/mergetool-automerge branch itself is already broken:
https://github.com/gitgitgadget/git/actions/runs/441233234

Yes, if you didn't have me blocked and read what I said a week ago in [1], you would have saved yourself that time.

Seth House has claimed the patch series though.

[1] https://lore.kernel.org/git/[email protected]/

Signature · 1 line
Felipe Contreras
#16 lore raw

v6 cover mergetool: add automerge configuration

In reply to Felipe Contreras

Sorry for the slow turnaround on this. I haven't used Git via email patches before now so it took me quite a few hours to read through tutorials, configure git-send-email and fight missing Perl libs. Please let me know if I did anything incorrectly! I should be able to contribute more quickly from now on.

Changes since v5:

* Add per-tool configuration that Felipe has a "deep philosophical" opposition to adding.

Felipe Contreras (1):
mergetool: add automerge configuration

Seth House (1):
mergetool: Add per-tool support for the autoMerge flag

 Documentation/config/mergetool.txt |  6 ++++++
 git-mergetool.sh                   | 13 +++++++++++++
 t/t7610-mergetool.sh               | 18 ++++++++++++++++++
 3 files changed, 37 insertions(+)
Signature · 1 line
2.29.2
#17 lore raw

PATCH v6 1/2 mergetool: add automerge configuration

From: Felipe Contreras <[email protected]>

It doesn't make sense to display easily-solvable conflicts in the different views of all mergetools.

Only the chunks that warrant conflict markers should be displayed.

In order to unobtrusively do this, add a new configuration:
mergetool.autoMerge.

See Seth House's blog post [1] for the idea, and the rationale.

[1] https://www.eseth.org/2020/mergetools.html

---
 Documentation/config/mergetool.txt |  3 +++
 git-mergetool.sh                   | 10 ++++++++++
 t/t7610-mergetool.sh               | 18 ++++++++++++++++++
 3 files changed, 31 insertions(+)
Documentation/config/mergetool.txt +3 -0
index 16a27443a3..43af7a96f9 100644
@@ -61,3 +61,6 @@ mergetool.writeToTemp::
 
 mergetool.prompt::
 	Prompt before each invocation of the merge resolution program.
+
+mergetool.autoMerge::
+	Automatically resolve conflicts that don't require user intervention.
git-mergetool.sh +10 -0
index e3f6d543fb..6e86d3b492 100755
@@ -274,6 +274,7 @@ merge_file () {
 		BASE=${BASE##*/}
 	fi
 
+	DIFF3="$MERGETOOL_TMPDIR/${BASE}_DIFF3_$$$ext"
 	BACKUP="$MERGETOOL_TMPDIR/${BASE}_BACKUP_$$$ext"
 	LOCAL="$MERGETOOL_TMPDIR/${BASE}_LOCAL_$$$ext"
 	REMOTE="$MERGETOOL_TMPDIR/${BASE}_REMOTE_$$$ext"
@@ -322,6 +323,15 @@ merge_file () {
 	checkout_staged_file 2 "$MERGED" "$LOCAL"
 	checkout_staged_file 3 "$MERGED" "$REMOTE"
 
+	if test "$(git config --bool mergetool.autoMerge)" = "true"
+	then
+		git merge-file --diff3 -q -p "$LOCAL" "$BASE" "$REMOTE" >"$DIFF3"
+		sed -e '/^<<<<<<< /,/^||||||| /d' -e '/^=======\r\?$/,/^>>>>>>> /d' "$DIFF3" >"$BASE"
+		sed -e '/^||||||| /,/^>>>>>>> /d' -e '/^<<<<<<< /d' "$DIFF3" >"$LOCAL"
+		sed -e '/^<<<<<<< /,/^=======\r\?$/d' -e '/^>>>>>>> /d' "$DIFF3" >"$REMOTE"
+		rm -- "$DIFF3"
+	fi
+
 	if test -z "$local_mode" || test -z "$remote_mode"
 	then
 		echo "Deleted merge conflict for '$MERGED':"
t/t7610-mergetool.sh +18 -0
index 70afdd06fa..b75c91199b 100755
@@ -828,4 +828,22 @@ test_expect_success 'mergetool -Oorder-file is honored' '
 	test_cmp expect actual
 '
 
+test_expect_success 'mergetool automerge' '
+	test_config mergetool.automerge true &&
+	test_when_finished "git reset --hard" &&
+	git checkout -b test${test_count}_b master &&
+	echo -e "base\n\na" >file1 &&
+	git commit -a -m "base" &&
+	echo -e "base\n\nc" >file1 &&
+	git commit -a -m "remote update" &&
+	git checkout -b test${test_count}_a HEAD~ &&
+	echo -e "local\n\nb" >file1 &&
+	git commit -a -m "local update" &&
+	test_must_fail git merge test${test_count}_b &&
+	yes "" | git mergetool file1 &&
+	echo -e "local\n\nc" >expect &&
+	test_cmp expect file1 &&
+	git commit -m "test resolved with mergetool"
+'
+
 test_done
Signature · 1 line
2.29.2
#18 lore raw

Re: [PATCH v6 1/2] mergetool: add automerge configuration

Quoted 8 lines · Seth House

...
See Seth House's blog post [1] for the idea, and the rationale.

[1] https://www.eseth.org/2020/mergetools.html

---

Missing Sign-off as a relayer.

t/t7610-mergetool.sh test_expect_success 'mergetool -Oorder-file is honored' ' · line 835

Quoted 13 lines
t/t7610-mergetool.sh +5 -0
index 70afdd06fa..b75c91199b 100755
@@ -828,4 +828,22 @@ test_expect_success 'mergetool -Oorder-file is honored' '
 	test_cmp expect actual
 '
 
+test_expect_success 'mergetool automerge' '
+	test_config mergetool.automerge true &&
+	test_when_finished "git reset --hard" &&
+	git checkout -b test${test_count}_b master &&
+	echo -e "base\n\na" >file1 &&

These do not seem to be taken from the version that has been improved by reviwer comments after v3.

Quoted 14 lines
t/t7610-mergetool.sh +13 -0
+	git commit -a -m "base" &&
+	echo -e "base\n\nc" >file1 &&
+	git commit -a -m "remote update" &&
+	git checkout -b test${test_count}_a HEAD~ &&
+	echo -e "local\n\nb" >file1 &&
+	git commit -a -m "local update" &&
+	test_must_fail git merge test${test_count}_b &&
+	yes "" | git mergetool file1 &&
+	echo -e "local\n\nc" >expect &&
+	test_cmp expect file1 &&
+	git commit -m "test resolved with mergetool"
+'
+
 test_done
#19 lore raw

Re: [PATCH v6 1/2] mergetool: add automerge configuration

Quoted 1 line · Junio C Hamano

Missing Sign-off as a relayer.

I haven't come across that in the docs on contributing to Git and my Google searches aren't helping. Do you mind pointing me to what to add?

Quoted 2 lines

These do not seem to be taken from the version that has been improved by reviwer comments after v3.

Whoops! Thanks for the catch. Seems I fished the wrong version out of my email. Created a new v7 based off the correct v5.

#20 lore raw

Re: [PATCH v6 1/2] mergetool: add automerge configuration

Quoted 5 lines · Seth House
Quoted 1 line · Junio C Hamano

Missing Sign-off as a relayer.

I haven't come across that in the docs on contributing to Git and my Google searches aren't helping. Do you mind pointing me to what to add?

Documentation/SubmittingPatches#sign-off

=== Certify your work by adding your Signed-off-by trailer

To improve tracking of who did what, we ask you to certify that you wrote the patch or have the right to pass it on under the same license as ours, by "signing off" your patch. Without sign-off, we cannot accept your patches.

If you can certify the below D-C-O:

[[dco]]
.Developer's Certificate of Origin 1.1
____
By making a contribution to this project, I certify that:

a. The contribution was created in whole or in part by me and I
have the right to submit it under the open source license indicated in the file; or

b. The contribution is based upon previous work that, to the best
of my knowledge, is covered under an appropriate open source license and I have the right under that license to submit that work with modifications, whether created in whole or in part by me, under the same open source license (unless I am permitted to submit under a different license), as indicated in the file; or

c. The contribution was provided directly to me by some other
person who certified (a), (b) or (c) and I have not modified it.

d. I understand and agree that this project and the contribution
are public and that a record of the contribution (including all personal information I submit with it, including my sign-off) is maintained indefinitely and may be redistributed consistent with this project or the open source license(s) involved.
____

you add a "Signed-off-by" trailer to your commit, that looks like this:

....

....

So, you'd add your own signed-off-by trailer at the end of the trailer list.

https://lore.kernel.org/git/[email protected]/

for an example where Johannes Schindelin picked up a patch written by Dennis Ameling and relayed it to the list.

Thanks.

#21 lore raw

PATCH v6 2/2 mergetool: Add per-tool support for the autoMerge flag

In reply to Seth House

Keep the global mergetool flag and add a per-tool override flag so that users may enable the flag for one tool and disable it for another.

---
 Documentation/config/mergetool.txt | 3 +++
 git-mergetool.sh                   | 5 ++++-
 2 files changed, 7 insertions(+), 1 deletion(-)
Documentation/config/mergetool.txt +3 -0
index 43af7a96f9..7f32281a61 100644
@@ -21,6 +21,9 @@ mergetool.<tool>.trustExitCode::
 	if the file has been updated, otherwise the user is prompted to
 	indicate the success of the merge.
 
+mergetool.<tool>.autoMerge::
+	Automatically resolve conflicts that don't require user intervention.
+
 mergetool.meld.hasOutput::
 	Older versions of `meld` do not support the `--output` option.
 	Git will attempt to detect whether `meld` supports `--output`
git-mergetool.sh +4 -1
index 6e86d3b492..81df301734 100755
@@ -323,7 +323,10 @@ merge_file () {
 	checkout_staged_file 2 "$MERGED" "$LOCAL"
 	checkout_staged_file 3 "$MERGED" "$REMOTE"
 
-	if test "$(git config --bool mergetool.autoMerge)" = "true"
+	if test "$(
+		git config --get --bool "mergetool.$merge_tool.automerge" ||
+		git config --get --bool "mergetool.automerge" ||
+		echo true)" = true
 	then
 		git merge-file --diff3 -q -p "$LOCAL" "$BASE" "$REMOTE" >"$DIFF3"
 		sed -e '/^<<<<<<< /,/^||||||| /d' -e '/^=======\r\?$/,/^>>>>>>> /d' "$DIFF3" >"$BASE"
Signature · 1 line
2.29.2
#22 lore raw

Re: [PATCH v6 2/2] mergetool: Add per-tool support for the autoMerge flag

Quoted 28 lines · Seth House

Keep the global mergetool flag and add a per-tool override flag so that users may enable the flag for one tool and disable it for another.

---
 Documentation/config/mergetool.txt | 3 +++
 git-mergetool.sh                   | 5 ++++-
 2 files changed, 7 insertions(+), 1 deletion(-)
Documentation/config/mergetool.txt +3 -0
index 43af7a96f9..7f32281a61 100644
@@ -21,6 +21,9 @@ mergetool.<tool>.trustExitCode::
 	if the file has been updated, otherwise the user is prompted to
 	indicate the success of the merge.
 
+mergetool.<tool>.autoMerge::
+	Automatically resolve conflicts that don't require user intervention.
+
 mergetool.meld.hasOutput::
 	Older versions of `meld` do not support the `--output` option.
 	Git will attempt to detect whether `meld` supports `--output`
git-mergetool.sh +0 -0
index 6e86d3b492..81df301734 100755

git-mergetool.sh merge_file () { · line 329

Quoted 8 lines
git-mergetool.sh +4 -1
@@ -323,7 +323,10 @@ merge_file () {
 	checkout_staged_file 2 "$MERGED" "$LOCAL"
 	checkout_staged_file 3 "$MERGED" "$REMOTE"
 
-	if test "$(git config --bool mergetool.autoMerge)" = "true"
+	if test "$(
+		git config --get --bool "mergetool.$merge_tool.automerge" ||
+		git config --get --bool "mergetool.automerge" ||
+		echo true)" = true

Your [v6 1/2] that you build this step on does not enable the feature by default, but this step does; it deserves to be documented and mentioned in the proposed log message.

But I think you'd want to build this step on top of newer one, if only to take the portability fix to the tests, and that patch enables the feature by default, so ...

git-mergetool.sh merge_file () { · line 332

Quoted 3 lines
git-mergetool.sh +0 -0
 	then
 		git merge-file --diff3 -q -p "$LOCAL" "$BASE" "$REMOTE" >"$DIFF3"
 		sed -e '/^<<<<<<< /,/^||||||| /d' -e '/^=======\r\?$/,/^>>>>>>> /d' "$DIFF3" >"$BASE"

Thanks.

#23 lore raw

v7 cover mergetool: add automerge configuration

In reply to Seth House

Changes since v6:

* Incorporated Junio's help and advice:

* Rebased v7 off of the correct v5 version.
* Signed off on Felipe's commit. (Although I have minor qualms with Felipe's various wording and even the name of the flag it is decidedly not worth burdening the list with bike-shedding.)

Changes since v5:

* Add per-tool configuration that Felipe has a "deep philosophical" opposition to adding.

Felipe Contreras (1):
mergetool: add automerge configuration

Seth House (1):
mergetool: Add per-tool support for the autoMerge flag

 Documentation/config/mergetool.txt |  6 ++++++
 git-mergetool.sh                   | 20 ++++++++++++++++++++
 t/t7610-mergetool.sh               | 18 ++++++++++++++++++
 3 files changed, 44 insertions(+)
Signature · 1 line
2.29.2
#24 lore raw

PATCH v7 1/2 mergetool: add automerge configuration

From: Felipe Contreras <[email protected]>

The purpose of mergetools is to resolve conflicts when git cannot automatically do so.

In order to do that git has added markers in the specific areas that need resolving, which the user must manually fix. The tool is supposed to help with that.

However, by passing the original BASE, LOCAL, and REMOTE files, many changes without conflict are presented to the user when in fact nothing needs to be done for those.

We can fix that by propagating the final version of the file with the automatic merge to all the panes of the mergetool (BASE, LOCAL, and REMOTE), and only make them differ on the places where there are actual conflicts.

As most people will want the new behavior, we enable it by default. Users that do not want the new behavior can set the new configuration mergetool.autoMerge to false.

See Seth House's blog post [1] for the idea, and the rationale.

[1] https://www.eseth.org/2020/mergetools.html

---
 Documentation/config/mergetool.txt |  3 +++
 git-mergetool.sh                   | 17 +++++++++++++++++
 t/t7610-mergetool.sh               | 18 ++++++++++++++++++
 3 files changed, 38 insertions(+)
Documentation/config/mergetool.txt +3 -0
index 16a27443a3..7ce6d0d3ac 100644
@@ -61,3 +61,6 @@ mergetool.writeToTemp::
 
 mergetool.prompt::
 	Prompt before each invocation of the merge resolution program.
+
+mergetool.autoMerge::
+	Remove lines without conflicts from all the files. Defaults to `true`.
git-mergetool.sh +17 -0
index e3f6d543fb..f4db0cac8d 100755
@@ -239,6 +239,17 @@ checkout_staged_file () {
 	fi
 }
 
+auto_merge () {
+	git merge-file --diff3 --marker-size=7 -q -p "$LOCAL" "$BASE" "$REMOTE" >"$DIFF3"
+	if test -s "$DIFF3"
+	then
+		sed -e '/^<<<<<<< /,/^||||||| /d' -e '/^=======\r\?$/,/^>>>>>>> /d' "$DIFF3" >"$BASE"
+		sed -e '/^||||||| /,/^>>>>>>> /d' -e '/^<<<<<<< /d' "$DIFF3" >"$LOCAL"
+		sed -e '/^<<<<<<< /,/^=======\r\?$/d' -e '/^>>>>>>> /d' "$DIFF3" >"$REMOTE"
+	fi
+	rm -- "$DIFF3"
+}
+
 merge_file () {
 	MERGED="$1"
 
@@ -274,6 +285,7 @@ merge_file () {
 		BASE=${BASE##*/}
 	fi
 
+	DIFF3="$MERGETOOL_TMPDIR/${BASE}_DIFF3_$$$ext"
 	BACKUP="$MERGETOOL_TMPDIR/${BASE}_BACKUP_$$$ext"
 	LOCAL="$MERGETOOL_TMPDIR/${BASE}_LOCAL_$$$ext"
 	REMOTE="$MERGETOOL_TMPDIR/${BASE}_REMOTE_$$$ext"
@@ -322,6 +334,11 @@ merge_file () {
 	checkout_staged_file 2 "$MERGED" "$LOCAL"
 	checkout_staged_file 3 "$MERGED" "$REMOTE"
 
+	if test "$(git config --bool mergetool.autoMerge)" != "false"
+	then
+		auto_merge
+	fi
+
 	if test -z "$local_mode" || test -z "$remote_mode"
 	then
 		echo "Deleted merge conflict for '$MERGED':"
t/t7610-mergetool.sh +18 -0
index 70afdd06fa..ccabd04823 100755
@@ -828,4 +828,22 @@ test_expect_success 'mergetool -Oorder-file is honored' '
 	test_cmp expect actual
 '
 
+test_expect_success 'mergetool automerge' '
+	test_config mergetool.automerge true &&
+	test_when_finished "git reset --hard" &&
+	git checkout -b test${test_count}_b master &&
+	test_write_lines >file1 base "" a &&
+	git commit -a -m "base" &&
+	test_write_lines >file1 base "" c &&
+	git commit -a -m "remote update" &&
+	git checkout -b test${test_count}_a HEAD~ &&
+	test_write_lines >file1 local "" b &&
+	git commit -a -m "local update" &&
+	test_must_fail git merge test${test_count}_b &&
+	yes "" | git mergetool file1 &&
+	test_write_lines >expect local "" c &&
+	test_cmp expect file1 &&
+	git commit -m "test resolved with mergetool"
+'
+
 test_done
Signature · 1 line
2.29.2
#25 lore raw

PATCH v7 2/2 mergetool: Add per-tool support for the autoMerge flag

In reply to Seth House

Keep the global mergetool flag and add a per-tool override flag so that users may enable the flag for one tool and disable it for another.

