Better support for customising context lines in --patch commands
#1915
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Welcome to GitGitGadgetHi @NinjaInShade, and welcome to GitGitGadget, the GitHub App to send patch series to the Git mailing list from GitHub Pull Requests. Please make sure that either:
You can CC potential reviewers by adding a footer to the PR description with the following syntax: NOTE: DO NOT copy/paste your CC list from a previous GGG PR's description, Also, it is a good idea to review the commit messages one last time, as the Git project expects them in a quite specific form:
It is in general a good idea to await the automated test ("Checks") in this Pull Request before contributing the patches, e.g. to avoid trivial issues such as unportable code. Contributing the patchesBefore you can contribute the patches, your GitHub username needs to be added to the list of permitted users. Any already-permitted user can do that, by adding a comment to your PR of the form Both the person who commented An alternative is the channel Once on the list of permitted usernames, you can contribute the patches to the Git mailing list by adding a PR comment If you want to see what email(s) would be sent for a After you submit, GitGitGadget will respond with another comment that contains the link to the cover letter mail in the Git mailing list archive. Please make sure to monitor the discussion in that thread and to address comments and suggestions (while the comments and suggestions will be mirrored into the PR by GitGitGadget, you will still want to reply via mail). If you do not want to subscribe to the Git mailing list just to be able to respond to a mail, you can download the mbox from the Git mailing list archive (click the curl -g --user "<EMailAddress>:<Password>" \
--url "imaps://imap.gmail.com/INBOX" -T /path/to/raw.txtTo iterate on your change, i.e. send a revised patch or patch series, you will first want to (force-)push to the same branch. You probably also want to modify your Pull Request description (or title). It is a good idea to summarize the revision by adding something like this to the cover letter (read: by editing the first comment on the PR, i.e. the PR description): To send a new iteration, just add another PR comment with the contents: Need help?New contributors who want advice are encouraged to join [email protected], where volunteers who regularly contribute to Git are willing to answer newbie questions, give advice, or otherwise provide mentoring to interested contributors. You must join in order to post or view messages, but anyone can join. You may also be able to find help in real time in the developer IRC channel, |
NinjaInShade
force-pushed
the
interactive-patch-context
branch
2 times, most recently
from
0484407 to
ff64c39
Compare
|
/allow |
|
User NinjaInShade is now allowed to use GitGitGadget. WARNING: NinjaInShade has no public email address set on GitHub; GitGitGadget needs an email address to Cc: you on your contribution, so that you receive any feedback on the Git mailing list. Go to https://github.com/settings/profile to make your preferred email public to let GitGitGadget know which email address to use. |
NinjaInShade
changed the title
--patch commands--patch commands
NinjaInShade
force-pushed
the
interactive-patch-context
branch
2 times, most recently
from
9b90dc6 to
b4b7854
Compare
|
/submit |
|
Submitted as [email protected] To fetch this version into To fetch this version to local tag |
![gitgitgadget[bot]](https://avatars.githubusercontent.com/in/12836?s=60&v=4){kind=small}
|
User |
|
User |
|
Thanks for letting me know, I'll make sure to amend that in v2 :)
…On Mon, 5 May 2025 at 21:33, gitgitgadget[bot] ***@***.***> wrote:
*gitgitgadget[bot]* left a comment (gitgitgadget/git#1915)
<#1915 (comment)>
User Eric Sunshine ***@***.***> has been added to the cc:
list.
—
Reply to this email directly, view it on GitHub
<#1915 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/APFXCEW367PGDBXHMJDZTHT247DJ7AVCNFSM6AAAAAB4M4W3LOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDQNJSGI3DIOBUGQ>
.
