All of lore.kernel.org
 help / color / mirror / Atom feed
From: Phillip Wood <phillip.wood123@gmail.com>
To: Rohit Ashiwal <rohit.ashiwal265@gmail.com>,
	Junio <gitster@pobox.com>, GIT Mailing List <git@vger.kernel.org>
Cc: Thomas <t.gummerer@gmail.com>, Elijah <newren@gmail.com>,
	Dscho <Johannes.Schindelin@gmx.de>,
	Martin <martin.agren@gmail.com>
Subject: Re: [GSoC][PATCHl 5/6] rebase -i: support --ignore-date
Date: Thu, 8 Aug 2019 17:53:09 +0100	[thread overview]
Message-ID: <24577efb-a959-04b1-c202-284571c5505e@gmail.com> (raw)
In-Reply-To: <0619f7b7-a7e5-54a2-3d28-ac359139bde7@gmail.com>

On 08/08/2019 12:42, Phillip Wood wrote:
> On 06/08/2019 18:36, Rohit Ashiwal wrote:
>> rebase am already has this flag to "lie" about the author date
>> by changing it to the committer (current) date. Let's add the same
>> for interactive machinery.
>>
>> Signed-off-by: Rohit Ashiwal <rohit.ashiwal265@gmail.com>
>> ---
 >> [...[
>>   static const char staged_changes_advice[] =
>>   N_("you have staged changes in your working tree\n"
>>   "If these changes are meant to be squashed into the previous commit, 
>> run:\n"
>> @@ -985,7 +997,7 @@ static int run_git_commit(struct repository *r,
>>   {
>>       struct child_process cmd = CHILD_PROCESS_INIT;
>> -    if (opts->committer_date_is_author_date &&
>> +    if (opts->committer_date_is_author_date && !opts->ignore_date &&
>>           setenv_committer_date_to_author_date())
>>           return 1;
> 
> We read the author script again just below, can set the committer date 
> there by parsing the author string, that would mean you could use the 
> same function that works on an author string in try_to_commit() (this 
> comment should be on patch 3 I think)

It's a bit more complicated as you've done this to avoid the duplication 
in the previous version. I think it should be possible to do it by 
reading the author identity upfront (protected by 'if 
is_rebase_i(opts)') and processing it appropriately in the two code 
paths. (You might need to refactor some of the functions like 
read_env_script() to do this.) An alternative would be to refactor 
try_to_commit() so that it can create the root commit and delete the 
code here that does that.