---
 Documentation/config/mergetool.txt | 3 +++
 git-mergetool.sh                   | 5 ++++-
 2 files changed, 7 insertions(+), 1 deletion(-)
Documentation/config/mergetool.txt +3 -0
index 7ce6d0d3ac..ef147fc118 100644
@@ -21,6 +21,9 @@ mergetool.<tool>.trustExitCode::
 	if the file has been updated, otherwise the user is prompted to
 	indicate the success of the merge.
 
+mergetool.<tool>.autoMerge::
+	Remove lines without conflicts from all the files. Defaults to `true`.
+
 mergetool.meld.hasOutput::
 	Older versions of `meld` do not support the `--output` option.
 	Git will attempt to detect whether `meld` supports `--output`
git-mergetool.sh +4 -1
index f4db0cac8d..e3c7d78d1d 100755
@@ -334,7 +334,10 @@ merge_file () {
 	checkout_staged_file 2 "$MERGED" "$LOCAL"
 	checkout_staged_file 3 "$MERGED" "$REMOTE"
 
-	if test "$(git config --bool mergetool.autoMerge)" != "false"
+	if test "$(
+		git config --get --bool "mergetool.$merge_tool.automerge" ||
+		git config --get --bool "mergetool.automerge" ||
+		echo true)" = true
 	then
 		auto_merge
 	fi
Signature · 1 line
2.29.2
#26 lore raw

RE: [PATCH v7 2/2] mergetool: Add per-tool support for the autoMerge flag

git-mergetool.sh merge_file () { · line 340

Quoted 13 lines · Seth House
git-mergetool.sh +4 -1
index f4db0cac8d..e3c7d78d1d 100755
@@ -334,7 +334,10 @@ merge_file () {
 	checkout_staged_file 2 "$MERGED" "$LOCAL"
 	checkout_staged_file 3 "$MERGED" "$REMOTE"
 
-	if test "$(git config --bool mergetool.autoMerge)" != "false"
+	if test "$(
+		git config --get --bool "mergetool.$merge_tool.automerge" ||
+		git config --get --bool "mergetool.automerge" ||
+		echo true)" = true

This is a per-tool user configuration.

Wasn't your argument that some tools would want to disable this flag? That is; the tool, not the user.

For example, the author of diffconflicts might want to disable this flag for all its users, or at least disable it by default.

How can the winmerge difftool disable this flag?

Quoted 3 lines
git-mergetool.sh +0 -0
 	then
 		auto_merge
 	fi
Signature · 1 line
Felipe Contreras
#27 lore raw

v8 cover mergetool: add automerge configuration

In reply to Seth House

Changes since v7:

* Add a tool-specific override function to setup_tool based on Junio's original patch feedback.

The implementation of initialize_merge_tool is very much not set in stone. Suggestions are very welcome for alternate approaches that are less invasive.

Felipe Contreras (1):
mergetool: add automerge configuration

Seth House (3):
mergetool: Add per-tool support for the autoMerge flag
mergetool: Break setup_tool out into separate initialization function
mergetool: Add automerge_enabled tool-specific override function

 Documentation/config/mergetool.txt   |  6 ++++++
 Documentation/git-mergetool--lib.txt |  4 ++++
 git-difftool--helper.sh              |  2 ++
 git-mergetool--lib.sh                | 11 ++++++++---
 git-mergetool.sh                     | 22 ++++++++++++++++++++++
 t/t7610-mergetool.sh                 | 18 ++++++++++++++++++
 6 files changed, 60 insertions(+), 3 deletions(-)
Signature · 1 line
2.29.2
#28 lore raw

PATCH v8 1/4 mergetool: add automerge configuration

From: Felipe Contreras <[email protected]>

The purpose of mergetools is to resolve conflicts when git cannot automatically do so.

In order to do that git has added markers in the specific areas that need resolving, which the user must manually fix. The tool is supposed to help with that.

However, by passing the original BASE, LOCAL, and REMOTE files, many changes without conflict are presented to the user when in fact nothing needs to be done for those.

We can fix that by propagating the final version of the file with the automatic merge to all the panes of the mergetool (BASE, LOCAL, and REMOTE), and only make them differ on the places where there are actual conflicts.

As most people will want the new behavior, we enable it by default. Users that do not want the new behavior can set the new configuration mergetool.autoMerge to false.

See Seth House's blog post [1] for the idea, and the rationale.

[1] https://www.eseth.org/2020/mergetools.html

---
 Documentation/config/mergetool.txt |  3 +++
 git-mergetool.sh                   | 17 +++++++++++++++++
 t/t7610-mergetool.sh               | 18 ++++++++++++++++++
 3 files changed, 38 insertions(+)
Documentation/config/mergetool.txt +3 -0
index 16a27443a3..7ce6d0d3ac 100644
@@ -61,3 +61,6 @@ mergetool.writeToTemp::
 
 mergetool.prompt::
 	Prompt before each invocation of the merge resolution program.
+
+mergetool.autoMerge::
+	Remove lines without conflicts from all the files. Defaults to `true`.
git-mergetool.sh +17 -0
index e3f6d543fb..f4db0cac8d 100755
@@ -239,6 +239,17 @@ checkout_staged_file () {
 	fi
 }
 
+auto_merge () {
+	git merge-file --diff3 --marker-size=7 -q -p "$LOCAL" "$BASE" "$REMOTE" >"$DIFF3"
+	if test -s "$DIFF3"
+	then
+		sed -e '/^<<<<<<< /,/^||||||| /d' -e '/^=======\r\?$/,/^>>>>>>> /d' "$DIFF3" >"$BASE"
+		sed -e '/^||||||| /,/^>>>>>>> /d' -e '/^<<<<<<< /d' "$DIFF3" >"$LOCAL"
+		sed -e '/^<<<<<<< /,/^=======\r\?$/d' -e '/^>>>>>>> /d' "$DIFF3" >"$REMOTE"
+	fi
+	rm -- "$DIFF3"
+}
+
 merge_file () {
 	MERGED="$1"
 
@@ -274,6 +285,7 @@ merge_file () {
 		BASE=${BASE##*/}
 	fi
 
+	DIFF3="$MERGETOOL_TMPDIR/${BASE}_DIFF3_$$$ext"
 	BACKUP="$MERGETOOL_TMPDIR/${BASE}_BACKUP_$$$ext"
 	LOCAL="$MERGETOOL_TMPDIR/${BASE}_LOCAL_$$$ext"
 	REMOTE="$MERGETOOL_TMPDIR/${BASE}_REMOTE_$$$ext"
@@ -322,6 +334,11 @@ merge_file () {
 	checkout_staged_file 2 "$MERGED" "$LOCAL"
 	checkout_staged_file 3 "$MERGED" "$REMOTE"
 
+	if test "$(git config --bool mergetool.autoMerge)" != "false"
+	then
+		auto_merge
+	fi
+
 	if test -z "$local_mode" || test -z "$remote_mode"
 	then
 		echo "Deleted merge conflict for '$MERGED':"
t/t7610-mergetool.sh +18 -0
index 70afdd06fa..ccabd04823 100755
@@ -828,4 +828,22 @@ test_expect_success 'mergetool -Oorder-file is honored' '
 	test_cmp expect actual
 '
 
+test_expect_success 'mergetool automerge' '
+	test_config mergetool.automerge true &&
+	test_when_finished "git reset --hard" &&
+	git checkout -b test${test_count}_b master &&
+	test_write_lines >file1 base "" a &&
+	git commit -a -m "base" &&
+	test_write_lines >file1 base "" c &&
+	git commit -a -m "remote update" &&
+	git checkout -b test${test_count}_a HEAD~ &&
+	test_write_lines >file1 local "" b &&
+	git commit -a -m "local update" &&
+	test_must_fail git merge test${test_count}_b &&
+	yes "" | git mergetool file1 &&
+	test_write_lines >expect local "" c &&
+	test_cmp expect file1 &&
+	git commit -m "test resolved with mergetool"
+'
+
 test_done
Signature · 1 line
2.29.2
#29 lore raw

Re: [PATCH v8 1/4] mergetool: add automerge configuration

Am 28.12.20 um 05:54 schrieb Seth House:

Documentation/config/mergetool.txt mergetool.writeToTemp:: · line 66

Quoted 11 lines
Documentation/config/mergetool.txt +3 -0
index 16a27443a3..7ce6d0d3ac 100644
@@ -61,3 +61,6 @@ mergetool.writeToTemp::
 
 mergetool.prompt::
 	Prompt before each invocation of the merge resolution program.
+
+mergetool.autoMerge::
+	Remove lines without conflicts from all the files. Defaults to `true`.

This text, starting with "Remove lines", sounds alarming. Isn't it more along the lines of:

Consolidate non-conflicting parts, so that only conflicting parts are presented to the merge tool. Defaults to true.

It would be great to keep the list of config entries sorted. I know that mergetool.prompt is not at the correct location, but that shouldn't be an excuse to make the situation worse.

-- Hannes

#30 lore raw

PATCH v8 2/4 mergetool: Add per-tool support for the autoMerge flag

In reply to Seth House

Keep the global mergetool flag and add a per-tool override flag so that users may enable the flag for one tool and disable it for another.

---
 Documentation/config/mergetool.txt | 3 +++
 git-mergetool.sh                   | 5 ++++-
 2 files changed, 7 insertions(+), 1 deletion(-)
Documentation/config/mergetool.txt +3 -0
index 7ce6d0d3ac..ef147fc118 100644
@@ -21,6 +21,9 @@ mergetool.<tool>.trustExitCode::
 	if the file has been updated, otherwise the user is prompted to
 	indicate the success of the merge.
 
+mergetool.<tool>.autoMerge::
+	Remove lines without conflicts from all the files. Defaults to `true`.
+
 mergetool.meld.hasOutput::
 	Older versions of `meld` do not support the `--output` option.
 	Git will attempt to detect whether `meld` supports `--output`
git-mergetool.sh +4 -1
index f4db0cac8d..e3c7d78d1d 100755
@@ -334,7 +334,10 @@ merge_file () {
 	checkout_staged_file 2 "$MERGED" "$LOCAL"
 	checkout_staged_file 3 "$MERGED" "$REMOTE"
 
-	if test "$(git config --bool mergetool.autoMerge)" != "false"
+	if test "$(
+		git config --get --bool "mergetool.$merge_tool.automerge" ||
+		git config --get --bool "mergetool.automerge" ||
+		echo true)" = true
 	then
 		auto_merge
 	fi
Signature · 1 line
2.29.2
#31 lore raw

Re: [PATCH v8 2/4] mergetool: Add per-tool support for the autoMerge flag

Quoted 2 lines

Subject: Re: [PATCH v8 2/4] mergetool: Add per-tool support for the autoMerge flag

"git shortlog --no-merges --since=2.months" may tell you this but our convention is not to capitalize the word after "<area>:" on the title.

Quoted 14 lines

Keep the global mergetool flag and add a per-tool override flag so that users may enable the flag for one tool and disable it for another.

---
 Documentation/config/mergetool.txt | 3 +++
 git-mergetool.sh                   | 5 ++++-
 2 files changed, 7 insertions(+), 1 deletion(-)
Documentation/config/mergetool.txt +0 -0
index 7ce6d0d3ac..ef147fc118 100644

Documentation/config/mergetool.txt mergetool.<tool>.trustExitCode:: · line 26

Quoted 6 lines
Documentation/config/mergetool.txt +3 -0
@@ -21,6 +21,9 @@ mergetool.<tool>.trustExitCode::
 	if the file has been updated, otherwise the user is prompted to
 	indicate the success of the merge.
 
+mergetool.<tool>.autoMerge::
+	Remove lines without conflicts from all the files. Defaults to `true`.
+

This entry needs to mention how it relates to the big red button mergetool.autoMerge and vice versa. E.g.

mergetool.autoMerge::
- Remove lines without conflicts from all the files. Defaults to true.
+ Remove lines without conflicts from all the files. Can be
+ overriden per-tool via mergetool.<tool>.autoMerge configuration
+ variable. Defaults to true.

would be a good update to the documentation introduced by the previous step. It is somewhat misleading for the per-tool entry added in this patch to say "Defaults to true", as the value of the big red button configuration would be the real default.

Remove ... the files when the mergetool '<tool>' is in use. See also mergetool.autoMerge.

or something like that, perhaps.

#32 lore raw

PATCH v8 3/4 mergetool: Break setup_tool out into separate initialization function

In reply to Seth House

The tool-specific functions are sometimes needed in scope earlier than when run_merge_tool is called.

---
 Documentation/git-mergetool--lib.txt | 4 ++++
 git-difftool--helper.sh              | 2 ++
 git-mergetool--lib.sh                | 7 ++++---
 git-mergetool.sh                     | 2 ++
 4 files changed, 12 insertions(+), 3 deletions(-)
Documentation/git-mergetool--lib.txt +4 -0
index 4da9d24096..3e8f59ac0e 100644
@@ -38,6 +38,10 @@ get_merge_tool_cmd::
 get_merge_tool_path::
 	returns the custom path for a merge tool.
 
+initialize_merge_tool::
+	bring merge tool specific functions into scope so they can be used or
+	overridden.
+
 run_merge_tool::
 	launches a merge tool given the tool name and a true/false
 	flag to indicate whether a merge base is present.
git-difftool--helper.sh +2 -0
index 46af3e60b7..c47a6d4253 100755
@@ -61,6 +61,7 @@ launch_merge_tool () {
 		export BASE
 		eval $GIT_DIFFTOOL_EXTCMD '"$LOCAL"' '"$REMOTE"'
 	else
+		initialize_merge_tool "$merge_tool"
 		run_merge_tool "$merge_tool"
 	fi
 }
@@ -79,6 +80,7 @@ if test -n "$GIT_DIFFTOOL_DIRDIFF"
 then
 	LOCAL="$1"
 	REMOTE="$2"
+	initialize_merge_tool "$merge_tool"
 	run_merge_tool "$merge_tool" false
 else
 	# Launch the merge tool on each path provided by 'git diff'
git-mergetool--lib.sh +4 -3
index 7225abd811..e059b3559e 100644
@@ -248,6 +248,10 @@ trust_exit_code () {
 	fi
 }
 
+initialize_merge_tool () {
+	# Bring tool-specific functions into scope
+	setup_tool "$1" || return 1
+}
 
 # Entry point for running tools
 run_merge_tool () {
@@ -259,9 +263,6 @@ run_merge_tool () {
 	merge_tool_path=$(get_merge_tool_path "$1") || exit
 	base_present="$2"
 
-	# Bring tool-specific functions into scope
-	setup_tool "$1" || return 1
-
 	if merge_mode
 	then
 		run_merge_cmd "$1"
git-mergetool.sh +2 -0
index e3c7d78d1d..929192d0f8 100755
@@ -334,6 +334,8 @@ merge_file () {
 	checkout_staged_file 2 "$MERGED" "$LOCAL"
 	checkout_staged_file 3 "$MERGED" "$REMOTE"
 
+	initialize_merge_tool "$merge_tool"
+
 	if test "$(
 		git config --get --bool "mergetool.$merge_tool.automerge" ||
 		git config --get --bool "mergetool.automerge" ||
Signature · 1 line
2.29.2
#33 lore raw

Re: [PATCH v8 3/4] mergetool: Break setup_tool out into separate initialization function

Am 28.12.20 um 05:54 schrieb Seth House:

Quoted 2 lines

The tool-specific functions are sometimes needed in scope earlier than when run_merge_tool is called.