You are receiving this because you were mentioned.Message ID:
***@***.***>
|
|
User |
|
User |
NinjaInShade
force-pushed
the
interactive-patch-context
branch
from
b4b7854 to
cf726da
Compare
Refactor to use the modern "test_grep" test utility instead of regular "grep" which provides better debug information if tests fail. This is a prerequisite to the commits that follow which add to both test files. Signed-off-by: Leon Michalak <[email protected]>
Refactor to use the modern "test_config" test utility instead of manual "git config" as the former provides clean up on test completion. This is a prerequisite to the commits that follow which add to this test file. Signed-off-by: Leon Michalak <[email protected]>
Various builtins that use add-patch infrastructure do not respect the user's diff.context and diff.interHunkContext file configurations. This patch fixes this inconsistency. Signed-off-by: Leon Michalak <[email protected]>
This patch compliments the previous commit, where builtins that use add-patch infrastructure now respect diff.context and diff.interHunkContext file configurations. In particular, this patch helps users who don't want to set persistent context configurations or just want a way to override them on a one-time basis, by allowing the relevant builtins to accept corresponding command line options that override the file configurations. This mimics commands such as diff and log, which allow for both context file configuration and command line overrides. Signed-off-by: Leon Michalak <[email protected]>
NinjaInShade
force-pushed
the
interactive-patch-context
branch
from
cf726da to
973dfad
Compare
|
/submit |
|
Submitted as [email protected] To fetch this version into To fetch this version to local tag |
| @@ -63,7 +63,7 @@ test_expect_success 'setup (initial)' ' | |||
| ' | |||
There was a problem hiding this comment.
On the Git mailing list, Junio C Hamano wrote (reply to this):
"Leon Michalak via GitGitGadget" <[email protected]> writes:
> From: Leon Michalak <[email protected]>
>
> Refactor to use the modern "test_grep" test utility instead of regular
> "grep" which provides better debug information if tests fail.
>
> This is a prerequisite to the commits that follow which add to both test
> files.
>
> Signed-off-by: Leon Michalak <[email protected]>
> ---
These mostly look sensible, but I would title & phrase the commit
description to 'use "test_grep"', not 'refactor to &'. It's shorter
and more direct ;-)
Thanks.There was a problem hiding this comment.
On the Git mailing list, Leon Michalak wrote (reply to this):
On Mon, 12 May 2025 at 14:42, Junio C Hamano <[email protected]> wrote:
> These mostly look sensible, but I would title & phrase the commit
> description to 'use "test_grep"', not 'refactor to &'. It's shorter
> and more direct ;-)
Thanks - will make sure to change that in v3 :)
| @@ -0,0 +1,10 @@ | |||
| `-U<n>`:: | |||
There was a problem hiding this comment.
On the Git mailing list, Junio C Hamano wrote (reply to this):
"Leon Michalak via GitGitGadget" <[email protected]> writes:
> diff --git a/Documentation/diff-context-options.adoc b/Documentation/diff-context-options.adoc
> new file mode 100644
> index 000000000000..e161260358ff
> --- /dev/null
> +++ b/Documentation/diff-context-options.adoc
> @@ -0,0 +1,10 @@
> +`-U<n>`::
> +`--unified=<n>`::
> + Generate diffs with _<n>_ lines of context. Defaults to `diff.context`
> + or 3 if the config option is unset.
> +
> +`--inter-hunk-context=<n>`::
> + Show the context between diff hunks, up to the specified _<number>_
> + of lines, thereby fusing hunks that are close to each other.
> + Defaults to `diff.interHunkContext` or 0 if the config option
> + is unset.
It might not be trivial to do but I wonder if we cannot do better
than this to share more of the same text across manual pages. These
two being options understood by `git diff`, we certainly have an
existing description for them, no?
Other than that, looking good to me.
Thanks.There was a problem hiding this comment.
On the Git mailing list, Leon Michalak wrote (reply to this):
On Mon, 12 May 2025 at 17:45, Junio C Hamano <[email protected]> wrote:
> It might not be trivial to do but I wonder if we cannot do better
> than this to share more of the same text across manual pages. These
> two being options understood by `git diff`, we certainly have an
> existing description for them, no?
Yes, I did of course notice documentation for `git diff` also has
these; ultimately my justification for not changing that to use this
new .adoc include as well was for a couple reasons:
- these two options are not together in the `git diff` documentation
(not *so* important, and they probably should actually be together?)