Best Wishes

Phillip

> 
>>       if ((flags & CREATE_ROOT_COMMIT) && !(flags & AMEND_MSG)) {
>> @@ -1013,10 +1025,18 @@ static int run_git_commit(struct repository *r,
>>           if (res <= 0)
>>               res = error_errno(_("could not read '%s'"), defmsg);
>> -        else
>> +        else {
>> +            if (opts->ignore_date) {
>> +                if (!author)
>> +                    BUG("ignore-date can only be used with "
>> +                        "rebase -i, which must set the "
> 
> I know it's only a bug message but it's not just 'rebase -i' but 'rebase 
> -k', 'rebase -m' ... that use this code path, it would be better just to 
> say 'rebase'
> 
>> +                        "author before committing the tree");
>> +                ignore_author_date(&author);
>> +            }
>>               res = commit_tree(msg.buf, msg.len, cache_tree_oid,
>>                         NULL, &root_commit, author,
>>                         opts->gpg_sign);
>> +        }
>>           strbuf_release(&msg);
>>           strbuf_release(&script);
>> @@ -1046,6 +1066,8 @@ static int run_git_commit(struct repository *r,
>>           argv_array_push(&cmd.args, "--amend");
>>       if (opts->gpg_sign)
>>           argv_array_pushf(&cmd.args, "-S%s", opts->gpg_sign);
>> +    if (opts->ignore_date)
>> +        argv_array_pushf(&cmd.args, "--date=%ld", time(NULL));
> 
> I think this is racy as it only sets the author date, the committer date 
> may end up being different.
> 
>>       if (defmsg)
>>           argv_array_pushl(&cmd.args, "-F", defmsg, NULL);
>>       else if (!(flags & EDIT_MSG))
>> @@ -1425,7 +1447,7 @@ static int try_to_commit(struct repository *r,
>>       if (parse_head(r, &current_head))
>>           return -1;
>>       if (!(flags & AMEND_MSG) && opts->committer_date_is_author_date &&
>> -        setenv_committer_date_to_author_date())
>> +        !opts->ignore_date && setenv_committer_date_to_author_date())
>>           return -1;
>>       if (flags & AMEND_MSG) {
>>           const char *exclude_gpgsig[] = { "gpgsig", NULL };
>> @@ -1447,7 +1469,7 @@ static int try_to_commit(struct repository *r,
>>               res = error(_("unable to parse commit author"));
>>               goto out;
>>           }
>> -        if (opts->committer_date_is_author_date) {
>> +        if (opts->committer_date_is_author_date && !opts->ignore_date) {
> 
> If we only handled committer_date_is_author_date in a single place it 
> wouldn't need to be changed twice in this patch
> 
>>               char *date;
>>               int len = strlen(author);
>>               char *idx = memchr(author, '>', len);
>> @@ -1507,6 +1529,11 @@ static int try_to_commit(struct repository *r,
>>       reset_ident_date();
>> +    if (opts->ignore_date) {
>> +        ignore_author_date(&author);
>> +        free(author_to_free);
>> +        author_to_free = (char *)author;
>> +    }
>>       if (commit_tree_extended(msg->buf, msg->len, &tree, parents,
>>                    oid, author, opts->gpg_sign, extra)) {
>>           res = error(_("failed to write commit object"));
>> @@ -2583,6 +2610,11 @@ static int read_populate_opts(struct 
>> replay_opts *opts)
>>               opts->committer_date_is_author_date = 1;
>>           }
>> +        if (file_exists(rebase_path_ignore_date())) {
>> +            opts->allow_ff = 0;
>> +            opts->ignore_date = 1;
>> +        }
>> +
>>           if (file_exists(rebase_path_reschedule_failed_exec()))
>>               opts->reschedule_failed_exec = 1;
>> @@ -2667,6 +2699,8 @@ int write_basic_state(struct replay_opts *opts, 
>> const char *head_name,
>>           write_file(rebase_path_signoff(), "--signoff\n");
>>       if (opts->committer_date_is_author_date)
>>           write_file(rebase_path_cdate_is_adate(), "%s", "");
>> +    if (opts->ignore_date)
>> +        write_file(rebase_path_ignore_date(), "%s", "");
>>       if (opts->reschedule_failed_exec)
>>           write_file(rebase_path_reschedule_failed_exec(), "%s", "");
>> @@ -3484,6 +3518,9 @@ static int do_merge(struct repository *r,
>>           argv_array_push(&cmd.args, git_path_merge_msg(r));
>>           if (opts->gpg_sign)
>>               argv_array_push(&cmd.args, opts->gpg_sign);
>> +        if (opts->ignore_date)
>> +            argv_array_pushf(&cmd.args,
>> +                     "GIT_AUTHOR_DATE=%ld", time(NULL));
>>           /* Add the tips to be merged */
>>           for (j = to_merge; j; j = j->next)
>> @@ -3756,7 +3793,8 @@ static int pick_commits(struct repository *r,
>>       if (opts->allow_ff)
>>           assert(!(opts->signoff || opts->no_commit ||
>>                   opts->record_origin || opts->edit ||
>> -                opts->committer_date_is_author_date));
>> +                opts->committer_date_is_author_date ||
>> +                opts->ignore_date));
>>       if (read_and_refresh_cache(r, opts))
>>           return -1;
>> diff --git a/sequencer.h b/sequencer.h
>> index e6cba468db..73d0515a3e 100644
>> --- a/sequencer.h
>> +++ b/sequencer.h
>> @@ -44,6 +44,7 @@ struct replay_opts {
>>       int quiet;
>>       int reschedule_failed_exec;
>>       int committer_date_is_author_date;
>> +    int ignore_date;
>>       int mainline;
>> diff --git a/t/t3433-rebase-options-compatibility.sh 
>> b/t/t3433-rebase-options-compatibility.sh
>> index ceab48a831..95d99c4b7b 100755
>> --- a/t/t3433-rebase-options-compatibility.sh
>> +++ b/t/t3433-rebase-options-compatibility.sh
>> @@ -81,4 +81,20 @@ test_expect_success 
>> '--committer-date-is-author-date works with interactive back
>>       test_cmp authortime committertime
>>   '
>> +# Checking for +0000 in author time is enough since default
>> +# timezone is UTC, but the timezone used while committing
>> +# sets to +0530.
> 
> That sounds potentially fragile but I guess the timezone is unlikely to 
> change in the future
> 
> Best Wishes
> 
> Phillip
> 
>> +test_expect_success '--ignore-date works with am backend' '
>> +    git commit --amend --date="$GIT_AUTHOR_DATE" &&
>> +    git rebase --ignore-date HEAD^ &&
>> +    git show HEAD --pretty="format:%ai" >authortime &&
>> +    grep "+0000" authortime
>> +'
>> +
>> +test_expect_success '--ignore-date works with interactive backend' '
>> +    git commit --amend --date="$GIT_AUTHOR_DATE" &&
>> +    git rebase --ignore-date -i HEAD^ &&
>> +    git show HEAD --pretty="format:%ai" >authortime &&
>> +    grep "+0000" authortime
>> +'
>>   test_done
>>

  reply	other threads:[~2019-08-08 16:53 UTC|newest]