You should answer why this change is needed. "are sometimes needed in scope earlier" cannot be true, else we would have a bug that is fixed by this change. But this isn't a bug fix, is it?

It would be ok to say "We are going to add another thing that we will need before run_merge_tool; this is a preparation" or something.

Which brings me to another point: I do not see that something is added to initialize_merge_tool in a later patch. You are only replacing setup_tool calls by initialize_merge_tool, which forwards to setup_tool. Why do we need a new function?

Quoted 14 lines
---
 Documentation/git-mergetool--lib.txt | 4 ++++
 git-difftool--helper.sh              | 2 ++
 git-mergetool--lib.sh                | 7 ++++---
 git-mergetool.sh                     | 2 ++
 4 files changed, 12 insertions(+), 3 deletions(-)
Documentation/git-mergetool--lib.txt +0 -0
index 4da9d24096..3e8f59ac0e 100644

Documentation/git-mergetool--lib.txt get_merge_tool_cmd:: · line 47

Quoted 10 lines
Documentation/git-mergetool--lib.txt +4 -0
@@ -38,6 +38,10 @@ get_merge_tool_cmd::
 get_merge_tool_path::
 	returns the custom path for a merge tool.
 
+initialize_merge_tool::
+	bring merge tool specific functions into scope so they can be used or
+	overridden.
+
 run_merge_tool::
 	launches a merge tool given the tool name and a true/false
 	flag to indicate whether a merge base is present.

[ swapped hunks for better sentence structure ]

Quoted 16 lines
git-mergetool--lib.sh +4 -0
index 7225abd811..e059b3559e 100644
@@ -248,6 +248,10 @@ trust_exit_code () {
 	fi
 }
 
+initialize_merge_tool () {
+	# Bring tool-specific functions into scope
+	setup_tool "$1" || return 1
+}
 
 # Entry point for running tools
 run_merge_tool () {

git-mergetool--lib.sh run_merge_tool () { · line 264

Quoted 6 lines
git-mergetool--lib.sh +0 -3
@@ -259,9 +263,6 @@ run_merge_tool () {
 	merge_tool_path=$(get_merge_tool_path "$1") || exit
 	base_present="$2"
 
-	# Bring tool-specific functions into scope
-	setup_tool "$1" || return 1
-

Before this change, run_merge_tool would exit here when there was an error during setup_tool. But...

git-difftool--helper.sh launch_merge_tool () { · line 66

Quoted 11 lines
git-difftool--helper.sh +1 -0
index 46af3e60b7..c47a6d4253 100755
@@ -61,6 +61,7 @@ launch_merge_tool () {
 		export BASE
 		eval $GIT_DIFFTOOL_EXTCMD '"$LOCAL"' '"$REMOTE"'
 	else
+		initialize_merge_tool "$merge_tool"
 		run_merge_tool "$merge_tool"
 	fi

... after the change we do not exit anymore. Does it matter?

Perhaps

		initialize_merge_tool "$merge_tool" &&

run_merge_tool "$merge_tool"

Quoted 25 lines
git-difftool--helper.sh +1 -0
 }
@@ -79,6 +80,7 @@ if test -n "$GIT_DIFFTOOL_DIRDIFF"
 then
 	LOCAL="$1"
 	REMOTE="$2"
+	initialize_merge_tool "$merge_tool"
 	run_merge_tool "$merge_tool" false
 else
 	# Launch the merge tool on each path provided by 'git diff'
git-mergetool.sh +2 -0
index e3c7d78d1d..929192d0f8 100755
@@ -334,6 +334,8 @@ merge_file () {
 	checkout_staged_file 2 "$MERGED" "$LOCAL"
 	checkout_staged_file 3 "$MERGED" "$REMOTE"
 
+	initialize_merge_tool "$merge_tool"
+
 	if test "$(
 		git config --get --bool "mergetool.$merge_tool.automerge" ||
 		git config --get --bool "mergetool.automerge" ||
#34 lore raw

PATCH v8 4/4 mergetool: Add automerge_enabled tool-specific override function

In reply to Seth House

Hat-tip to Junio C Hamano for the implementation.

---
 git-mergetool--lib.sh | 4 ++++
 git-mergetool.sh      | 2 +-
 2 files changed, 5 insertions(+), 1 deletion(-)
git-mergetool--lib.sh +4 -0
index e059b3559e..5084ceffeb 100644
@@ -164,6 +164,10 @@ setup_tool () {
 		return 1
 	}
 
+	automerge_enabled () {
+		true
+	}
+
 	translate_merge_tool_path () {
 		echo "$1"
 	}
git-mergetool.sh +1 -1
index 929192d0f8..a44afd3822 100755
@@ -336,7 +336,7 @@ merge_file () {
 
 	initialize_merge_tool "$merge_tool"
 
-	if test "$(
+	if automerge_enabled && test "$(
 		git config --get --bool "mergetool.$merge_tool.automerge" ||
 		git config --get --bool "mergetool.automerge" ||
 		echo true)" = true
Signature · 1 line
2.29.2
#35 lore raw

Re: [PATCH v8 4/4] mergetool: Add automerge_enabled tool-specific override function

Am 28.12.20 um 05:54 schrieb Seth House:

Quoted 1 line

Hat-tip to Junio C Hamano for the implementation.

We usually write this as

Quoted 12 lines
---
 git-mergetool--lib.sh | 4 ++++
 git-mergetool.sh      | 2 +-
 2 files changed, 5 insertions(+), 1 deletion(-)
git-mergetool--lib.sh +0 -0
index e059b3559e..5084ceffeb 100644

git-mergetool--lib.sh setup_tool () { · line 168

Quoted 5 lines
git-mergetool--lib.sh +2 -0
@@ -164,6 +164,10 @@ setup_tool () {
 		return 1
 	}
 
+	automerge_enabled () {
+		true

I would have written this as return 0 instead of true like some of the functions above this hunk.

Quoted 11 lines
git-mergetool--lib.sh +2 -0
+	}
+
 	translate_merge_tool_path () {
 		echo "$1"
 	}
git-mergetool.sh +0 -0
index 929192d0f8..a44afd3822 100755

git-mergetool.sh merge_file () { · line 342

Quoted 8 lines
git-mergetool.sh +1 -1
@@ -336,7 +336,7 @@ merge_file () {
 
 	initialize_merge_tool "$merge_tool"
 
-	if test "$(
+	if automerge_enabled && test "$(
 		git config --get --bool "mergetool.$merge_tool.automerge" ||
 		git config --get --bool "mergetool.automerge" ||
 		echo true)" = true

-- Hannes

#36 lore raw

Re: [PATCH v8 4/4] mergetool: Add automerge_enabled tool-specific override function

In reply to Seth House
Quoted 1 line · Seth House

Hat-tip to Junio C Hamano for the implementation.

That is the least interesting thing the log message for this commit can talk about. Instead readers should be able to learn things like these from the log message (I am not saying the log message should have these in an enumerated list; I am just enumerating these as samples):

- Why does this exist?

- What does a tool author want to use the mechanism for, and how does the tool author use it?

- This mechanism allows tool authors to say "never allow autoMerge for this tool", but there is no provision to let them say "always use autoMerge without allowing users to turn it off".

It also needs a bit of documentation update to mention that individual mergetool backend can choose not to trigger the feature at all, even if the user configures it with mergetool.autoMerge and mergetool.<tool>.autoMerge options.

Thanks.

Quoted 34 lines
---
 git-mergetool--lib.sh | 4 ++++
 git-mergetool.sh      | 2 +-
 2 files changed, 5 insertions(+), 1 deletion(-)
git-mergetool--lib.sh +4 -0
index e059b3559e..5084ceffeb 100644
@@ -164,6 +164,10 @@ setup_tool () {
 		return 1
 	}
 
+	automerge_enabled () {
+		true
+	}
+
 	translate_merge_tool_path () {
 		echo "$1"
 	}
git-mergetool.sh +1 -1
index 929192d0f8..a44afd3822 100755
@@ -336,7 +336,7 @@ merge_file () {
 
 	initialize_merge_tool "$merge_tool"
 
-	if test "$(
+	if automerge_enabled && test "$(
 		git config --get --bool "mergetool.$merge_tool.automerge" ||
 		git config --get --bool "mergetool.automerge" ||
 		echo true)" = true
#37 lore raw

v9 cover mergetool: add automerge configuration

In reply to Seth House

Thank you Junio and Johannes for patiently pointing out all my newbie mistakes.

Changes since v8:

- Improve documentation of what the autoMerge flag does.
- Add documentation note for the automerge_enabled override.

- Junio, is there another place this should also live?

- Cross-reference merge.autoMerge and merge.<tool>.autoMerge docs.
- Sort the list of mergetool options. Added as a standalone commit in case this change doesn't belong as part of this patchset. I'd be happy to move this to a standalone patch if that's preferable.
- Improve all commit messages with full explanations and rationale:

- Johannes, I didn't directly respond to your initialize_merge_tool question because you're right that explanation should be part of the commit message. Please let me know if that now answers your question and if not we can discuss further in a thread.

- Fix omitted exit codes when running initialize_merge_tool.
- Fix automerge_enabled return value for consistency.
- Update commit message capitalization to conform to repo norms.
- Rephrase commit message 'thanks' as Helped-by markers.

Felipe Contreras (1):
mergetool: add automerge configuration

Seth House (4):
mergetool: alphabetize the mergetool config docs
mergetool: add per-tool support for the autoMerge flag
mergetool: break setup_tool out into separate initialization function
mergetool: add automerge_enabled tool-specific override function

 Documentation/config/mergetool.txt   | 28 ++++++++++++++++++++++------
 Documentation/git-mergetool--lib.txt |  4 ++++
 git-difftool--helper.sh              |  2 ++
 git-mergetool--lib.sh                | 11 ++++++++---
 git-mergetool.sh                     | 22 ++++++++++++++++++++++
 t/t7610-mergetool.sh                 | 18 ++++++++++++++++++
 6 files changed, 76 insertions(+), 9 deletions(-)
Signature · 1 line
2.29.2
#38 lore raw

PATCH v9 1/5 mergetool: add automerge configuration

From: Felipe Contreras <[email protected]>

The purpose of mergetools is to resolve conflicts when git cannot automatically do so.

In order to do that git has added markers in the specific areas that need resolving, which the user must manually fix. The tool is supposed to help with that.

However, by passing the original BASE, LOCAL, and REMOTE files, many changes without conflict are presented to the user when in fact nothing needs to be done for those.

We can fix that by propagating the final version of the file with the automatic merge to all the panes of the mergetool (BASE, LOCAL, and REMOTE), and only make them differ on the places where there are actual conflicts.

As most people will want the new behavior, we enable it by default. Users that do not want the new behavior can set the new configuration mergetool.autoMerge to false.

See Seth House's blog post [1] for the idea, and the rationale.

[1] https://www.eseth.org/2020/mergetools.html

---
 Documentation/config/mergetool.txt |  3 +++
 git-mergetool.sh                   | 17 +++++++++++++++++
 t/t7610-mergetool.sh               | 18 ++++++++++++++++++
 3 files changed, 38 insertions(+)
Documentation/config/mergetool.txt +3 -0
index 16a27443a3..7ce6d0d3ac 100644
@@ -61,3 +61,6 @@ mergetool.writeToTemp::
 
 mergetool.prompt::
 	Prompt before each invocation of the merge resolution program.
+
+mergetool.autoMerge::
+	Remove lines without conflicts from all the files. Defaults to `true`.
git-mergetool.sh +17 -0
index e3f6d543fb..f4db0cac8d 100755
@@ -239,6 +239,17 @@ checkout_staged_file () {
 	fi
 }
 
+auto_merge () {
+	git merge-file --diff3 --marker-size=7 -q -p "$LOCAL" "$BASE" "$REMOTE" >"$DIFF3"
+	if test -s "$DIFF3"
+	then
+		sed -e '/^<<<<<<< /,/^||||||| /d' -e '/^=======\r\?$/,/^>>>>>>> /d' "$DIFF3" >"$BASE"
+		sed -e '/^||||||| /,/^>>>>>>> /d' -e '/^<<<<<<< /d' "$DIFF3" >"$LOCAL"
+		sed -e '/^<<<<<<< /,/^=======\r\?$/d' -e '/^>>>>>>> /d' "$DIFF3" >"$REMOTE"
+	fi
+	rm -- "$DIFF3"
+}
+
 merge_file () {
 	MERGED="$1"
 
@@ -274,6 +285,7 @@ merge_file () {
 		BASE=${BASE##*/}
 	fi
 
+	DIFF3="$MERGETOOL_TMPDIR/${BASE}_DIFF3_$$$ext"
 	BACKUP="$MERGETOOL_TMPDIR/${BASE}_BACKUP_$$$ext"
 	LOCAL="$MERGETOOL_TMPDIR/${BASE}_LOCAL_$$$ext"
 	REMOTE="$MERGETOOL_TMPDIR/${BASE}_REMOTE_$$$ext"
@@ -322,6 +334,11 @@ merge_file () {
 	checkout_staged_file 2 "$MERGED" "$LOCAL"
 	checkout_staged_file 3 "$MERGED" "$REMOTE"
 
+	if test "$(git config --bool mergetool.autoMerge)" != "false"
+	then
+		auto_merge
+	fi
+
 	if test -z "$local_mode" || test -z "$remote_mode"
 	then
 		echo "Deleted merge conflict for '$MERGED':"
t/t7610-mergetool.sh +18 -0
index 70afdd06fa..ccabd04823 100755
@@ -828,4 +828,22 @@ test_expect_success 'mergetool -Oorder-file is honored' '
 	test_cmp expect actual
 '
 
+test_expect_success 'mergetool automerge' '
+	test_config mergetool.automerge true &&
+	test_when_finished "git reset --hard" &&
+	git checkout -b test${test_count}_b master &&
+	test_write_lines >file1 base "" a &&
+	git commit -a -m "base" &&
+	test_write_lines >file1 base "" c &&
+	git commit -a -m "remote update" &&
+	git checkout -b test${test_count}_a HEAD~ &&
+	test_write_lines >file1 local "" b &&
+	git commit -a -m "local update" &&
+	test_must_fail git merge test${test_count}_b &&
+	yes "" | git mergetool file1 &&
+	test_write_lines >expect local "" c &&
+	test_cmp expect file1 &&
+	git commit -m "test resolved with mergetool"
+'
+
 test_done
Signature · 1 line
2.29.2
#39 lore raw

PATCH v9 2/5 mergetool: alphabetize the mergetool config docs

In reply to Seth House

The ordering in this file has drifted a little. Let's make things better while we're adding new entres. :)

---
 Documentation/config/mergetool.txt | 20 ++++++++++----------
 1 file changed, 10 insertions(+), 10 deletions(-)
Documentation/config/mergetool.txt +10 -10
index 7ce6d0d3ac..3291fa7102 100644
@@ -1,7 +1,3 @@
-mergetool.<tool>.path::
-	Override the path for the given tool.  This is useful in case
-	your tool is not in the PATH.
-
 mergetool.<tool>.cmd::
 	Specify the command to invoke the specified merge tool.  The
 	specified command is evaluated in shell with the following
@@ -13,6 +9,10 @@ mergetool.<tool>.cmd::
 	merged; 'MERGED' contains the name of the file to which the merge
 	tool should write the results of a successful merge.
 
+mergetool.<tool>.path::
+	Override the path for the given tool.  This is useful in case
+	your tool is not in the PATH.
+
 mergetool.<tool>.trustExitCode::
 	For a custom merge command, specify whether the exit code of
 	the merge command can be used to determine whether the merge was
@@ -40,6 +40,9 @@ mergetool.meld.useAutoMerge::
 	value of `false` avoids using `--auto-merge` altogether, and is the
 	default value.
 