- there is an extra if def which adds on "implies --patch" text in the
`git diff` documentation which isn't the behaviour the add-patch
commands are going for, so that makes the intent a little different
here
But would be good to hear if anyone else has any thoughts.
|
This patch series was integrated into seen via git@0945e76. |
|
This patch series was integrated into seen via git@3512d8e. |
|
This patch series was integrated into seen via git@d808817. |
add-interactive.c
Outdated
| @@ -39,8 +39,12 @@ static void init_color(struct repository *r, struct add_i_state *s, | |||
| void init_add_i_state(struct add_i_state *s, struct repository *r) | |||
There was a problem hiding this comment.
On the Git mailing list, Phillip Wood wrote (reply to this):
Hi Leon
On 10/05/2025 14:46, Leon Michalak via GitGitGadget wrote:
> From: Leon Michalak <[email protected]>
> > Various builtins that use add-patch infrastructure do not respect
> the user's diff.context and diff.interHunkContext file configurations.
We could expand this slightly by adding
This is because the plumbing commands used by "git add -p" to generate
the diff do not read those config settings. Fix this by reading the
config before generating the patch and passing it along to the diff
command with the "-U" and "--inter-hunk-context" command-line options.
> This patch fixes this inconsistency.
> > Signed-off-by: Leon Michalak <[email protected]>
> ---
> @@ -78,6 +82,19 @@ void init_add_i_state(struct add_i_state *s, struct repository *r)
> repo_config_get_string(r, "diff.algorithm",
> &s->interactive_diff_algorithm);
> > + if (!repo_config_get_int(r, "diff.context", &context)) {
> + if (context < 0)
> + die(_("%s cannot be negative"), "diff.context");
> + else
> + s->context = context;
> + };
> + if (!repo_config_get_int(r, "diff.interHunkContext", &interhunkcontext)) {
> + if (interhunkcontext < 0)
> + die(_("%s cannot be negative"), "diff.interHunkContext");
> + else
> + s->interhunkcontext = interhunkcontext;
> + };
Thanks for changing this. This iteration of the code changes looks good
> diff --git a/t/t4055-diff-context.sh b/t/t4055-diff-context.sh
> index 1384a8195705..c4b861c360cc 100755
> --- a/t/t4055-diff-context.sh
> +++ b/t/t4055-diff-context.sh
> @@ -52,6 +52,46 @@ test_expect_success 'diff.context honored by "log"' '
> test_grep "^ firstline" output
> '
It's great that you have written tests for this patch but as I said
last time I think the new tests should be in t3701-add-interactive.sh
as we're interested in testing whether "git add -p" passes on
diff.context to "git diff" , not whether "git diff" respects
diff.context. I still think there are too many tests here as we know
that all the different "-p" commands share a single code path. Our
test suite is slow enough already so we do not want to add new tests
that do not increase our code coverage. I would suggest removing
these tests and instead add the following in t3701
test_expect_success 'add -p respects diff.context' '
test_write_lines a b c d e f g h i j k l m >file &&
git add file &&
test_write_lines a b c d e f G h i j k l m >file &&
echo y | git -c diff.context=5 add -p >actual &&
test_grep "@@ -2,11 +2,11 @@" actual
'
test_expect_success 'add -p respects diff.interHunkContext' '
test_write_lines a b c d e f g h i j k l m n o p q r s >file &&
git add file &&
test_write_lines a b c d E f g i i j k l m N o p q r s >file &&
echo y | git -c diff.interhunkcontext=2 add -p >actual &&
test_grep "@@ -2,16 +2,16 @@" actual
'
> +test_expect_success 'negative integer config parsing by "add"' '
Perhaps "add -p rejects negative diff.context" would be clearer?
> + test_config diff.context -1 &&
> + test_must_fail git add -p 2>output &&
> + test_grep "diff.context cannot be negative" output
> +'
This is great but again we only need to test a single command and we
should do so in t3701. We should also check that negative values of
diff.interHunkContext are also rejected.
Best Wishes
PhillipThere was a problem hiding this comment.
On the Git mailing list, Junio C Hamano wrote (reply to this):
Phillip Wood <[email protected]> writes:
> Hi Leon
>
> On 10/05/2025 14:46, Leon Michalak via GitGitGadget wrote:
>> From: Leon Michalak <[email protected]>
>> Various builtins that use add-patch infrastructure do not respect
>> the user's diff.context and diff.interHunkContext file configurations.