Thread overview: 96+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-08-06 17:36 [GSoC][PATCHl 0/6] rebase -i: support more options Rohit Ashiwal
2019-08-06 17:36 ` [GSoC][PATCHl 1/6] rebase -i: add --ignore-whitespace flag Rohit Ashiwal
2019-08-07  5:38   ` Junio C Hamano
2019-08-07 20:25     ` Rohit Ashiwal
2019-08-08 16:44   ` Phillip Wood
2019-08-12 17:43     ` Rohit Ashiwal
2019-08-06 17:36 ` [GSoC][PATCHl 2/6] sequencer: add NULL checks under read_author_script Rohit Ashiwal
2019-08-06 17:36 ` [GSoC][PATCHl 3/6] rebase -i: support --committer-date-is-author-date Rohit Ashiwal
2019-08-08 11:29   ` Phillip Wood
2019-08-08 16:00     ` Junio C Hamano
2019-08-06 17:36 ` [GSoC][PATCHl 4/6] sequencer: rename amend_author to author_to_rename Rohit Ashiwal
2019-08-08 11:30   ` Phillip Wood
2019-08-06 17:36 ` [GSoC][PATCHl 5/6] rebase -i: support --ignore-date Rohit Ashiwal
2019-08-07 19:41   ` Johannes Schindelin
2019-08-07 20:22     ` Junio C Hamano
2019-08-07 20:33       ` Rohit Ashiwal
2019-08-08 11:42   ` Phillip Wood
2019-08-08 16:53     ` Phillip Wood [this message]
2019-08-06 17:36 ` [GSoC][PATCHl 6/6] rebase: add --author-date-is-committer-date Rohit Ashiwal
2019-08-08 11:42   ` Phillip Wood
2019-08-12 19:42 ` [GSoC][PATCH v2 0/6] rebase -i: support more options Rohit Ashiwal
2019-08-12 19:42   ` [GSoC][PATCH v2 1/6] rebase -i: add --ignore-whitespace flag Rohit Ashiwal
2019-08-13 12:07     ` Phillip Wood
2019-08-12 19:42   ` [GSoC][PATCH v2 2/6] sequencer: add NULL checks under read_author_script Rohit Ashiwal
2019-08-12 19:42   ` [GSoC][PATCH v2 3/6] rebase -i: support --committer-date-is-author-date Rohit Ashiwal
2019-08-13 10:38     ` Phillip Wood
2019-08-13 12:09       ` Phillip Wood
2019-08-13 17:06       ` Junio C Hamano
2019-08-14 18:38         ` Phillip Wood
2019-08-13 13:35     ` Phillip Wood
2019-08-12 19:42   ` [GSoC][PATCH v2 4/6] sequencer: rename amend_author to author_to_rename Rohit Ashiwal
2019-08-13 13:29     ` Phillip Wood
2019-08-12 19:42   ` [GSoC][PATCH v2 5/6] rebase -i: support --ignore-date Rohit Ashiwal
2019-08-13 13:28     ` Phillip Wood
2019-08-13 17:21       ` Junio C Hamano
2019-08-14 18:47         ` Phillip Wood
2019-08-13 21:45     ` Junio C Hamano
2019-08-14 18:51       ` Phillip Wood
2019-08-14 19:33         ` Junio C Hamano
2019-08-17  9:28           ` Phillip Wood
2019-08-12 19:43   ` [GSoC][PATCH v2 6/6] rebase: add --author-date-is-committer-date Rohit Ashiwal
2019-08-13 17:28     ` Junio C Hamano
2019-08-20  3:45 ` [PATCH v3 0/6] rebase -i: support more options Rohit Ashiwal
2019-08-20  3:45   ` [PATCH v3 1/6] rebase -i: add --ignore-whitespace flag Rohit Ashiwal
2019-08-20 18:40     ` Phillip Wood
2019-08-20 18:47       ` Rohit Ashiwal
2019-08-20  3:45   ` [PATCH v3 2/6] sequencer: add NULL checks under read_author_script Rohit Ashiwal
2019-08-23 15:20     ` Junio C Hamano
2019-08-20  3:45   ` [PATCH v3 3/6] rebase -i: support --committer-date-is-author-date Rohit Ashiwal
2019-08-20 13:32     ` Phillip Wood
2019-08-20  3:45   ` [PATCH v3 4/6] sequencer: rename amend_author to author_to_rename Rohit Ashiwal