+mergetool.autoMerge::
+	Remove lines without conflicts from all the files. Defaults to `true`.
+
 mergetool.keepBackup::
 	After performing a merge, the original file with conflict markers
 	can be saved as a file with a `.orig` extension.  If this variable
@@ -53,14 +56,11 @@ mergetool.keepTemporaries::
 	preserved, otherwise they will be removed after the tool has
 	exited. Defaults to `false`.
 
+mergetool.prompt::
+	Prompt before each invocation of the merge resolution program.
+
 mergetool.writeToTemp::
 	Git writes temporary 'BASE', 'LOCAL', and 'REMOTE' versions of
 	conflicting files in the worktree by default.  Git will attempt
 	to use a temporary directory for these files when set `true`.
 	Defaults to `false`.
-
-mergetool.prompt::
-	Prompt before each invocation of the merge resolution program.
-
-mergetool.autoMerge::
-	Remove lines without conflicts from all the files. Defaults to `true`.
Signature · 1 line
2.29.2
#40 lore raw

PATCH v9 3/5 mergetool: add per-tool support for the autoMerge flag

In reply to Seth House

Keep the global mergetool flag and add a per-tool override flag so that users may enable the flag for one tool and disable it for another.

---
 Documentation/config/mergetool.txt | 15 ++++++++++++++-
 git-mergetool.sh                   |  5 ++++-
 2 files changed, 18 insertions(+), 2 deletions(-)
Documentation/config/mergetool.txt +14 -1
index 3291fa7102..bde472d49a 100644
@@ -1,3 +1,9 @@
+mergetool.<tool>.autoMerge::
+	A mergetool-specific override for the global `mergetool.autoMerge`
+	configuration flag. This allows individual mergetools to enable or
+	disable the flag regardless of the global setting. See
+	`mergetool.autoMerge` for the full description.
+
 mergetool.<tool>.cmd::
 	Specify the command to invoke the specified merge tool.  The
 	specified command is evaluated in shell with the following
@@ -41,7 +47,14 @@ mergetool.meld.useAutoMerge::
 	default value.
 
 mergetool.autoMerge::
-	Remove lines without conflicts from all the files. Defaults to `true`.
+	During a merge Git will automatically resolve as many conflicts as
+	possible and then wrap conflict markers around any conflicts that it
+	cannot resolve. This flag consolidates the non-conflicting parts into
+	the corresponding 'LOCAL' and 'REMOTE' files so that only the
+	unresolved conflicts are presented to the merge tool. Can be overriden
+	per-tool via the `mergetool.<tool>.autoMerge` configuration variable.
+	Note: individual mergetool scripts can elect to ignore user preferences
+	entirely. Defaults to `true`.
 
 mergetool.keepBackup::
 	After performing a merge, the original file with conflict markers
git-mergetool.sh +4 -1
index f4db0cac8d..e3c7d78d1d 100755
@@ -334,7 +334,10 @@ merge_file () {
 	checkout_staged_file 2 "$MERGED" "$LOCAL"
 	checkout_staged_file 3 "$MERGED" "$REMOTE"
 
-	if test "$(git config --bool mergetool.autoMerge)" != "false"
+	if test "$(
+		git config --get --bool "mergetool.$merge_tool.automerge" ||
+		git config --get --bool "mergetool.automerge" ||
+		echo true)" = true
 	then
 		auto_merge
 	fi
Signature · 1 line
2.29.2
#41 lore raw

PATCH v9 4/5 mergetool: break setup_tool out into separate initialization function

In reply to Seth House

This is preparation for the following commit where we need to source the mergetool shell script to look for overrides before run_merge_tool is called. Previously run_merge_tool both sourced that script and invoked the mergetool.

In the case of the following commit, we need the result of the automerge_enabled override, if it exists, well before we actually run run_merge_tool.

A new function initialize_merge_tool was chosen for consistency with run_merge_tool since setup_tool and setup_user_tool are not exposed or called directly.

---
 Documentation/git-mergetool--lib.txt | 4 ++++
 git-difftool--helper.sh              | 2 ++
 git-mergetool--lib.sh                | 7 ++++---
 git-mergetool.sh                     | 2 ++
 4 files changed, 12 insertions(+), 3 deletions(-)
Documentation/git-mergetool--lib.txt +4 -0
index 4da9d24096..3e8f59ac0e 100644
@@ -38,6 +38,10 @@ get_merge_tool_cmd::
 get_merge_tool_path::
 	returns the custom path for a merge tool.
 
+initialize_merge_tool::
+	bring merge tool specific functions into scope so they can be used or
+	overridden.
+
 run_merge_tool::
 	launches a merge tool given the tool name and a true/false
 	flag to indicate whether a merge base is present.
git-difftool--helper.sh +2 -0
index 46af3e60b7..234dd6944e 100755
@@ -61,6 +61,7 @@ launch_merge_tool () {
 		export BASE
 		eval $GIT_DIFFTOOL_EXTCMD '"$LOCAL"' '"$REMOTE"'
 	else
+		initialize_merge_tool "$merge_tool" &&
 		run_merge_tool "$merge_tool"
 	fi
 }
@@ -79,6 +80,7 @@ if test -n "$GIT_DIFFTOOL_DIRDIFF"
 then
 	LOCAL="$1"
 	REMOTE="$2"
+	initialize_merge_tool "$merge_tool" &&
 	run_merge_tool "$merge_tool" false
 else
 	# Launch the merge tool on each path provided by 'git diff'
git-mergetool--lib.sh +4 -3
index 7225abd811..e059b3559e 100644
@@ -248,6 +248,10 @@ trust_exit_code () {
 	fi
 }
 
+initialize_merge_tool () {
+	# Bring tool-specific functions into scope
+	setup_tool "$1" || return 1
+}
 
 # Entry point for running tools
 run_merge_tool () {
@@ -259,9 +263,6 @@ run_merge_tool () {
 	merge_tool_path=$(get_merge_tool_path "$1") || exit
 	base_present="$2"
 
-	# Bring tool-specific functions into scope
-	setup_tool "$1" || return 1
-
 	if merge_mode
 	then
 		run_merge_cmd "$1"
git-mergetool.sh +2 -0
index e3c7d78d1d..929192d0f8 100755
@@ -334,6 +334,8 @@ merge_file () {
 	checkout_staged_file 2 "$MERGED" "$LOCAL"
 	checkout_staged_file 3 "$MERGED" "$REMOTE"
 
+	initialize_merge_tool "$merge_tool"
+
 	if test "$(
 		git config --get --bool "mergetool.$merge_tool.automerge" ||
 		git config --get --bool "mergetool.automerge" ||
Signature · 1 line
2.29.2
#42 lore raw

Re: [PATCH v9 4/5] mergetool: break setup_tool out into separate initialization function

Am 28.12.20 um 20:29 schrieb Seth House:

git-mergetool.sh merge_file () { · line 337

Quoted 9 lines
git-mergetool.sh +1 -0
index e3c7d78d1d..929192d0f8 100755
@@ -334,6 +334,8 @@ merge_file () {
 	checkout_staged_file 2 "$MERGED" "$LOCAL"
 	checkout_staged_file 3 "$MERGED" "$REMOTE"
 
+	initialize_merge_tool "$merge_tool"

In my earlier review, I was not explicit about the lack of error handling of this invocation, because I hoped that you would notice it yourself. initialize_merge_tool does have a few failure modes via setup_tool that are not really unlikely; ignoring errors would be wrong, I think.

Before this change, the errors would be handled as part of the failing run_merge_tool call at the end of function merge_file. But now errors are ignored. Just appending || return would not be appropriate at this point because a lot has already happened before the call that has to be rewound. Would it be possible to move the call above the mergetool_tmpdir_init call, so that nothing has to be rewound?

git-mergetool.sh merge_file () { · line 341

Quoted 5 lines
git-mergetool.sh +1 -0
+
 	if test "$(
 		git config --get --bool "mergetool.$merge_tool.automerge" ||
 		git config --get --bool "mergetool.automerge" ||

-- Hannes

#43 lore raw

Re: [PATCH v9 4/5] mergetool: break setup_tool out into separate initialization function

Quoted 2 lines · Johannes Sixt

Would it be possible to move the call above the mergetool_tmpdir_init call, so that nothing has to be rewound?

Ah, I see. Good suggestion. Yes, moving that call higher works just fine. E.g.:

git-mergetool.sh +2 -2
index a44afd3822..36c1920dd6 100755
@@ -276,6 +276,8 @@ merge_file () {
 		ext=
 	esac
 
+	initialize_merge_tool "$merge_tool" || return
+
 	mergetool_tmpdir_init
 
 	if test "$MERGETOOL_TMPDIR" != "."
@@ -334,8 +336,6 @@ merge_file () {
 	checkout_staged_file 2 "$MERGED" "$LOCAL"
 	checkout_staged_file 3 "$MERGED" "$REMOTE"
 
-	initialize_merge_tool "$merge_tool"
-
 	if automerge_enabled && test "$(
 		git config --get --bool "mergetool.$merge_tool.automerge" ||
 		git config --get --bool "mergetool.automerge" ||

Thanks. I'll roll that change into a v10 patch series later today or tomorrow to give a little more time for any other feedback.

#44 lore raw

PATCH v9 5/5 mergetool: add automerge_enabled tool-specific override function

In reply to Seth House

The author or maintainer of a mergetool may optionally elect disable (or enable) the autoMerge feature for that mergetool even if the user has chosen differently using the mergetool.autoMerge and mergetool.<tool>.autoMerge options.

To add a tool-specific override, edit the mergetools/<tool> shell script for that tool and add an automerge_enabled function:

    automerge_enabled () {
        return 1
    }

Disabling may be desirable if the mergetool wants or needs access to the original, unmodified 'LOCAL', 'REMOTE', and 'BASE' versions of the conflicted file. For example:

- A tool may use a custom conflict resolution algorithm and prefer to ignore the results of Git's conflict resolution.
- A tool may want to visually compare/constrast the version of the file from before the merge (saved to 'LOCAL', 'REMOTE', and 'BASE') with Git's conflict resolution results (saved to 'MERGED').
- A student or researcher working on a new algorithm may want to directly compare the result of that algorithm with the result of Git's algorithm.

---
 git-mergetool--lib.sh | 4 ++++
 git-mergetool.sh      | 2 +-
 2 files changed, 5 insertions(+), 1 deletion(-)
git-mergetool--lib.sh +4 -0
index e059b3559e..567991abbc 100644
@@ -164,6 +164,10 @@ setup_tool () {
 		return 1
 	}
 
+	automerge_enabled () {
+		return 0
+	}
+
 	translate_merge_tool_path () {
 		echo "$1"
 	}
git-mergetool.sh +1 -1
index 929192d0f8..a44afd3822 100755
@@ -336,7 +336,7 @@ merge_file () {
 
 	initialize_merge_tool "$merge_tool"
 
-	if test "$(
+	if automerge_enabled && test "$(
 		git config --get --bool "mergetool.$merge_tool.automerge" ||
 		git config --get --bool "mergetool.automerge" ||
 		echo true)" = true
Signature · 1 line
2.29.2
#45 lore raw

RE: [PATCH v9 5/5] mergetool: add automerge_enabled tool-specific override function

Quoted 6 lines · Seth House

Disabling may be desirable if the mergetool wants or needs access to the original, unmodified 'LOCAL', 'REMOTE', and 'BASE' versions of the conflicted file. For example:

- A tool may use a custom conflict resolution algorithm and prefer to ignore the results of Git's conflict resolution.

If git's conflict resolution decides there are no conflicts, how would such tool "ignore" that?

Quoted 3 lines

- A tool may want to visually compare/constrast the version of the file from before the merge (saved to 'LOCAL', 'REMOTE', and 'BASE') with Git's conflict resolution results (saved to 'MERGED').

Can't such tool use "git checkout-index" for that?

Quoted 3 lines

- A student or researcher working on a new algorithm may want to directly compare the result of that algorithm with the result of Git's algorithm.

1. If git's algorithm decides there are no conflicts, and the new algorithm decides there are conflicts, how would such researcher find that out?

2. Can't such researcher simply do:
git -c mergetool.automerge=false mergetool?

Cheers.

Signature · 1 line
Felipe Contreras
#46 lore raw

Re: [PATCH v9 5/5] mergetool: add automerge_enabled tool-specific override function

In reply to Seth House

git-mergetool.sh merge_file () { · line 342

Quoted 13 lines · Seth House
git-mergetool.sh +1 -1
index 929192d0f8..a44afd3822 100755
@@ -336,7 +336,7 @@ merge_file () {
 
 	initialize_merge_tool "$merge_tool"
 
-	if test "$(
+	if automerge_enabled && test "$(
 		git config --get --bool "mergetool.$merge_tool.automerge" ||
 		git config --get --bool "mergetool.automerge" ||
 		echo true)" = true

This allows the tool author to say "nobody ever is allowed to use my tool with the automerge feature". I know I may have suggested something like that, but I am not sure if we want to be all that draconian.

If the user explicitly says "I want the new behaviour enabled for this particular merge tool", we are better off letting the user use it and take responsibility for the possible breakage.

My preference would probably be

- if "mergetool.$merge_tool.automerge" is set to 'true' or 'false', that's final.

- Your automerge_enabled helper that is by default 'true' (but allows individual merge_tool to return 'false') is asked, and if it says 'false', that's final. But 'true' from automerge_enabled is not final at this step.

- if "mergetool.automerge" is set to 'true' or 'false', that's final.

- otherwise, your automerge_enabled helper's answer (either 'true' or 'false') gives the final answer.

That way, those who use a broad "mergetool.automerge = true/false" would still honor what automerge_enabled yields (which is "enabled by default but individual merge_tool can set its default to be disabled"), while individual mergetool.$merge_tool.automerge configuration would always win.

Hmm?

#47 lore raw

Re: [PATCH v9 5/5] mergetool: add automerge_enabled tool-specific override function

Quoted 3 lines · Junio C Hamano

If the user explicitly says "I want the new behaviour enabled for this particular merge tool", we are better off letting the user use it and take responsibility for the possible breakage.

Good suggestion. Agreed on all counts. I'll roll that preference hierarchy into the v10 patch set.

#48 lore raw

Re: [PATCH v9 5/5] mergetool: add automerge_enabled tool-specific override function

Quoted 7 lines · Seth House
Quoted 3 lines · Junio C Hamano

If the user explicitly says "I want the new behaviour enabled for this particular merge tool", we are better off letting the user use it and take responsibility for the possible breakage.

Good suggestion. Agreed on all counts. I'll roll that preference hierarchy into the v10 patch set.

By the way, do you have any idea why we see test breakages only on macos when this topic is merged to 'seen'?

https://github.com/git/git/runs/1659807735?check_suite_focus=true#step:4:1641
https://github.com/git/git/runs/1659807735?check_suite_focus=true#step:5:2641

Thanks.

#49 lore raw

Re: [PATCH v9 5/5] mergetool: add automerge_enabled tool-specific override function

Quoted 2 lines · Junio C Hamano

By the way, do you have any idea why we see test breakages only on macos when this topic is merged to 'seen'?

Thanks for those links. I have an OSX machine nearby and will investigate tomorrow.

Related: are the Windows tests affected by this patch? I wanted to check for myself but I've been struggling with getting Git-for-Windows installed in a VM.

#50 lore raw

Re: [PATCH v9 5/5] mergetool: add automerge_enabled tool-specific override function

Quoted 10 lines · Seth House
Quoted 2 lines · Junio C Hamano

By the way, do you have any idea why we see test breakages only on macos when this topic is merged to 'seen'?

Thanks for those links. I have an OSX machine nearby and will investigate tomorrow.

Related: are the Windows tests affected by this patch? I wanted to check for myself but I've been struggling with getting Git-for-Windows installed in a VM.

On the left hand side of the page I gave the links to, it shows that 'windows-build' job is failing (and windows-test jobs are not run as a consequence). I am not sure why it failed, but I have a feeling that the build machinery hasn't even seen the code being built when it errored out.

cf. https://github.com/git/git/runs/1659807855?check_suite_focus=true#step:3:40

So we cannot tell (yet).

#51 lore raw

Re: [PATCH v9 5/5] mergetool: add automerge_enabled tool-specific override function

Hi Junio,

Quoted 19 lines · Junio C Hamano
Quoted 10 lines · Seth House
Quoted 2 lines · Junio C Hamano

By the way, do you have any idea why we see test breakages only on macos when this topic is merged to 'seen'?