>
> We could expand this slightly by adding
>
> This is because the plumbing commands used by "git add -p" to generate
> the diff do not read those config settings. Fix this by reading the
> config before generating the patch and passing it along to the diff
> command with the "-U" and "--inter-hunk-context" command-line options.
>
>> This patch fixes this inconsistency.
>> Signed-off-by: Leon Michalak <[email protected]>
>> ---
>
>> @@ -78,6 +82,19 @@ void init_add_i_state(struct add_i_state *s, struct repository *r)
>> repo_config_get_string(r, "diff.algorithm",
>> &s->interactive_diff_algorithm);
>> + if (!repo_config_get_int(r, "diff.context", &context)) {
>> + if (context < 0)
>> + die(_("%s cannot be negative"), "diff.context");
>> + else
>> + s->context = context;
>> + };
>> + if (!repo_config_get_int(r, "diff.interHunkContext", &interhunkcontext)) {
>> + if (interhunkcontext < 0)
>> + die(_("%s cannot be negative"), "diff.interHunkContext");
>> + else
>> + s->interhunkcontext = interhunkcontext;
>> + };
>
> Thanks for changing this. This iteration of the code changes looks good
Lose the ';' (semicolon) after closing {brace}s.
This is C; you do not need an empty statement after a {block}.
Everything in your review I am very happy to see. Thanks for giving
a great review.
There was a problem hiding this comment.
On the Git mailing list, Phillip Wood wrote (reply to this):
On 13/05/2025 16:47, Junio C Hamano wrote:
> Phillip Wood <[email protected]> writes:
>> On 10/05/2025 14:46, Leon Michalak via GitGitGadget wrote:
>>
>>> @@ -78,6 +82,19 @@ void init_add_i_state(struct add_i_state *s, struct repository *r)
>>> repo_config_get_string(r, "diff.algorithm",
>>> &s->interactive_diff_algorithm);
>>> + if (!repo_config_get_int(r, "diff.context", &context)) {
>>> + if (context < 0)
>>> + die(_("%s cannot be negative"), "diff.context");
>>> + else
>>> + s->context = context;
>>> + };
>>> + if (!repo_config_get_int(r, "diff.interHunkContext", &interhunkcontext)) {
>>> + if (interhunkcontext < 0)
>>> + die(_("%s cannot be negative"), "diff.interHunkContext");
>>> + else
>>> + s->interhunkcontext = interhunkcontext;
>>> + };
>>
>> Thanks for changing this. This iteration of the code changes looks good
> > Lose the ';' (semicolon) after closing {brace}s.
> This is C; you do not need an empty statement after a {block}.
Oh well spotted, I'd missed that
Thanks
PhillipThere was a problem hiding this comment.
On the Git mailing list, Junio C Hamano wrote (reply to this):
Phillip Wood <[email protected]> writes:
> On 13/05/2025 16:47, Junio C Hamano wrote:
>> Phillip Wood <[email protected]> writes:
>>> On 10/05/2025 14:46, Leon Michalak via GitGitGadget wrote:
>>>
>>>> @@ -78,6 +82,19 @@ void init_add_i_state(struct add_i_state *s, struct repository *r)
>>>> repo_config_get_string(r, "diff.algorithm",
>>>> &s->interactive_diff_algorithm);
>>>> + if (!repo_config_get_int(r, "diff.context", &context)) {
>>>> + if (context < 0)
>>>> + die(_("%s cannot be negative"), "diff.context");
>>>> + else
>>>> + s->context = context;
>>>> + };
>>>> + if (!repo_config_get_int(r, "diff.interHunkContext", &interhunkcontext)) {
>>>> + if (interhunkcontext < 0)
>>>> + die(_("%s cannot be negative"), "diff.interHunkContext");
>>>> + else
>>>> + s->interhunkcontext = interhunkcontext;
>>>> + };
>>>
>>> Thanks for changing this. This iteration of the code changes looks good
>> Lose the ';' (semicolon) after closing {brace}s.
>> This is C; you do not need an empty statement after a {block}.
>
> Oh well spotted, I'd missed that
Heh, with enough number of eyeballs, all the bugs are shallow.
| @@ -0,0 +1,10 @@ | |||
| `-U<n>`:: | |||
There was a problem hiding this comment.