2019-08-20  3:45   ` [PATCH v3 5/6] rebase -i: support --ignore-date Rohit Ashiwal
2019-08-20 13:45     ` Phillip Wood
2019-08-20 17:42     ` Junio C Hamano
2019-08-20 18:30       ` Phillip Wood
2019-08-20  3:45   ` [GSoC][PATCH v2 6/6] rebase: add --author-date-is-committer-date Rohit Ashiwal
2019-08-20  4:00     ` Rohit Ashiwal
2019-08-20  3:45   ` [PATCH v3 6/6] rebase: add --reset-author-date Rohit Ashiwal
2019-08-20  3:54   ` [PATCH v3 0/6] rebase -i: support more options Rohit Ashiwal
2019-08-20 13:56   ` Phillip Wood
2019-08-20 17:53     ` Junio C Hamano
2019-08-20 18:37       ` Phillip Wood
2019-09-07 11:50 ` [PATCH v4 " Rohit Ashiwal
2019-09-07 11:50   ` [PATCH v4 1/6] rebase -i: add --ignore-whitespace flag Rohit Ashiwal
2019-10-04  9:29     ` Phillip Wood
2019-10-05 18:12       ` Elijah Newren
2019-10-06 17:57       ` Rohit Ashiwal
2019-09-07 11:50   ` [PATCH v4 2/6] sequencer: allow callers of read_author_script() to ignore fields Rohit Ashiwal
2019-09-07 11:50   ` [PATCH v4 3/6] rebase -i: support --committer-date-is-author-date Rohit Ashiwal
2019-10-04  9:37     ` Phillip Wood
2019-10-06 17:57       ` Rohit Ashiwal
2019-10-24 13:28     ` Phillip Wood
2019-09-07 11:50   ` [PATCH v4 4/6] sequencer: rename amend_author to author_to_rename Rohit Ashiwal
2019-09-07 11:50   ` [PATCH v4 5/6] rebase -i: support --ignore-date Rohit Ashiwal
2019-09-07 20:56     ` Rohit Ashiwal
2019-09-27 10:00     ` Phillip Wood
2019-10-06 17:57       ` Rohit Ashiwal
2019-10-24 20:36         ` Phillip Wood
2019-09-07 11:50   ` [PATCH v4 6/6] rebase: add --reset-author-date Rohit Ashiwal
2019-09-09 18:02   ` [PATCH v4 0/6] rebase -i: support more options Junio C Hamano
2019-09-09 18:51     ` Phillip Wood
2019-09-09 19:24       ` Junio C Hamano
2019-11-01 13:59 ` [PATCH v5 " Rohit Ashiwal
2019-11-01 13:59   ` [PATCH v5 1/6] rebase -i: add --ignore-whitespace flag Rohit Ashiwal
2019-11-01 13:59   ` [PATCH v5 2/6] sequencer: allow callers of read_author_script() to ignore fields Rohit Ashiwal
2019-11-01 14:00   ` [PATCH v5 3/6] rebase -i: support --committer-date-is-author-date Rohit Ashiwal
2019-11-01 14:00   ` [PATCH v5 4/6] sequencer: rename amend_author to author_to_rename Rohit Ashiwal
2019-11-01 14:00   ` [PATCH v5 5/6] rebase -i: support --ignore-date Rohit Ashiwal
2019-11-02  7:32     ` Junio C Hamano
2019-11-02  7:48       ` Junio C Hamano
2019-11-01 14:00   ` [PATCH v5 6/6] rebase: add --reset-author-date Rohit Ashiwal
2019-11-02  7:34     ` Junio C Hamano
2019-11-21  6:14   ` [PATCH v5 0/6] rebase -i: support more options Junio C Hamano
2019-11-21  8:17     ` Alban Gruin
2019-11-22  6:32       ` Junio C Hamano
2019-11-28 11:14   ` Phillip Wood

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=24577efb-a959-04b1-c202-284571c5505e@gmail.com \
    --to=phillip.wood123@gmail.com \
    --cc=Johannes.Schindelin@gmx.de \
    --cc=git@vger.kernel.org \
    --cc=gitster@pobox.com \
    --cc=martin.agren@gmail.com \
    --cc=newren@gmail.com \
    --cc=rohit.ashiwal265@gmail.com \
    --cc=t.gummerer@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.