Thanks for those links. I have an OSX machine nearby and will investigate tomorrow.

Related: are the Windows tests affected by this patch? I wanted to check for myself but I've been struggling with getting Git-for-Windows installed in a VM.

On the left hand side of the page I gave the links to, it shows that 'windows-build' job is failing (and windows-test jobs are not run as a consequence). I am not sure why it failed, but I have a feeling that the build machinery hasn't even seen the code being built when it errored out.

Quoted 3 lines

cf. https://github.com/git/git/runs/1659807855?check_suite_focus=true#step:3:40

So we cannot tell (yet).

There are unfortunately intermittent failures while downloading git-sdk-64-minimal; That's what this job is seeing. I restarted that build:
https://github.com/git/git/runs/1659807855?check_suite_focus=true#step:3:40

Ciao,
Dscho

#52 lore raw

v10 cover mergetool: add hideResolved configuration (was automerge)

In reply to Seth House

Changes since v9:

- Rename automerge to hideResolved.

Several mergetools have a feature that they call "automerge", "auto merge", or "auto solve" and Git has a mergetool.meld.useAutoMerge flag so I think it's better to avoid potential confusion. Plus Git already performed the merge and this doesn't do any additional merging, instead it just "hides" conflicts that Git already resolved.

- Reworked and consolidated commits and commit messages.

- Add preference hierarchy.

Followed Junio's suggestions:
https://lore.kernel.org/git/[email protected]/

- Switch from sed to two merge-file calls.

Thanks to everyone who helped with all the suggestions and fixup!s to get sed working cross-platform. Unfortunately there's not a great, portable method to preserve carriage returns when using both autocrlf and MSYS2: https://lore.kernel.org/git/20210120232447.GA35105@ellen/

Although calling merge-file twice is (much) less efficient than sed it's still fairly quick for small files. For large files it's likely opening those files in a mergetool will have a higher overhead than the merge-file invocations:
https://lore.kernel.org/git/20210122010902.GA48178@ellen/

A potential future optimisation could be to augment the C implementation (xmerge.c ?) with a flag to write two files as the merge is being performed instead of writing conflict markers.

- Kept initialize_merge_tool wrapper.

I updated the commit message where initialize_merge_tool is introduced to try and better explain my thinking for not simply exposing setup_tool instead. I'm happy to switch that if anyone still feels it should be switched.

Seth House (3):
mergetool: add hideResolved configuration
mergetool: break setup_tool out into separate initialization function
mergetool: add per-tool support and overrides for the hideResolved
flag

 Documentation/config/mergetool.txt   | 15 +++++++++++++++
 Documentation/git-mergetool--lib.txt |  4 ++++
 git-difftool--helper.sh              |  6 ++++++
 git-mergetool--lib.sh                | 11 ++++++++---
 git-mergetool.sh                     | 26 ++++++++++++++++++++++++++
 t/t7610-mergetool.sh                 | 18 ++++++++++++++++++
 6 files changed, 77 insertions(+), 3 deletions(-)
Signature · 1 line
2.29.2
#53 lore raw

PATCH v10 1/3 mergetool: add hideResolved configuration

The purpose of a mergetool is to help the user resolve any conflicts that Git cannot automatically resolve. If there is a conflict that must be resolved manually Git will write a file named MERGED which contains everything Git was able to resolve by itself and also everything that it was not able to resolve wrapped in conflict markers.

One way to think of MERGED is as a two- or three-way diff. If each "side" of the conflict markers is separately extracted an external tool can represent those conflicts as a side-by-side diff.

However many mergetools instead diff LOCAL and REMOTE both of which contain versions of the file from before the merge. Since the conflicts Git resolved automatically are not present it forces the user to manually re-resolve those conflicts. Some mergetools also show MERGED but often only for reference and not as the focal point to resolve the conflicts.

This adds a mergetool.hideResolved flag that will overwrite LOCAL and REMOTE with each corresponding "side" of a conflicted file and thus hide all conflicts that Git was able to resolve itself. Overwriting these files will immediately benefit any mergetool that uses them without requiring any changes.

No adverse effects were noted in a small survey of popular mergetools[1] so this behavior defaults to true. However it can be globally disabled by setting mergetool.hideResolved to false.

[1] https://www.eseth.org/2020/mergetools.html

    https://github.com/whiteinge/eseth/blob/c884424769fffb05d87afb33b2cf80cecb4044c3/2020/mergetools.md
---
 Documentation/config/mergetool.txt |  9 +++++++++
 git-mergetool.sh                   | 14 ++++++++++++++
 t/t7610-mergetool.sh               | 18 ++++++++++++++++++
 3 files changed, 41 insertions(+)
Documentation/config/mergetool.txt +9 -0
index 16a27443a3..3171bacf91 100644
@@ -40,6 +40,15 @@ mergetool.meld.useAutoMerge::
 	value of `false` avoids using `--auto-merge` altogether, and is the
 	default value.
 
+mergetool.hideResolved::
+	During a merge Git will automatically resolve as many conflicts as
+	possible and then wrap conflict markers around any conflicts that it
+	cannot resolve. This flag writes the non-conflicting parts into the
+	corresponding 'LOCAL' and 'REMOTE' files so that only the unresolved
+	conflicts are presented to the merge tool. Can be overriden per-tool
+	via the `mergetool.<tool>.hideResolved` configuration variable.
+	Defaults to `true`.
+
 mergetool.keepBackup::
 	After performing a merge, the original file with conflict markers
 	can be saved as a file with a `.orig` extension.  If this variable
git-mergetool.sh +14 -0
index e3f6d543fb..5b0d15ed89 100755
@@ -239,6 +239,13 @@ checkout_staged_file () {
 	fi
 }
 
+hide_resolved () {
+	git merge-file --ours -q -p "$LOCAL" "$BASE" "$REMOTE" >"$LCONFL"
+	git merge-file --theirs -q -p "$LOCAL" "$BASE" "$REMOTE" >"$RCONFL"
+	mv -- "$LCONFL" "$LOCAL"
+	mv -- "$RCONFL" "$REMOTE"
+}
+
 merge_file () {
 	MERGED="$1"
 
@@ -276,7 +283,9 @@ merge_file () {
 
 	BACKUP="$MERGETOOL_TMPDIR/${BASE}_BACKUP_$$$ext"
 	LOCAL="$MERGETOOL_TMPDIR/${BASE}_LOCAL_$$$ext"
+	LCONFL="$MERGETOOL_TMPDIR/${BASE}_LOCAL_LCONFL_$$$ext"
 	REMOTE="$MERGETOOL_TMPDIR/${BASE}_REMOTE_$$$ext"
+	RCONFL="$MERGETOOL_TMPDIR/${BASE}_REMOTE_RCONFL_$$$ext"
 	BASE="$MERGETOOL_TMPDIR/${BASE}_BASE_$$$ext"
 
 	base_mode= local_mode= remote_mode=
@@ -322,6 +331,11 @@ merge_file () {
 	checkout_staged_file 2 "$MERGED" "$LOCAL"
 	checkout_staged_file 3 "$MERGED" "$REMOTE"
 
+	if test "$(git config --get mergetool.hideResolved)" != "false"
+	then
+		hide_resolved
+	fi
+
 	if test -z "$local_mode" || test -z "$remote_mode"
 	then
 		echo "Deleted merge conflict for '$MERGED':"
t/t7610-mergetool.sh +18 -0
index 70afdd06fa..0e34b87e37 100755
@@ -828,4 +828,22 @@ test_expect_success 'mergetool -Oorder-file is honored' '
 	test_cmp expect actual
 '
 
+test_expect_success 'mergetool hideResolved' '
+	test_config mergetool.hideResolved true &&
+	test_when_finished "git reset --hard" &&
+	git checkout -b test${test_count}_b master &&
+	test_write_lines >file1 base "" a &&
+	git commit -a -m "base" &&
+	test_write_lines >file1 base "" c &&
+	git commit -a -m "remote update" &&
+	git checkout -b test${test_count}_a HEAD~ &&
+	test_write_lines >file1 local "" b &&
+	git commit -a -m "local update" &&
+	test_must_fail git merge test${test_count}_b &&
+	yes "" | git mergetool file1 &&
+	test_write_lines >expect local "" c &&
+	test_cmp expect file1 &&
+	git commit -m "test resolved with mergetool"
+'
+
 test_done
Signature · 1 line
2.29.2
#54 lore raw

Re: [PATCH v10 1/3] mergetool: add hideResolved configuration

Documentation/config/mergetool.txt mergetool.meld.useAutoMerge:: · line 50

Quoted 8 lines · Seth House
Documentation/config/mergetool.txt +8 -0
+mergetool.hideResolved::
+	During a merge Git will automatically resolve as many conflicts as
+	possible and then wrap conflict markers around any conflicts that it
+	cannot resolve. This flag writes the non-conflicting parts into the
+	corresponding 'LOCAL' and 'REMOTE' files so that only the unresolved
+	conflicts are presented to the merge tool. Can be overriden per-tool
+	via the `mergetool.<tool>.hideResolved` configuration variable.
+	Defaults to `true`.

This description makes the readers expect that the configuration variable is a boolean, and setting it to 'no' would disable the feature, but ...

git-mergetool.sh merge_file () { · line 334

Quoted 5 lines
git-mergetool.sh +1 -0
@@ -322,6 +331,11 @@ merge_file () {
 	checkout_staged_file 2 "$MERGED" "$LOCAL"
 	checkout_staged_file 3 "$MERGED" "$REMOTE"
 
+	if test "$(git config --get mergetool.hideResolved)" != "false"

... without --type=bool, any boolean 'false' value that is not exactly spelled 'false' won't be normalized and fail this test.

I haven't read the remaining 2 patches, so I cannot yet tell if I can just insert "--type=bool" here and everything would be fine, or if there are other fallouts for doing so.

Quoted 12 lines
git-mergetool.sh +4 -0
+	then
+		hide_resolved
+	fi
+
 	if test -z "$local_mode" || test -z "$remote_mode"
 	then
 		echo "Deleted merge conflict for '$MERGED':"
t/t7610-mergetool.sh +0 -0
index 70afdd06fa..0e34b87e37 100755

t/t7610-mergetool.sh test_expect_success 'mergetool -Oorder-file is honored' ' · line 834

Quoted 7 lines
t/t7610-mergetool.sh +4 -0
@@ -828,4 +828,22 @@ test_expect_success 'mergetool -Oorder-file is honored' '
 	test_cmp expect actual
 '
 
+test_expect_success 'mergetool hideResolved' '
+	test_config mergetool.hideResolved true &&
+	test_when_finished "git reset --hard" &&
+	git checkout -b test${test_count}_b master &&

As a new feature, this should work with the tip of 'master', but I think the t7610 test forces the initial branch name to be 'main'.

I'll tweak locally while queuing.

Quoted 15 lines
t/t7610-mergetool.sh +14 -0
+	test_write_lines >file1 base "" a &&
+	git commit -a -m "base" &&
+	test_write_lines >file1 base "" c &&
+	git commit -a -m "remote update" &&
+	git checkout -b test${test_count}_a HEAD~ &&
+	test_write_lines >file1 local "" b &&
+	git commit -a -m "local update" &&
+	test_must_fail git merge test${test_count}_b &&
+	yes "" | git mergetool file1 &&
+	test_write_lines >expect local "" c &&
+	test_cmp expect file1 &&
+	git commit -m "test resolved with mergetool"
+'
+
 test_done
#55 lore raw

PATCH v10 2/3 mergetool: break setup_tool out into separate initialization function

In reply to Seth House

This is preparation for the following commit where we need to source the mergetool shell script to look for overrides before run_merge_tool is called. Previously run_merge_tool both sourced that script and invoked the mergetool.

In the case of the following commit, we need the result of the hide_resolved override, if present, before we actually run run_merge_tool.

The new initialize_merge_tool wrapper is exposed and documented as a public interface for consistency with the existing run_merge_tool which is also public. Although setup_tool could instead be exposed directly, the related setup_user_tool would probably also want to be elevated to match and this felt the cleanest to me.

---
 Documentation/git-mergetool--lib.txt | 4 ++++
 git-difftool--helper.sh              | 6 ++++++
 git-mergetool--lib.sh                | 7 ++++---
 git-mergetool.sh                     | 2 ++
 4 files changed, 16 insertions(+), 3 deletions(-)
Documentation/git-mergetool--lib.txt +4 -0
index 4da9d24096..3e8f59ac0e 100644
@@ -38,6 +38,10 @@ get_merge_tool_cmd::
 get_merge_tool_path::
 	returns the custom path for a merge tool.
 
+initialize_merge_tool::
+	bring merge tool specific functions into scope so they can be used or
+	overridden.
+
 run_merge_tool::
 	launches a merge tool given the tool name and a true/false
 	flag to indicate whether a merge base is present.
git-difftool--helper.sh +6 -0
index 46af3e60b7..992124cc67 100755
@@ -61,6 +61,9 @@ launch_merge_tool () {
 		export BASE
 		eval $GIT_DIFFTOOL_EXTCMD '"$LOCAL"' '"$REMOTE"'
 	else
+		initialize_merge_tool "$merge_tool"
+		# ignore the error from the above --- run_merge_tool
+		# will diagnose unusable tool by itself
 		run_merge_tool "$merge_tool"
 	fi
 }
@@ -79,6 +82,9 @@ if test -n "$GIT_DIFFTOOL_DIRDIFF"
 then
 	LOCAL="$1"
 	REMOTE="$2"
+	initialize_merge_tool "$merge_tool"
+	# ignore the error from the above --- run_merge_tool
+	# will diagnose unusable tool by itself
 	run_merge_tool "$merge_tool" false
 else
 	# Launch the merge tool on each path provided by 'git diff'
git-mergetool--lib.sh +4 -3
index 7225abd811..e059b3559e 100644
@@ -248,6 +248,10 @@ trust_exit_code () {
 	fi
 }
 
+initialize_merge_tool () {
+	# Bring tool-specific functions into scope
+	setup_tool "$1" || return 1
+}
 
 # Entry point for running tools
 run_merge_tool () {
@@ -259,9 +263,6 @@ run_merge_tool () {
 	merge_tool_path=$(get_merge_tool_path "$1") || exit
 	base_present="$2"
 
-	# Bring tool-specific functions into scope
-	setup_tool "$1" || return 1
-
 	if merge_mode
 	then
 		run_merge_cmd "$1"
git-mergetool.sh +2 -0
index 5b0d15ed89..865f12551a 100755
@@ -272,6 +272,8 @@ merge_file () {
 		ext=
 	esac
 
+	initialize_merge_tool "$merge_tool" || return
+
 	mergetool_tmpdir_init
 
 	if test "$MERGETOOL_TMPDIR" != "."
Signature · 1 line
2.29.2
#56 lore raw

PATCH v10 3/3 mergetool: add per-tool support and overrides for the hideResolved flag

In reply to Seth House

Keep the global mergetool flag and add a per-tool override flag so that users may enable the flag for one tool and disable it for another. In addition, the author or maintainer of a mergetool may optionally elect to set the default hideResolved value for that mergetool.

To disable the feature for a specific tool, edit the mergetools/<tool> shell script for that tool and add a hide_resolved_enabled function:

    hide_resolved_enabled () {
        return 1
    }

Disabling may be desirable if the mergetool wants or needs access to the original, unmodified 'LOCAL' and 'REMOTE' versions of the conflicted file. For example:

- A tool may use a custom conflict resolution algorithm and prefer to ignore the results of Git's conflict resolution.
- A tool may want to visually compare/constrast the version of the file from before the merge (saved to 'LOCAL', 'REMOTE', and 'BASE') with Git's conflict resolution results (saved to 'MERGED').

---
 Documentation/config/mergetool.txt |  6 ++++++
 git-mergetool--lib.sh              |  4 ++++
 git-mergetool.sh                   | 14 ++++++++++++--
 3 files changed, 22 insertions(+), 2 deletions(-)
Documentation/config/mergetool.txt +6 -0
index 3171bacf91..046816fb07 100644
@@ -13,6 +13,12 @@ mergetool.<tool>.cmd::
 	merged; 'MERGED' contains the name of the file to which the merge
 	tool should write the results of a successful merge.
 