On the Git mailing list, Phillip Wood wrote (reply to this):
Hi Leon
On 10/05/2025 14:46, Leon Michalak via GitGitGadget wrote:
> From: Leon Michalak <[email protected]>
> > This patch compliments the previous commit, where builtins that use
> add-patch infrastructure now respect diff.context and
> diff.interHunkContext file configurations.
> > In particular, this patch helps users who don't want to set persistent
> context configurations or just want a way to override them on a one-time
> basis, by allowing the relevant builtins to accept corresponding command
> line options that override the file configurations.
> > This mimics commands such as diff and log, which allow for both context
> file configuration and command line overrides.
The code changes here mostly look good, I've left a few comments
below. I think the tests could be improved, I've left some suggestions
on limiting the number of tests while improving the coverage. The new
tests I'm suggesting that check invalid option combinations are the
basis for most of my code comments.
There is still the issue of what to do with -U0. As I mentioned
previously "git apply" will fail when we try to apply the patch. We
can either pass the appropriate flag when the context is zero or
possibly use -U0 to mean the default number of context lines.
> diff --git a/Documentation/diff-context-options.adoc b/Documentation/diff-context-options.adoc
> new file mode 100644
> index 000000000000..e161260358ff
> --- /dev/null
> +++ b/Documentation/diff-context-options.adoc
> @@ -0,0 +1,10 @@
> +`-U<n>`::
> +`--unified=<n>`::
> + Generate diffs with _<n>_ lines of context. Defaults to `diff.context`
> + or 3 if the config option is unset.
> +
> +`--inter-hunk-context=<n>`::
> + Show the context between diff hunks, up to the specified _<number>_
> + of lines, thereby fusing hunks that are close to each other.
> + Defaults to `diff.interHunkContext` or 0 if the config option
> + is unset.
Nice - we reuse the same text for all the "-p" commands.
> diff --git a/add-interactive.c b/add-interactive.c
> [...]
> @@ -98,6 +99,17 @@ void init_add_i_state(struct add_i_state *s, struct repository *r)
> repo_config_get_bool(r, "interactive.singlekey", &s->use_single_key);
> if (s->use_single_key)
> setbuf(stdin, NULL);
> +
> + if (add_p_opt->context != -1) {
> + if (add_p_opt->context < 0)
> + die(_("%s cannot be negative"), "--unified");
> + s->context = add_p_opt->context;
> + }
> + if (add_p_opt->interhunkcontext != -1) {
> + if (add_p_opt->interhunkcontext < 0)
> + die(_("%s cannot be negative"), "--inter-hunk-context");
> + s->interhunkcontext = add_p_opt->interhunkcontext;
> + }
Centralizing these checks like this is a good idea.
> @@ -1031,10 +1047,13 @@ static int run_diff(struct add_i_state *s, const struct pathspec *ps,
> if (count > 0) {
> struct child_process cmd = CHILD_PROCESS_INIT;
> > - strvec_pushl(&cmd.args, "git", "diff", "-p", "--cached",
> - oid_to_hex(!is_initial ? &oid :
> - s->r->hash_algo->empty_tree),
> - "--", NULL);
> + strvec_pushl(&cmd.args, "git", "diff", "-p", "--cached", NULL);
> + if (s->context != -1)
> + strvec_pushf(&cmd.args, "--unified=%i", s->context);
> + if (s->interhunkcontext != -1)
> + strvec_pushf(&cmd.args, "--inter-hunk-context=%i", s->interhunkcontext);
> + strvec_pushl(&cmd.args, oid_to_hex(!is_initial ? &oid :
> + s->r->hash_algo->empty_tree), "--", NULL);
This is good - we propagate the values we were given on the
command-line.
> diff --git a/builtin/checkout.c b/builtin/checkout.c
> [...]