+mergetool.<tool>.hideResolved::
+	A mergetool-specific override for the global `mergetool.hideResolved`
+	configuration flag. This allows individual mergetools to enable or
+	disable the flag regardless of the global setting. See
+	`mergetool.hideResolved` for the full description.
+
 mergetool.<tool>.trustExitCode::
 	For a custom merge command, specify whether the exit code of
 	the merge command can be used to determine whether the merge was
git-mergetool--lib.sh +4 -0
index e059b3559e..11f00dde41 100644
@@ -164,6 +164,10 @@ setup_tool () {
 		return 1
 	}
 
+	hide_resolved_enabled () {
+		return 0
+	}
+
 	translate_merge_tool_path () {
 		echo "$1"
 	}
git-mergetool.sh +12 -2
index 865f12551a..6cf3884277 100755
@@ -333,9 +333,19 @@ merge_file () {
 	checkout_staged_file 2 "$MERGED" "$LOCAL"
 	checkout_staged_file 3 "$MERGED" "$REMOTE"
 
-	if test "$(git config --get mergetool.hideResolved)" != "false"
+	# hideResolved preferences hierarchy:
+	# First respect user's tool-specific configuration if exists.
+	if test "$(git config --get "mergetool.$merge_tool.hideResolved")" != "false"
 	then
-		hide_resolved
+		# Next respect tool-specified configuration.
+		if hide_resolved_enabled
+		then
+			# Finally respect if user has a global disable.
+			if test "$(git config --get "mergetool.hideResolved")" != "false"
+			then
+				hide_resolved
+			fi
+		fi
 	fi
 
 	if test -z "$local_mode" || test -z "$remote_mode"
Signature · 1 line
2.29.2
#57 lore raw

Re: [PATCH v10 3/3] mergetool: add per-tool support and overrides for the hideResolved flag

Quoted 4 lines · Seth House

Keep the global mergetool flag and add a per-tool override flag so that users may enable the flag for one tool and disable it for another. In addition, the author or maintainer of a mergetool may optionally elect to set the default hideResolved value for that mergetool.

OK.

Quoted 10 lines

To disable the feature for a specific tool, edit the mergetools/<tool> shell script for that tool and add a hide_resolved_enabled function:

    hide_resolved_enabled () {
        return 1
    }

Disabling may be desirable if the mergetool wants or needs access to the original, unmodified 'LOCAL' and 'REMOTE' versions of the conflicted file.

The above sounds as if it is a hint/help for end users, but it is unreasonable to expect all end users of a particular <tool> to edit part of their Git installation. I suspect that you didn't mean it that way, and instead it is meant to advise (new) tool authors who will add mergetools/<tool> for their own tool, and when read with that in mind, it does make sort-of sense (except that when you are author of this thing, you won't "edit" as if you are modifying something that already exists---you'd be the one who is adding the <tool> under mergetools/ directory).

For an end-user, to disable the feature for a tool, you'd just configure mergetool.<tool>.hideResolved to 'false', right?

Quoted 22 lines

For example:

- A tool may use a custom conflict resolution algorithm and prefer to ignore the results of Git's conflict resolution.
- A tool may want to visually compare/constrast the version of the file from before the merge (saved to 'LOCAL', 'REMOTE', and 'BASE') with Git's conflict resolution results (saved to 'MERGED').

---
 Documentation/config/mergetool.txt |  6 ++++++
 git-mergetool--lib.sh              |  4 ++++
 git-mergetool.sh                   | 14 ++++++++++++--
 3 files changed, 22 insertions(+), 2 deletions(-)
Documentation/config/mergetool.txt +0 -0
index 3171bacf91..046816fb07 100644

Documentation/config/mergetool.txt mergetool.<tool>.cmd:: · line 20

Quoted 8 lines
Documentation/config/mergetool.txt +5 -0
@@ -13,6 +13,12 @@ mergetool.<tool>.cmd::
 	merged; 'MERGED' contains the name of the file to which the merge
 	tool should write the results of a successful merge.
 
+mergetool.<tool>.hideResolved::
+	A mergetool-specific override for the global `mergetool.hideResolved`
+	configuration flag. This allows individual mergetools to enable or
+	disable the flag regardless of the global setting. See
+	`mergetool.hideResolved` for the full description.

This description is iffy.

The configuration allows "users" to enable or disable the feature for individual mergetools, overriding the 'mergetool.hideResolved' global setting, no? The above paragraph reads as if the tool author is enabling/disabling it.

Quoted 20 lines
git-mergetool--lib.sh +4 -0
index e059b3559e..11f00dde41 100644
@@ -164,6 +164,10 @@ setup_tool () {
 		return 1
 	}
 
+	hide_resolved_enabled () {
+		return 0
+	}
+
 	translate_merge_tool_path () {
 		echo "$1"
 	}
git-mergetool.sh +0 -0
index 865f12551a..6cf3884277 100755

git-mergetool.sh merge_file () { · line 338

Quoted 7 lines
git-mergetool.sh +3 -1
@@ -333,9 +333,19 @@ merge_file () {
 	checkout_staged_file 2 "$MERGED" "$LOCAL"
 	checkout_staged_file 3 "$MERGED" "$REMOTE"
 
-	if test "$(git config --get mergetool.hideResolved)" != "false"
+	# hideResolved preferences hierarchy:
+	# First respect user's tool-specific configuration if exists.
+	if test "$(git config --get "mergetool.$merge_tool.hideResolved")" != "false"

The same "--type=bool" comment applies to this step, too.

git-mergetool.sh merge_file () { · line 349

Quoted 11 lines
git-mergetool.sh +9 -0
 	then
+		# Next respect tool-specified configuration.
+		if hide_resolved_enabled
+		then
+			# Finally respect if user has a global disable.
+			if test "$(git config --get "mergetool.hideResolved")" != "false"
+			then
+				hide_resolved
+			fi
+		fi
 	fi

I am not sure if I understand this logic.

If the user says "for the tool <tool>, set hideresolved to true/false" explicitly, I think it should be final. Even if the tool's author expresses that s/he prefers not to have to work on a pre-munged input by setting hide_resolved_enabled to false, if the end-user says s/he wants to use it on that tool, we do not want to help the tool to override the user's wish.

If the user says "use hideresolved feature, as I like it in general" by setting mergetool.hideResolved, on the other hand, it may be also reasonable to heed "no, I recommend against it for this tool" for individual tool whose hide_resolved_enabled returns false. And if the global is set to 'false', the user says "I do not want it", and it may be iffy to let individual tool to countermand it.

WHen dealing with either of these variables, therefore, you'd need to know if the variable is not set at all, or if the variable is set to true, or to false. Even if we default to enabled, we need to be able to tell if the user didn't say anything (and we enabled the feature for the user because of our default choice), or if the user explicitly said s/he wants it.

In other words, you'd need to treat mergetool.hideResolved and mergetool.$merge_tool.hideResolved as tristates.

Here is how "git config --type=bool" can be used to normalize various ways to spell true/false and tell between "not set" and "set to some value":

    $ git -c a.b config --type=bool a.b; echo $?
    true
    0
    $ git -c a.b=yes config --type=bool a.b; echo $?
    true
    0
    $ git -c a.b=0 config --type=bool a.b; echo $?
    false
    0
    $ git config --type=bool a.b; echo $?
    1

IOW, if "git config --type=bool" fails, the user does not have the variable set. If it succeeds, you'd get normalized 'true/false' string on its standard output.

Using that technique, here is my attempt to rewrite the above logic, with commentary.

global_config=mergetool.hideResolved tool_config=mergetool.$merge_tool.hideResolved

if enabled=$(git config --type=bool "$tool_config") then
# The user explicitly says true or false, so there # is no point in asking any other source of preferences ;
elif enabled=$(git config --type=bool "$global_config") then
# There is a blanket preference for all tools, and 'true' # means "I like the hide-resolved in general, so use it # when appropriate" by the user. We can let the tool # author to override and disable, though.
#
# On the other hand, when set to 'false', it is "I really # don't like the feature in general, so do not use it # anywhere", which we take it as final, without letting # the tool override it.
if test "$enabled" = true && hide_resolved_enabled then
enabled=true
else
enabled=false
fi
else
# The user does not have preference. Ask the tool if hide_resolved_enabled
then
enabled=true
else
enabled=false
fi
fi

# Now we know if the feature should be used.
if test "$enabled" = true
then
hide_resolved
fi

#58 lore raw

v11 cover mergetool: add hideResolved configuration (was automerge)

In reply to Seth House

Changes since v10:

- Update the calls to git config to return normalized strings.

Junio, thank you for explaining the existence/omission and normalized strings tristate. I missed that in the docs and that's perfect.

- Adopt Junio's replacement preference hierarchy conditionals to respect opt-ins and not just opt-outs.

Your suggested code worked out-of-box in all the scenarios I could think to test. \o/

- Tweak the mergetool.hideResolved docs to call out the role of LOCAL and REMOTE.

- Reword commit messages and docs to better differentiate between config flags users set and code that merge tool maintainers write.

Seth House (3):
mergetool: add hideResolved configuration
mergetool: break setup_tool out into separate initialization function
mergetool: add per-tool support and overrides for the hideResolved
flag

 Documentation/config/mergetool.txt   | 14 ++++++++
 Documentation/git-mergetool--lib.txt |  4 +++
 git-difftool--helper.sh              |  6 ++++
 git-mergetool--lib.sh                | 11 ++++--
 git-mergetool.sh                     | 52 ++++++++++++++++++++++++++++
 t/t7610-mergetool.sh                 | 18 ++++++++++
 6 files changed, 102 insertions(+), 3 deletions(-)
Signature · 1 line
2.29.2
#59 lore raw

PATCH v11 1/3 mergetool: add hideResolved configuration

The purpose of a mergetool is to help the user resolve any conflicts that Git cannot automatically resolve. If there is a conflict that must be resolved manually Git will write a file named MERGED which contains everything Git was able to resolve by itself and also everything that it was not able to resolve wrapped in conflict markers.

One way to think of MERGED is as a two- or three-way diff. If each "side" of the conflict markers is separately extracted an external tool can represent those conflicts as a side-by-side diff.

However many mergetools instead diff LOCAL and REMOTE both of which contain versions of the file from before the merge. Since the conflicts Git resolved automatically are not present it forces the user to manually re-resolve those conflicts. Some mergetools also show MERGED but often only for reference and not as the focal point to resolve the conflicts.

This adds a mergetool.hideResolved flag that will overwrite LOCAL and REMOTE with each corresponding "side" of a conflicted file and thus hide all conflicts that Git was able to resolve itself. Overwriting these files will immediately benefit any mergetool that uses them without requiring any changes to the tool.

No adverse effects were noted in a small survey of popular mergetools[1] so this behavior defaults to true. However it can be globally disabled by setting mergetool.hideResolved to false.

[1] https://www.eseth.org/2020/mergetools.html

    https://github.com/whiteinge/eseth/blob/c884424769fffb05d87afb33b2cf80cecb4044c3/2020/mergetools.md
---
 Documentation/config/mergetool.txt | 10 ++++++++++
 git-mergetool.sh                   | 14 ++++++++++++++
 t/t7610-mergetool.sh               | 18 ++++++++++++++++++
 3 files changed, 42 insertions(+)
Documentation/config/mergetool.txt +10 -0
index 16a27443a3..b858191970 100644
@@ -40,6 +40,16 @@ mergetool.meld.useAutoMerge::
 	value of `false` avoids using `--auto-merge` altogether, and is the
 	default value.
 
+mergetool.hideResolved::
+	During a merge Git will automatically resolve as many conflicts as
+	possible and write the 'MERGED' file containing conflict markers around
+	any conflicts that it cannot resolve; 'LOCAL' and 'REMOTE' normally
+	represent the versions of the file from before Git's conflict
+	resolution. This flag causes 'LOCAL' and 'REMOTE' to be overwriten so
+	that only the unresolved conflicts are presented to the merge tool. Can
+	be configured per-tool via the `mergetool.<tool>.hideResolved`
+	configuration variable. Defaults to `true`.
+
 mergetool.keepBackup::
 	After performing a merge, the original file with conflict markers
 	can be saved as a file with a `.orig` extension.  If this variable
git-mergetool.sh +14 -0
index e3f6d543fb..40a103443d 100755
@@ -239,6 +239,13 @@ checkout_staged_file () {
 	fi
 }
 
+hide_resolved () {
+	git merge-file --ours -q -p "$LOCAL" "$BASE" "$REMOTE" >"$LCONFL"
+	git merge-file --theirs -q -p "$LOCAL" "$BASE" "$REMOTE" >"$RCONFL"
+	mv -- "$LCONFL" "$LOCAL"
+	mv -- "$RCONFL" "$REMOTE"
+}
+
 merge_file () {
 	MERGED="$1"
 
@@ -276,7 +283,9 @@ merge_file () {
 
 	BACKUP="$MERGETOOL_TMPDIR/${BASE}_BACKUP_$$$ext"
 	LOCAL="$MERGETOOL_TMPDIR/${BASE}_LOCAL_$$$ext"
+	LCONFL="$MERGETOOL_TMPDIR/${BASE}_LOCAL_LCONFL_$$$ext"
 	REMOTE="$MERGETOOL_TMPDIR/${BASE}_REMOTE_$$$ext"
+	RCONFL="$MERGETOOL_TMPDIR/${BASE}_REMOTE_RCONFL_$$$ext"
 	BASE="$MERGETOOL_TMPDIR/${BASE}_BASE_$$$ext"
 
 	base_mode= local_mode= remote_mode=
@@ -322,6 +331,11 @@ merge_file () {
 	checkout_staged_file 2 "$MERGED" "$LOCAL"
 	checkout_staged_file 3 "$MERGED" "$REMOTE"
 
+	if test "$(git config --type=bool mergetool.hideResolved)" != "false"
+	then
+		hide_resolved
+	fi
+
 	if test -z "$local_mode" || test -z "$remote_mode"
 	then
 		echo "Deleted merge conflict for '$MERGED':"
t/t7610-mergetool.sh +18 -0
index 04b0095072..cec4a860ef 100755
@@ -842,4 +842,22 @@ test_expect_success 'mergetool --tool-help shows recognized tools' '
 	grep meld mergetools
 '
 
+test_expect_success 'mergetool hideResolved' '
+	test_config mergetool.hideResolved true &&
+	test_when_finished "git reset --hard" &&
+	git checkout -b test${test_count}_b master &&
+	test_write_lines >file1 base "" a &&
+	git commit -a -m "base" &&
+	test_write_lines >file1 base "" c &&
+	git commit -a -m "remote update" &&
+	git checkout -b test${test_count}_a HEAD~ &&
+	test_write_lines >file1 local "" b &&
+	git commit -a -m "local update" &&
+	test_must_fail git merge test${test_count}_b &&
+	yes "" | git mergetool file1 &&
+	test_write_lines >expect local "" c &&
+	test_cmp expect file1 &&
+	git commit -m "test resolved with mergetool"
+'
+
 test_done
Signature · 1 line
2.29.2
#60 lore raw

PATCH 1/1 mergetool: do not enable hideResolved by default

A typical mergetool uses four panes, showing the content of the file being resolved from MERGE_BASE ('BASE'), HEAD ('LOCAL'), MERGE_HEAD ('REMOTE'), and the working copy. This allows understanding the conflicts in context: by seeing the entire content of the file from MERGE_HEAD, say, we can see the full intent of the code we are pulling in and understand what they were trying to do that conflicted with our own changes.

Sometimes, though, the exact content of these three competing versions of a file is not so important. Especially if the mergetool supports folding unchanged lines, the new 'mergetool.hideResolved' feature can be helpful for allowing a person resolving a merge to focus on the portion with conflicts. For sections of the file where BASE matched LOCAL or REMOTE, this feature makes all three versions match the resolved version, so that the user resolving can focus exclusively on the portions with conflicts. In other words, hideResolved makes a multi-pane merge tool show a similar amount of information to the file with conflict markers with conflictstyle=diff3, saving the operator from having to pay attention to parts that resolved cleanly.

98ea309b3f (mergetool: add hideResolved configuration, 2021-02-09) which introduced this setting enabled it by default, explaining:

No adverse effects were noted in a small survey of popular mergetools[1] so this behavior defaults to true. However it can be globally disabled by setting mergetool.hideResolved to false.

In practice, however, this has proved confusing for users. No indication is shown in the UI that the base, local, and remote versions shown have been modified by additional resolution. Especially in cases where conflicts involve elements beyond textual conflict, it has resulted in incorrect resolutions and wasted work to figure out what happened. Flip the default back to the traditional behavior of false: although the old behavior involves slightly slower merges in the only-textual-conflicts case, it prevents this kind of painful moment of betrayal by one's tools, which is more important.