> @@ -564,8 +575,13 @@ static int checkout_paths(const struct checkout_opts *opts,
> else
> BUG("either flag must have been set, worktree=%d, index=%d",
> opts->checkout_worktree, opts->checkout_index);
> - return !!run_add_p(the_repository, patch_mode, rev,
> - &opts->pathspec);
> + return !!run_add_p(the_repository, patch_mode, &add_p_opt,
> + rev, &opts->pathspec);
> + } else {
> + if (opts->patch_context != -1)
> + die(_("the option '%s' requires '%s'"), "--unified", "--patch");
> + if (opts->patch_interhunk_context != -1)
> + die(_("the option '%s' requires '%s'"), "--inter-hunk-context", "--patch");
> }
This does not catch "git checkout -U 7" because this code is only run
if we're checking out paths. I think you need to check this is
checkout_main() instead.
> diff --git a/builtin/stash.c b/builtin/stash.c
> [...]
> @@ -1826,8 +1831,15 @@ static int push_stash(int argc, const char **argv, const char *prefix,
> die(_("the option '%s' requires '%s'"), "--pathspec-file-nul", "--pathspec-from-file");
> }
> > + if (!patch_mode) {
> + if (add_p_opt.context != -1)
> + die(_("the option '%s' requires '%s'"), "--unified", "--patch");
> + if (add_p_opt.interhunkcontext != -1)
> + die(_("the option '%s' requires '%s'"), "--inter-hunk-context", "--patch");
> + }
> +
This needs to die on invalid context values as "git stash" seems to
ignore the exit code of the subprocess that checks for negative values.
> @@ -1877,8 +1892,17 @@ static int save_stash(int argc, const char **argv, const char *prefix,
> stash_msg = strbuf_join_argv(&stash_msg_buf, argc, argv, ' ');
> > memset(&ps, 0, sizeof(ps));
> +
> + if (!patch_mode) {
> + if (add_p_opt.context != -1)
> + die(_("the option '%s' requires '%s'"), "--unified", "--patch");
> + if (add_p_opt.interhunkcontext != -1)
> + die(_("the option '%s' requires '%s'"), "--inter-hunk-context", "--patch");
This needs to die on invalid context values as "git stash" seems to
ignore the exit code of the subprocess that checks for negative values.
> diff --git a/commit.h b/commit.h
> [...]
> #include "object.h"
> +#include "add-interactive.h"
> struct signature_check;
> struct strbuf;
Lets not add this. Instead lets just add a declaration for "struct
add_p_opt" like the ones in the context line so that we don't end up
including everything from add-interactive.h when we only need a single
struct declaration.
> diff --git a/parse-options.h b/parse-options.h
> [...]
> +#define OPT_DIFF_UNIFIED(v) OPT_INTEGER_F('U', "unified", v,
> N_("generate diffs with <n> lines context"), PARSE_OPT_NONEG)
This looks good
> +#define OPT_DIFF_INTERHUNK_CONTEXT(v) OPT_INTEGER_F(0, "inter-hunk-context", v, N_("show context between diff hunks up to the specified number of lines"), PARSE_OPT_NONEG)
This is a bit verbose but it matches what is in diff.c.
> diff --git a/t/t4032-diff-inter-hunk-context.sh b/t/t4032-diff-inter-hunk-context.sh
> index bada0cbd32f7..d5aad6e143a7 100755
> --- a/t/t4032-diff-inter-hunk-context.sh
> +++ b/t/t4032-diff-inter-hunk-context.sh
> @@ -47,6 +47,31 @@ t() {
> "
> }
> > +t_patch() {
> + use_config=
> + git config --unset diff.interHunkContext
> +
> + case $# in
> + 4) hunks=$4; cmd="add -p -U$3";;
> + 5) hunks=$5; cmd="add -p -U$3 --inter-hunk-context=$4";;
> + 6) hunks=$5; cmd="add -p -U$3"; git config diff.interHunkContext $4; use_config="(diff.interHunkContext=$4) ";;
> + esac
> + label="$use_config$cmd, $1 common $2"
> + file=f$1
> +
> + if ! test -f $file
> + then
> + f A $1 B >$file
> + git add $file
> + git commit -q -m. $file
> + f X $1 Y >$file
> + fi
> +
> + test_expect_success "$label: count hunks ($hunks)" "
> + test $(test_write_lines q | git $cmd $file | sed -n 's/^([0-9]*\/\([0-9]*\)) Stage this hunk.*/\1/p') = $hunks
> + "
> +}
> +
> cat <<EOF >expected.f1.0.1 || exit 1
> diff --git a/f1 b/f1
> --- a/f1
> @@ -107,6 +132,42 @@ t 3 lines 1 2 1 config
> t 9 lines 3 2 2 config
> t 9 lines 3 3 1 config
> > +# common lines ctx intrctx hunks
> +t_patch 1 line 0 2
> +t_patch 1 line 0 0 2
> +t_patch 1 line 0 1 1
> +t_patch 1 line 0 2 1
> +t_patch 1 line 1 1
> +
> +t_patch 2 lines 0 2
> +t_patch 2 lines 0 0 2
> +t_patch 2 lines 0 1 2
> +t_patch 2 lines 0 2 1
> +t_patch 2 lines 1 1
> +
> +t_patch 3 lines 1 2
> +t_patch 3 lines 1 0 2
> +t_patch 3 lines 1 1 1
> +t_patch 3 lines 1 2 1
> +
> +t_patch 9 lines 3 2
> +t_patch 9 lines 3 2 2
> +t_patch 9 lines 3 3 1
> +
> +# use diff.interHunkContext?