Should we want to migrate to hideResolved=true in the future, we still can. It just requires a more careful migration, including a period where "git mergetool" shows a warning or errors out in affected cases.

---
Hi,

Quoted 3 lines · Seth House

No adverse effects were noted in a small survey of popular mergetools[1] so this behavior defaults to true. However it can be globally disabled by setting mergetool.hideResolved to false.

Thanks much for protecting this by a flag. We tried this out internally at Google when it hit "next" and not too long later realized that the new default of "true" is not workable for us. I don't believe it's the right default for Git, either, hence this patch.

Thanks for working on the merge resolution workflow; it's much appreciated.

Sincerely,
Jonathan

 Documentation/config/mergetool.txt | 2 +-
 git-mergetool.sh                   | 9 ++-------
 2 files changed, 3 insertions(+), 8 deletions(-)
Documentation/config/mergetool.txt +1 -1
index 90f76f5b9b..cafbbef46a 100644
@@ -53,7 +53,7 @@ mergetool.hideResolved::
 	resolution. This flag causes 'LOCAL' and 'REMOTE' to be overwriten so
 	that only the unresolved conflicts are presented to the merge tool. Can
 	be configured per-tool via the `mergetool.<tool>.hideResolved`
-	configuration variable. Defaults to `true`.
+	configuration variable. Defaults to `false`.
 
 mergetool.keepBackup::
 	After performing a merge, the original file with conflict markers
git-mergetool.sh +2 -7
index 911470a5b2..f751d9cfe2 100755
@@ -358,13 +358,8 @@ merge_file () {
 		    enabled=false
 		fi
 	else
-		# The user does not have a preference. Ask the tool.
-		if hide_resolved_enabled
-		then
-		    enabled=true
-		else
-		    enabled=false
-		fi
+		# The user does not have a preference. Default to disabled.
+		enabled=false
 	fi
 
 	if test "$enabled" = true
Signature · 1 line
2.31.0.rc1.246.gcd05c9c855
#61 lore raw

Re: [PATCH] mergetool: do not enable hideResolved by default

Quoted 7 lines · Jonathan Nieder

A typical mergetool uses four panes, showing the content of the file being resolved from MERGE_BASE ('BASE'), HEAD ('LOCAL'), MERGE_HEAD ('REMOTE'), and the working copy. This allows understanding the conflicts in context: by seeing the entire content of the file from MERGE_HEAD, say, we can see the full intent of the code we are pulling in and understand what they were trying to do that conflicted with our own changes.

Well said. Agreed on all counts.

The very early days of these patch sets touched on this exact discussion point. (I'd link to it but that early discussion was a tad...unfocused.) I make semi-frequent reference of those versions of the conflicted file in the way you describe and have disabled hideResolved for a merge tool I maintain for that reason.

Quoted 7 lines

No adverse effects were noted in a small survey of popular mergetools[1] so this behavior defaults to true. However it can be globally disabled by setting mergetool.hideResolved to false.

In practice, however, this has proved confusing for users. No indication is shown in the UI that the base, local, and remote versions shown have been modified by additional resolution.

Compelling point. This flag drastically changes what LOCAL and REMOTE represent with little to no explanation.

There are three options to achieve the same end-goal of hideResolved that I've thought of:

1. Individual merge tools should do this work, not Git.

A merge tool already has all the information needed to hide already-resolved conflicts since that is what MERGED represents. Conflict markers *are* a two-way diff and a merge tool should display them as such, rather than display the textual markers verbatim.

In many ways this is the ideal approach -- all merge tools could be doing this with existing Git right now but none have seemingly thought of doing so yet.

2. Git could pass six versions of the conflicted file to a merge tool, rather than the current four.

Merge tools could accept LOCAL, REMOTE, BASE, MERGED (as most currently do), and also LCONFL and RCONFL files. The latter two being copies of MERGED but "pre split" by Git into the left conflicts and the right conflicts.

This would spare the merge tool the work of splitting MERGED. It may encourage them to continue displaying LOCAL and REMOTE as useful context but also make it easy to diff LCONFL with RCONFL and use that diff to actually resolve the conflict. It could also make things worse, as many tools simply diff _every_ file Git gives them regardless if that makes sense or not (>_<).

3. Git could overwrite LOCAL and REMOTE to display only unresolved conflicts.

(The current hideResolved addition.) This has the pragmatic benefit of requiring the least amount of change for all merge tools, but to your point above, *destroys* valuable data -- the additional context to help understand where the conflicts came from -- and that data can't be viewd without running additional Git commands to fetch it.

Defaulting hideResolved to off is a fine change IMO. We don't have a way to communicate to the end-user that LOCAL and REMOTE represent something markedly different than what they have traditionally represented, so having this be an opt-in will force the user to read the docs and understand the ramifications.

I really appreciate your thoughts that accompanied this patch. Sorry for the long response but your email made me want to ask the question:

Does the need to default hideResolved to off mean that it is the wrong approach?

Thinking through an end-user's workflow: would a user want to configure two copies of the same merge tool -- one with hideResolved and one without? An easy conflict could benefit from the former but if it's a tricky conflict the user would have to exit the tool and reopen the same tool without the flag. That sounds like an annoying workflow, and although the user would now have that extra, valuable context it would also put them squarely back into the current state of viewing already-resolved conflicts.

I know the Option 3, hideResolved, is merged and has that momentum and this patch looks good to me -- but perhaps Option 2 is more "correct", or Option 1, or yet another option I haven't thought of. Thoughts?

#62 lore raw

Re: [PATCH] mergetool: do not enable hideResolved by default

Hi,

Quoted 5 lines · Seth House

The very early days of these patch sets touched on this exact discussion point. (I'd link to it but that early discussion was a tad...unfocused.) I make semi-frequent reference of those versions of the conflicted file in the way you describe and have disabled hideResolved for a merge tool I maintain for that reason.

Thanks. Do you have a public example of a merge that was produced in such a way? It might help focus the discussion.

For concreteness' sake: in the repository that Dana mentioned, one can see some merges from before hideResolved at
https://android.googlesource.com/platform/tools/idea/+log/mirror-goog-studio-master-dev/build.txt.

The xml files there (I'm not sure these are the right ones for me to focus on, just commenting as I observe) remind me of other routine conflicts with xml I've had to resolve in the past, e.g. at
https://git.eclipse.org/r/c/jgit/jgit/+/134451/3. Having information from each side of the merge and not a mixture can be very helpful in this kind of case. That's especially true when the three-way merge algorithm didn't end up lining up the files correctly, which has happened from time to time to me in files with repetitive structure.

[...]

Quoted 11 lines

There are three options to achieve the same end-goal of hideResolved that I've thought of:

1. Individual merge tools should do this work, not Git.

A merge tool already has all the information needed to hide already-resolved conflicts since that is what MERGED represents. Conflict markers *are* a two-way diff and a merge tool should display them as such, rather than display the textual markers verbatim.

Quoted 3 lines

In many ways this is the ideal approach -- all merge tools could be doing this with existing Git right now but none have seemingly thought of doing so yet.

One obstacle to this is that a merge tool can't count on the file in the worktree containing pristine conflict markers, because the user may have already started to work on the merge resolution.

Quoted 8 lines

2. Git could pass six versions of the conflicted file to a merge tool, rather than the current four.

Merge tools could accept LOCAL, REMOTE, BASE, MERGED (as most currently do), and also LCONFL and RCONFL files. The latter two being copies of MERGED but "pre split" by Git into the left conflicts and the right conflicts.

Quoted 6 lines

This would spare the merge tool the work of splitting MERGED. It may encourage them to continue displaying LOCAL and REMOTE as useful context but also make it easy to diff LCONFL with RCONFL and use that diff to actually resolve the conflict. It could also make things worse, as many tools simply diff _every_ file Git gives them regardless if that makes sense or not (>_<).

Interesting! I kind of like this, especially if it were something the tool could opt in to. That said, I'm not the best person to ask, since I never ended up finding a good workflow using mergetool for my own use; instead, I tend to do the work of a merge tool "by hand":

- gradually resolving the merge in each diff3-style conflict hunk by removing common lines from base+local and base+remote until there is nothing left in base

- in harder cases, making the worktree match the local version, putting the diff from base to remote in a temporary file, and then hunk by hunk applying it

- in even harder cases, using git-imerge
<https://github.com/mhagger/git-imerge>

[...]

Quoted 5 lines

3. Git could overwrite LOCAL and REMOTE to display only unresolved conflicts.

(The current hideResolved addition.) This has the pragmatic benefit of requiring the least amount of change for all merge tools,

That's a good argument for having the option available, *as long as the user explicitly turns it on*.

[...]

Quoted 2 lines

Does the need to default hideResolved to off mean that it is the wrong approach?

One disadvantage relative to (1) is that the mergetool has no way to visually distinguish the automatically resolved portion. For that reason, I suspect this will never be something we can make the default. But in principle I'm not against it existing.

The implementation is concise and maintainable. The documentation adds a little user-facing complexity; I think as long as we're able to keep it clear and well maintained, that should be okay.

git-mergetool.txt probably ought to mention the hideResolved setting. Otherwise, users can have a confusing experience if they set the config once and forget about it later.

[...]

Quoted 9 lines

Thinking through an end-user's workflow: would a user want to configure two copies of the same merge tool -- one with hideResolved and one without? An easy conflict could benefit from the former but if it's a tricky conflict the user would have to exit the tool and reopen the same tool without the flag. That sounds like an annoying workflow, and although the user would now have that extra, valuable context it would also put them squarely back into the current state of viewing already-resolved conflicts.

Quoted 3 lines

I know the Option 3, hideResolved, is merged and has that momentum and this patch looks good to me -- but perhaps Option 2 is more "correct", or Option 1, or yet another option I haven't thought of. Thoughts?

I suspect option 1 is indeed more correct. Dana mentions that some mergetools (p4merge?) use different colors to highlight the 'automatically resolved' portions, something that isn't possible using option 3.

Thanks,
Jonathan

#63 lore raw

Re: [PATCH] mergetool: do not enable hideResolved by default

In reply to Jonathan Nieder
Quoted 7 lines · Jonathan Nieder
git-mergetool.sh +0 -0
index 911470a5b2..f751d9cfe2 100755
@@ -358,13 +358,8 @@ merge_file () {
 		    enabled=false
 		fi

git-mergetool.sh merge_file () { · line 362

Quoted 10 lines
git-mergetool.sh +2 -7
 	else
-		# The user does not have a preference. Ask the tool.
-		if hide_resolved_enabled
-		then
-		    enabled=true
-		else
-		    enabled=false
-		fi
+		# The user does not have a preference. Default to disabled.
+		enabled=false

OK. So the logic used to be

- If the user has preference for a specific backend, use it;

- If the user says the feature is unwanted, that is final;

- If the user says it generally is OK to use the feature, let each backend set the preference;

- If there is no preference, let each backend set the preference.

As we want to disable the feature for any backend when the user does not explicitly say the feature is wanted (either in general, or for a specific backend), the change in the above hunk is exactly want we want to see.

Looking good. Let's not revert the series and disable by default.

Should I expect an updated log message, though? What was in the proposed log message sounded more unsubstantiated complaint than giving readable reasons why the feature is unwanted, but both the response by Seth and your response to Seth's response had material that made it more convincing why we would want to disable this by default, e.g. "with little to no explanation", "We don't have a way to communicate to the end-user" (both by Seth), "when ... didn't end up lining up the files correctly", "no way to visually distinguish" (yours) are all good ingredients to explain why this feature is prone to subtly and silently give wrong information to the end-users.

Thanks.

#64 lore raw

Re: [PATCH] mergetool: do not enable hideResolved by default

Quoted 7 lines · Junio C Hamano

As we want to disable the feature for any backend when the user does not explicitly say the feature is wanted (either in general, or for a specific backend), the change in the above hunk is exactly want we want to see.

Looking good. Let's not revert the series and disable by default.

Quoted 11 lines

Should I expect an updated log message, though? What was in the proposed log message sounded more unsubstantiated complaint than giving readable reasons why the feature is unwanted, but both the response by Seth and your response to Seth's response had material that made it more convincing why we would want to disable this by default, e.g. "with little to no explanation", "We don't have a way to communicate to the end-user" (both by Seth), "when ... didn't end up lining up the files correctly", "no way to visually distinguish" (yours) are all good ingredients to explain why this feature is prone to subtly and silently give wrong information to the end-users.

For tonight's pushout, I'll use the patch as-is and merge it in 'seen'.

Thanks.

#65 lore raw

Re: [PATCH] mergetool: do not enable hideResolved by default

Quoted 21 lines · Junio C Hamano
Quoted 7 lines · Junio C Hamano

As we want to disable the feature for any backend when the user does not explicitly say the feature is wanted (either in general, or for a specific backend), the change in the above hunk is exactly want we want to see.

Looking good. Let's not revert the series and disable by default.

Quoted 11 lines

Should I expect an updated log message, though? What was in the proposed log message sounded more unsubstantiated complaint than giving readable reasons why the feature is unwanted, but both the response by Seth and your response to Seth's response had material that made it more convincing why we would want to disable this by default, e.g. "with little to no explanation", "We don't have a way to communicate to the end-user" (both by Seth), "when ... didn't end up lining up the files correctly", "no way to visually distinguish" (yours) are all good ingredients to explain why this feature is prone to subtly and silently give wrong information to the end-users.

Quoted 2 lines

For tonight's pushout, I'll use the patch as-is and merge it in 'seen'.

Any progress here?

#66 lore raw

Re: [PATCH] mergetool: do not enable hideResolved by default

Quoted 11 lines · Junio C Hamano
Quoted 10 lines · Junio C Hamano
Quoted 8 lines · Junio C Hamano

As we want to disable the feature for any backend when the user does not explicitly say the feature is wanted (either in general, or for a specific backend), the change in the above hunk is exactly want we want to see.

Looking good. Let's not revert the series and disable by default.

Should I expect an updated log message, though?

[...]

Quoted 4 lines
Quoted 2 lines

For tonight's pushout, I'll use the patch as-is and merge it in 'seen'.

Any progress here?

Sorry for the delay. I should be able to send out an improved log message (more concise and summarizing the supporting info from this thread) later this afternoon.

Thanks,
Jonathan

#67 lore raw

Re: [PATCH] mergetool: do not enable hideResolved by default

Quoted 5 lines · Jonathan Nieder
Quoted 1 line

Any progress here?

Sorry for the delay. I should be able to send out an improved log message (more concise and summarizing the supporting info from this thread) later this afternoon.

Thanks. I think this is the last known regression in the -rc, and an update before the final happens on coming Monday is very much appreciated.

#68 lore raw

v2 cover mergetool: do not enable hideResolved by default

Quoted 9 lines · Junio C Hamano
Quoted 3 lines · Jonathan Nieder

I should be able to send out an improved log
message (more concise and summarizing the supporting info from this thread) later this afternoon.

Thanks. I think this is the last known regression in the -rc, and an update before the final happens on coming Monday is very much appreciated.

A little late, but here it is. Thoughts of all kinds welcome, as always.

Jonathan Nieder (2):
mergetool: do not enable hideResolved by default
doc: describe mergetool configuration in git-mergetool page

 Documentation/config/mergetool.txt | 2 +-
 Documentation/git-mergetool.txt    | 4 ++++
 git-mergetool.sh                   | 9 ++-------
 3 files changed, 7 insertions(+), 8 deletions(-)
#69 lore raw

PATCH v2 1/2 mergetool: do not enable hideResolved by default

When 98ea309b3f (mergetool: add hideResolved configuration, 2021-02-09) introduced the mergetool.hideResolved setting to reduce the clutter in viewing non-conflicted sections of files in a mergetool, it enabled it by default, explaining:

No adverse effects were noted in a small survey of popular mergetools[1] so this behavior defaults to true.

In practice, alas, adverse effects do appear. A few issues:

1. No indication is shown in the UI that the base, local, and remote versions shown have been modified by additional resolution. This is inherent in the design: the idea of mergetool.hideResolved is to convince a mergetool that expects pristine local, base, and remote files to show partially resolved verisons of those files instead; there is no additional source of information accessible to the mergetool to see where the resolution has happened.

(By contrast, a mergetool generating the partial resolution from conflict markers for itself would be able to hilight the resolved sections with a different color.)

A user accustomed to seeing the files without partial resolution gets no indication that this behavior has changed when they upgrade Git.

2. If the computed merge did not line up the files correctly (for example due to repeated sections in the file), the partially resolved files can be misleading and do not have enough information to reconstruct what happened and compute the correct merge result.