> +t_patch 1 line 0 0 2 config
> +t_patch 1 line 0 1 1 config
> +t_patch 1 line 0 2 1 config
> +t_patch 9 lines 3 3 1 config
> +t_patch 2 lines 0 0 2 config
> +t_patch 2 lines 0 1 2 config
> +t_patch 2 lines 0 2 1 config
> +t_patch 3 lines 1 0 2 config
> +t_patch 3 lines 1 1 1 config
> +t_patch 3 lines 1 2 1 config
> +t_patch 9 lines 3 2 2 config
> +t_patch 9 lines 3 3 1 config
> +
There are 29 tests here and yet more below. I think we can
get the test coverage we need much more efficiently with the following
added to t3701
for cmd in add checkout restore 'commit -m file'
do
test_expect_success "${cmd%% *} accepts -U and --inter-hunk-context" "
test_write_lines a b c d e f g h i j k l m n o p q r s t u v >file &&
git add file &&
test_write_lines a b c d e F g h i j k l m n o p Q r s t u v >file &&
echo y | git -c diff.context=5 -c diff.interhunkcontext=1 \
$cmd -p -U 4 --inter-hunk-context 2 >actual &&
test_grep \"@@ -2,20 +2,20 @@\" actual
"
done
test_expect_success 'reset accepts -U and --inter-hunk-context' '
test_write_lines a b c d e f g h i j k l m n o p q r s t u v >file &&
git commit -m file file &&
test_write_lines a b c d e F g h i j k l m n o p Q r s t u v >file &&
git add file &&
echo y | git -c diff.context=5 -c diff.interhunkcontext=1 \
reset -p -U 4 --inter-hunk-context 2 >actual &&
test_grep "@@ -2,20 +2,20 @@" actual
'
test_expect_success 'stash accepts -U and --inter-hunk-context' '
test_write_lines a b c d e F g h i j k l m n o p Q r s t u v >file &&
git commit -m file file &&
test_write_lines a b c d e f g h i j k l m n o p q r s t u v >file &&
echo y | git -c diff.context=5 -c diff.interhunkcontext=1 \
stash -p -U 4 --inter-hunk-context 2 >actual &&
test_grep "@@ -2,20 +2,20 @@" actual
'
Those tests will fail if any of the commands that accept "-p" do not
accept "-U" or "--inter-hunk-context" or if command-line arguments do
not override the config settings. We should also add tests in t3701 to
check that invalid option combinations and values are rejected like so
for cmd in add checkout commit reset restore 'stash save' 'stash push'
do
test_expect_success "$cmd rejects invalid context options" "
test_must_fail git $cmd -p -U -3 2>actual &&
test_grep -e \"--unified cannot be negative\" actual &&
test_must_fail git $cmd -p --inter-hunk-context -3 2>actual &&
test_grep -e \"--inter-hunk-context cannot be negative\" actual &&
test_must_fail git $cmd -U 7 2>actual &&
test_grep -E \".--unified. requires .(--interactive/)?--patch.\" actual &&
test_must_fail git $cmd --inter-hunk-context 2 2>actual &&
test_grep -E \".--inter-hunk-context. requires .(--interactive/)?--patch.\" actual
"
done
The "checkout", "stash save" and "stash push" tests above currently
fail because the implementation does not implement those checks
properly.
With a few tweaks this series will be looking very good
Best Wishes
PhillipThere was a problem hiding this comment.
On the Git mailing list, Phillip Wood wrote (reply to this):
On 13/05/2025 14:52, Phillip Wood wrote:
>> diff --git a/builtin/stash.c b/builtin/stash.c
>> [...]
>> @@ -1826,8 +1831,15 @@ static int push_stash(int argc, const char >> **argv, const char *prefix,
>> die(_("the option '%s' requires '%s'"), "--pathspec-file- >> nul", "--pathspec-from-file");
>> }
>> + if (!patch_mode) {
>> + if (add_p_opt.context != -1)
>> + die(_("the option '%s' requires '%s'"), "--unified", "-- >> patch");
>> + if (add_p_opt.interhunkcontext != -1)
>> + die(_("the option '%s' requires '%s'"), "--inter-hunk- >> context", "--patch");
>> + }
>> +
> > This needs to die on invalid context values as "git stash" seems to
> ignore the exit code of the subprocess that checks for negative values.
Looking more closely the problem is that it quits if there are no changes to stash before validating -U or --inter-hunk-context. I think it should validate the options before checking if there is anything to stash.
Best Wishes
Phillip
There was a problem hiding this comment.
On the Git mailing list, Leon Michalak wrote (reply to this):
Hey, thanks for the thorough review Philip. I will properly digest
this when I get some free time, but I just wanted to say (I probably
should have mentioned this so my bad) that the reason I didn't change
to test just the singular command (yet, anyway) is that someone else
thought this was a good idea testing all of them, so I wasn't sure
whether to touch it or not in the end, and thought I'd just submit
this v2 and gather more opinions. Was this perhaps the wrong approach
though?There was a problem hiding this comment.
On the Git mailing list, [email protected] wrote (reply to this):
Hi Leon
On 13/05/2025 16:05, Leon Michalak wrote:
> Hey, thanks for the thorough review Philip. I will properly digest
> this when I get some free time, but I just wanted to say (I probably
> should have mentioned this so my bad) that the reason I didn't change
> to test just the singular command (yet, anyway) is that someone else
> thought this was a good idea testing all of them,
I'd missed that message - have you got a link to it please
> so I wasn't sure
> whether to touch it or not in the end, and thought I'd just submit
> this v2 and gather more opinions. Was this perhaps the wrong approach
> though?
If you get conflicting advise then it is a good idea to mention that in the cover letter and explain which option you went with and why.
Best Wishes
Phillip
|
This branch is now known as |
|
This patch series was integrated into seen via git@499dc4d. |
|
This patch series was integrated into seen via git@cc0cba3. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This series of patches attempt to give
--interactive/--patchcompatible builtins ("add", "commit", "checkout", "reset", "restore" and "stash") better support and nicer experience for configuring how many context lines are shown in diffs through a variety of ways.Prior to these patches, the user could not choose how many context lines they saw in
--patchcommands (apart from one workaround by usingGIT_DIFF_OPTS=-u<number> ..., however this isn't a good user experience or a persistent solution). Additionally, the behaviour around reading from thediff.contextanddiff.interHunkContextconfigs was also inconsistent with other diff generating commands such as "log -p".The summarised changes below hopefully make this experience better and fix some inconsistencies:
diff.contextanddiff.interHunkContextconfigs are now respected by--patchcompatible commands--unifiedand--inter-hunk-contextcommand line options have been added to--patchcompatible commands (which take prescendence over file configs)--interactivemode now expose a new "context" subcommand which configures the amount of context lines you wish to see in subsequent diffs generated from other subcommands such as "patch" or "diff"The original discussion for this can be read at:
Changes since v1:
test_grepandtest_configutilscc: "Kristoffer Haugsbakk" [email protected]
cc: Eric Sunshine [email protected]
cc: Christian Couder [email protected]
cc: Phillip Wood [email protected]