3. Resolving a conflict can involve information beyond the textual conflict. For example, if the local and remote versions added overlapping functionality in different ways, seeing the full unresolved versions of each alongside the base gives information about each side's intent that makes it possible to come up with a resolution that combines those two intents. By contrast, when starting with partially resolved versions of those files, one can produce a subtly wrong resolution that includes redundant extra code added by one side that is not needed in the approach taken on the other.

All that said, a user wanting to focus on textual conflicts with reduced clutter can still benefit from mergetool.hideResolved=true as a way to deemphasize sections of the code that resolve cleanly without requiring any changes to the invoked mergetool. The caveats described above are reduced when the user has explicitly turned this on, because then the user is aware of them.

Flip the default to 'false'.

---
Only difference from v1 is the commit message.

 Documentation/config/mergetool.txt | 2 +-
 git-mergetool.sh                   | 9 ++-------
 2 files changed, 3 insertions(+), 8 deletions(-)
Documentation/config/mergetool.txt +1 -1
index 90f76f5b9ba..cafbbef46ae 100644
@@ -53,7 +53,7 @@ mergetool.hideResolved::
 	resolution. This flag causes 'LOCAL' and 'REMOTE' to be overwriten so
 	that only the unresolved conflicts are presented to the merge tool. Can
 	be configured per-tool via the `mergetool.<tool>.hideResolved`
-	configuration variable. Defaults to `true`.
+	configuration variable. Defaults to `false`.
 
 mergetool.keepBackup::
 	After performing a merge, the original file with conflict markers
git-mergetool.sh +2 -7
index 911470a5b2c..f751d9cfe20 100755
@@ -358,13 +358,8 @@ merge_file () {
 		    enabled=false
 		fi
 	else
-		# The user does not have a preference. Ask the tool.
-		if hide_resolved_enabled
-		then
-		    enabled=true
-		else
-		    enabled=false
-		fi
+		# The user does not have a preference. Default to disabled.
+		enabled=false
 	fi
 
 	if test "$enabled" = true
Signature · 1 line
2.31.0.rc2.261.g7f71774620
#70 lore raw

PATCH v2 2/2 doc: describe mergetool configuration in git-mergetool(1)

In reply to Jonathan Nieder

In particular, this describes mergetool.hideResolved, which can help users discover this setting (either because it may be useful to them or in order to understand mergetool's behavior if they have forgotten setting it in the past).

Tested by running

make -C Documentation git-mergetool.1
man Documentation/git-mergetool.1

and reading through the page.

---
 Documentation/git-mergetool.txt | 4 ++++
 1 file changed, 4 insertions(+)

Thanks for reading.

Documentation/git-mergetool.txt +4 -0
index 6b14702e784..e587c7763a7 100644
@@ -99,6 +99,10 @@ success of the resolution after the custom tool has exited.
 	(see linkgit:git-config[1]).  To cancel `diff.orderFile`,
 	use `-O/dev/null`.
 
+CONFIGURATION
+-------------
+include::config/mergetool.txt[]
+
 TEMPORARY FILES
 ---------------
 `git mergetool` creates `*.orig` backup files while resolving merges.
Signature · 1 line
2.31.0.rc2.261.g7f71774620
#71 lore raw

Re: [PATCH 2/2] doc: describe mergetool configuration in git-mergetool(1)

Documentation/git-mergetool.txt success of the resolution after the custom tool has exited. · line 105

Quoted 12 lines · Jonathan Nieder
Documentation/git-mergetool.txt +4 -0
index 6b14702e784..e587c7763a7 100644
@@ -99,6 +99,10 @@ success of the resolution after the custom tool has exited.
 	(see linkgit:git-config[1]).  To cancel `diff.orderFile`,
 	use `-O/dev/null`.
 
+CONFIGURATION
+-------------
+include::config/mergetool.txt[]
+

It is a nice touch. We don't have much explanation other than the description below ...

mergetool.hideResolved::
During a merge Git will automatically resolve as many conflicts as possible and write the 'MERGED' file containing conflict markers around any conflicts that it cannot resolve; 'LOCAL' and 'REMOTE' normally represent the versions of the file from before Git's conflict resolution. This flag causes 'LOCAL' and 'REMOTE' to be overwriten so that only the unresolved conflicts are presented to the merge tool. Can be configured per-tool via the mergetool.<tool>.hideResolved configuration variable. Defaults to false.

... which appears in the included file on the feature.

Thanks.

#72 lore raw

Re: [PATCH 2/2] doc: describe mergetool configuration in git-mergetool(1)

In reply to Jonathan Nieder
Quoted 6 lines · Jonathan Nieder

Tested by running

make -C Documentation git-mergetool.1
man Documentation/git-mergetool.1

and reading through the page.

Nice. Also applying this step and running

	cd Documentation && ./doc-diff HEAD^ HEAD

would was a trivial way to see the change ;-)

#73 lore raw

PATCH v11 2/3 mergetool: break setup_tool out into separate initialization function

In reply to Seth House

This is preparation for the following commit where we need to source the mergetool shell script to look for overrides before run_merge_tool is called. Previously run_merge_tool both sourced that script and invoked the mergetool.

In the case of the following commit, we need the result of the hide_resolved override, if present, before we actually run run_merge_tool.

The new initialize_merge_tool wrapper is exposed and documented as a public interface for consistency with the existing run_merge_tool which is also public. Although setup_tool could instead be exposed directly, the related setup_user_tool would probably also want to be elevated to match and this felt the cleanest to me.

---
 Documentation/git-mergetool--lib.txt | 4 ++++
 git-difftool--helper.sh              | 6 ++++++
 git-mergetool--lib.sh                | 7 ++++---
 git-mergetool.sh                     | 2 ++
 4 files changed, 16 insertions(+), 3 deletions(-)
Documentation/git-mergetool--lib.txt +4 -0
index 4da9d24096..3e8f59ac0e 100644
@@ -38,6 +38,10 @@ get_merge_tool_cmd::
 get_merge_tool_path::
 	returns the custom path for a merge tool.
 
+initialize_merge_tool::
+	bring merge tool specific functions into scope so they can be used or
+	overridden.
+
 run_merge_tool::
 	launches a merge tool given the tool name and a true/false
 	flag to indicate whether a merge base is present.
git-difftool--helper.sh +6 -0
index 46af3e60b7..992124cc67 100755
@@ -61,6 +61,9 @@ launch_merge_tool () {
 		export BASE
 		eval $GIT_DIFFTOOL_EXTCMD '"$LOCAL"' '"$REMOTE"'
 	else
+		initialize_merge_tool "$merge_tool"
+		# ignore the error from the above --- run_merge_tool
+		# will diagnose unusable tool by itself
 		run_merge_tool "$merge_tool"
 	fi
 }
@@ -79,6 +82,9 @@ if test -n "$GIT_DIFFTOOL_DIRDIFF"
 then
 	LOCAL="$1"
 	REMOTE="$2"
+	initialize_merge_tool "$merge_tool"
+	# ignore the error from the above --- run_merge_tool
+	# will diagnose unusable tool by itself
 	run_merge_tool "$merge_tool" false
 else
 	# Launch the merge tool on each path provided by 'git diff'
git-mergetool--lib.sh +4 -3
index 78f3647ed9..4a8e36c792 100644
@@ -250,6 +250,10 @@ trust_exit_code () {
 	fi
 }
 
+initialize_merge_tool () {
+	# Bring tool-specific functions into scope
+	setup_tool "$1" || return 1
+}
 
 # Entry point for running tools
 run_merge_tool () {
@@ -261,9 +265,6 @@ run_merge_tool () {
 	merge_tool_path=$(get_merge_tool_path "$1") || exit
 	base_present="$2"
 
-	# Bring tool-specific functions into scope
-	setup_tool "$1" || return 1
-
 	if merge_mode
 	then
 		run_merge_cmd "$1"
git-mergetool.sh +2 -0
index 40a103443d..e5eac935f3 100755
@@ -272,6 +272,8 @@ merge_file () {
 		ext=
 	esac
 
+	initialize_merge_tool "$merge_tool" || return
+
 	mergetool_tmpdir_init
 
 	if test "$MERGETOOL_TMPDIR" != "."
Signature · 1 line
2.29.2
#74 lore raw

PATCH v11 3/3 mergetool: add per-tool support and overrides for the hideResolved flag

In reply to Seth House

Add a per-tool override flag so that users may enable the flag for one tool and disable it for another by setting mergetool.<tool>.hideResolved to false.

In addition, the author or maintainer of a mergetool may optionally override the default hideResolved value for that mergetool. If the mergetools/<tool> shell script contains a hide_resolved_enabled function it will be called when the mergetool is invoked and the return value will be used as the default for the hideResolved flag.

    hide_resolved_enabled () {
        return 1
    }

Disabling may be desirable if the mergetool wants or needs access to the original, unmodified 'LOCAL' and 'REMOTE' versions of the conflicted file. For example:

- A tool may use a custom conflict resolution algorithm and prefer to ignore the results of Git's conflict resolution.
- A tool may want to visually compare/constrast the version of the file from before the merge (saved to 'LOCAL', 'REMOTE', and 'BASE') with Git's conflict resolution results (saved to 'MERGED').

---
 Documentation/config/mergetool.txt |  5 +++++
 git-mergetool--lib.sh              |  4 ++++
 git-mergetool.sh                   | 36 +++++++++++++++++++++++++++++-
 3 files changed, 44 insertions(+), 1 deletion(-)
Documentation/config/mergetool.txt +5 -0
index b858191970..90f76f5b9b 100644
@@ -13,6 +13,11 @@ mergetool.<tool>.cmd::
 	merged; 'MERGED' contains the name of the file to which the merge
 	tool should write the results of a successful merge.
 
+mergetool.<tool>.hideResolved::
+	Allows the user to override the global `mergetool.hideResolved` value
+	for a specific tool. See `mergetool.hideResolved` for the full
+	description.
+
 mergetool.<tool>.trustExitCode::
 	For a custom merge command, specify whether the exit code of
 	the merge command can be used to determine whether the merge was
git-mergetool--lib.sh +4 -0
index 4a8e36c792..542a6a75eb 100644
@@ -166,6 +166,10 @@ setup_tool () {
 		return 1
 	}
 
+	hide_resolved_enabled () {
+		return 0
+	}
+
 	translate_merge_tool_path () {
 		echo "$1"
 	}
git-mergetool.sh +35 -1
index e5eac935f3..911470a5b2 100755
@@ -333,7 +333,41 @@ merge_file () {
 	checkout_staged_file 2 "$MERGED" "$LOCAL"
 	checkout_staged_file 3 "$MERGED" "$REMOTE"
 
-	if test "$(git config --type=bool mergetool.hideResolved)" != "false"
+	# hideResolved preferences hierarchy.
+	global_config="mergetool.hideResolved"
+	tool_config="mergetool.${merge_tool}.hideResolved"
+
+	if enabled=$(git config --type=bool "$tool_config")
+	then
+		# The user has a specific preference for a specific tool and no
+		# other preferences should override that.
+		: ;
+	elif enabled=$(git config --type=bool "$global_config")
+	then
+		# The user has a general preference for all tools.
+		#
+		# 'true' means the user likes the feature so we should use it
+		# where possible but tool authors can still override.
+		#
+		# 'false' means the user doesn't like the feature so we should
+		# not use it anywhere.
+		if test "$enabled" = true && hide_resolved_enabled
+		then
+		    enabled=true
+		else
+		    enabled=false
+		fi
+	else
+		# The user does not have a preference. Ask the tool.
+		if hide_resolved_enabled
+		then
+		    enabled=true
+		else
+		    enabled=false
+		fi
+	fi
+
+	if test "$enabled" = true
 	then
 		hide_resolved
 	fi
Signature · 1 line
2.29.2
#75 lore raw

Re: [PATCH v11 0/3] mergetool: add hideResolved configuration (was automerge)

In reply to Seth House
Quoted 6 lines · Seth House

Changes since v10:
Seth House (3):
mergetool: add hideResolved configuration
mergetool: break setup_tool out into separate initialization function
mergetool: add per-tool support and overrides for the hideResolved
flag

Thanks for all these iterations. The resuling series looks good to me.

#76 lore raw

Re: [PATCH v11 0/3] mergetool: add hideResolved configuration (was automerge)

Quoted 2 lines · Junio C Hamano

Thanks for all these iterations. The resuling series looks good to me.

Woot! Thanks for all the great feedback and for walking me through the process. I'm really happy with where this ended up.

#77 lore raw

Re: [PATCH v7 0/2] mergetool: add automerge configuration

In reply to Seth House
Quoted 3 lines · Seth House

* Signed off on Felipe's commit. (Although I have minor qualms with Felipe's various wording and even the name of the flag it is decidedly not worth burdening the list with bike-shedding.)

Even when the original is a horrible patch in your opinion that is laden with bugs, as long as the original author signed it off (which means that the original author certifies that it can be included in and distributed by the project under our licensing terms, and agrees to the fact that the original author did so will be recorded in perpetuity), you can relay such a patch as-is, and you are required (i.e. SubmittingPatches is pretty clear that without your sign-off we cannot accept) to sign it off to record the provenance of the code.

The other side of the above coin is that you are not endorsing or vounching for the patch when you sign it off, so your name is not smudged by wording and flag name chosen in a way that you may consider poor. So "Although..." part is not a good objection against signing it off.

In other words, sign-off is not about assuring quality.

Also, instead of relaying as-is, you can relay a patch with your improvements rolled into the same patch (i.e. not as follow-up fixes). Some (or major) parts of the original patch may still remain in the edited result and you'd need to keep original author's sign-off as-is [*1*].

In this topic's case, 2/2 would be a feature enhancement on top of 1/2, so relaying 1/2 as-is would be OK, but in a case where an promising patch was sent with sign-off and bugs, then gets abandoned by the original author, fixing the bug in the patch you relay in place (i.e. not as follow-up patches) may even be necessary to keep bisectability. When you do so, you'd typically do:

Subject: [PATCH] title of the patch

... original author's log message, possibly copyedited ... by you

+ <Comment on what you did on top of the original can come here>

+ [or brief comment here]
+ Signed-off-by: Your Name <[email protected]>

(1) add your sign-off at the end
(2) explain what you changed relative to the original, either inside [] on the line before your sign-off, or at the end of the log message proper.

to indicate that it is not relayed as-is; this allows you to take responsibility of an unintended breakage your "fixes" might have caused.

[Footnote]

*1* The result may become something that no longer aligns the original author's opinion, but that is OK. The sign-off by the original author just says that the original author has the right to contribute (the remaining part of) the patch and the original author agrees that the record of author's involvement in the patch (including sign-off) will be kept.

It is not about assuring quality of the final work by the original author, either.

#78 lore raw

Re: [PATCH v7 0/2] mergetool: add automerge configuration

Quoted 5 lines · Junio C Hamano

Also, instead of relaying as-is, you can relay a patch with your improvements rolled into the same patch (i.e. not as follow-up fixes). Some (or major) parts of the original patch may still remain in the edited result and you'd need to keep original author's sign-off as-is [*1*].

Yes, you *can*, but doing so in this case would be against the author's wishes, and a violation of the Developer Certificate of Origin.

Quoted 6 lines

In this topic's case, 2/2 would be a feature enhancement on top of 1/2, so relaying 1/2 as-is would be OK, but in a case where an promising patch was sent with sign-off and bugs, then gets abandoned by the original author, fixing the bug in the patch you relay in place (i.e. not as follow-up patches) may even be necessary to keep bisectability. When you do so, you'd typically do:

If the author doesn't object (which is usually the case), this makes sense.

But if the author objects, you would be violating clause (d) of the DCO.

Just because the only way to do X is to violate laws, terms, or agreements doesn't mean that's what you should do. You can simply not do X.

Signature · 1 line
Felipe Contreras
#79 lore raw

RE: [PATCH v6 0/1] mergetool: add automerge configuration

In reply to Seth House
Quoted 3 lines · Seth House

Sorry for the slow turnaround on this. I haven't used Git via email patches before now so it took me quite a few hours to read through tutorials, configure git-send-email and fight missing Perl libs.

What distribution are you using?

Quoted 1 line

Changes since v5:

This is not v6 of my patch series; it's v1 of yours, which I think should have a different title.

What happens when I want to do v6?

Other than that (and the fact that you initially used the wrong version as a baseline), I'm fine with your approach of doing a patch on top of mine.

Cheers.

Signature · 1 line
Felipe Contreras