git/list[1] front-page[2] threads[3] people[4] search[5] about
 

Re: [GSoC][PATCH v2 5/6] rebase -i: support --ignore-date

From
PWPhillip Wood <phillip.wood123@gmail.com>
Date
Aug 13, 2019, 13:28 UTC
Message-ID
<b59f0aa2-1dc3-905a-0094-5f63718dffcf@gmail.com>
In-Reply-To
<20190812194301.5655-6-rohit.ashiwal265@gmail.com>
Hi Rohit

This is looking better - there are a couple of memory management issues and minor nit-picks but apart from that it looks good.

On 12/08/2019 20:42, Rohit Ashiwal wrote:
Show 24 quoted lines
> 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>
> ---
>   Documentation/git-rebase.txt            |  6 +--
>   builtin/rebase.c                        | 21 +++++++---
>   sequencer.c                             | 55 ++++++++++++++++++++++++-
>   sequencer.h                             |  1 +
>   t/t3433-rebase-options-compatibility.sh | 16 +++++++
>   5 files changed, 88 insertions(+), 11 deletions(-)
> 
> diff --git a/Documentation/git-rebase.txt b/Documentation/git-rebase.txt
> index 697ce8e6ff..24ad2dda0b 100644
> --- a/Documentation/git-rebase.txt
> +++ b/Documentation/git-rebase.txt
> @@ -388,8 +388,8 @@ See also INCOMPATIBLE OPTIONS below.
>   	as the committer date. This implies --force-rebase.
>   
>   --ignore-date::
> -	This flag is passed to 'git am' to change the author date
> -	of the rebased commits (see linkgit:git-am[1]).
> +	Instead of using the given author date, re-set
reset is not hyphenated
> it to the value
> +	same
> as committer (current) date. This implies --force-rebase.
s/value same as committer/same value as the committer/
Show 49 quoted lines
>   +
>   See also INCOMPATIBLE OPTIONS below.
>   
> @@ -526,7 +526,6 @@ INCOMPATIBLE OPTIONS
>   
>   The following options:
>   
> - * --ignore-date
>    * --whitespace
>    * -C
>   
> @@ -552,6 +551,7 @@ In addition, the following pairs of options are incompatible:
>    * --preserve-merges and --rebase-merges
>    * --preserve-merges and --ignore-whitespace
>    * --preserve-merges and --committer-date-is-author-date
> + * --preserve-merges and --ignore-date
>    * --rebase-merges and --ignore-whitespace
>    * --rebase-merges and --strategy
>    * --rebase-merges and --strategy-option
> diff --git a/builtin/rebase.c b/builtin/rebase.c
> index b1039f8db0..ed58ca8e5a 100644
> --- a/builtin/rebase.c
> +++ b/builtin/rebase.c
> @@ -83,6 +83,7 @@ struct rebase_options {
>   	char *gpg_sign_opt;
>   	int autostash;
>   	int committer_date_is_author_date;
> +	int ignore_date;
>   	char *cmd;
>   	int allow_empty_message;
>   	int rebase_merges, rebase_cousins;
> @@ -117,6 +118,7 @@ static struct replay_opts get_replay_opts(const struct rebase_options *opts)
>   	replay.reschedule_failed_exec = opts->reschedule_failed_exec;
>   	replay.committer_date_is_author_date =
>   					opts->committer_date_is_author_date;
> +	replay.ignore_date = opts->ignore_date;
>   	replay.gpg_sign = xstrdup_or_null(opts->gpg_sign_opt);
>   	replay.strategy = opts->strategy;
>   
> @@ -536,7 +538,10 @@ int cmd_rebase__interactive(int argc, const char **argv, const char *prefix)
>   		warning(_("--[no-]rebase-cousins has no effect without "
>   			  "--rebase-merges"));
>   
> -	if (opts.committer_date_is_author_date)
> +	if (opts.ignore_date)
> +		opts.committer_date_is_author_date = 0;
> +	if (opts.committer_date_is_author_date ||
> +	    opts.ignore_date)
>   		opts.flags |= REBASE_FORCE;
Is any of this used by rebase--preserve-merges.sh?
Show 28 quoted lines
>   	return !!run_rebase_interactive(&opts, command);
> @@ -989,6 +994,8 @@ static int run_am(struct rebase_options *opts)
>   		argv_array_push(&am.args, "--ignore-whitespace");
>   	if (opts->committer_date_is_author_date)
>   		argv_array_push(&opts->git_am_opts, "--committer-date-is-author-date");
> +	if (opts->ignore_date)
> +		argv_array_push(&opts->git_am_opts, "--ignore-date");
>   	if (opts->action && !strcmp("continue", opts->action)) {
>   		argv_array_push(&am.args, "--resolved");
>   		argv_array_pushf(&am.args, "--resolvemsg=%s", resolvemsg);
> @@ -1435,8 +1442,8 @@ int cmd_rebase(int argc, const char **argv, const char *prefix)
>   		OPT_BOOL(0, "committer-date-is-author-date",
>   			 &options.committer_date_is_author_date,
>   			 N_("make committer date match author date")),
> -		OPT_PASSTHRU_ARGV(0, "ignore-date", &options.git_am_opts, NULL,
> -				  N_("passed to 'git am'"), PARSE_OPT_NOARG),
> +		OPT_BOOL(0, "ignore-date", &options.ignore_date,
> +			 "ignore author date and use current date"),
>   		OPT_PASSTHRU_ARGV('C', NULL, &options.git_am_opts, N_("n"),
>   				  N_("passed to 'git apply'"), 0),
>   		OPT_BOOL(0, "ignore-whitespace", &options.ignore_whitespace,
> @@ -1705,13 +1712,15 @@ int cmd_rebase(int argc, const char **argv, const char *prefix)
>   		    state_dir_base, cmd_live_rebase, buf.buf);
>   	}
>   
> -	if (options.committer_date_is_author_date)
> +	if (options.ignore_date)
> +		options.committer_date_is_author_date = 0;

We should probably print an error if the user gives --committer-date-is-author-date and --author-date-is-committer-date (and in cmd_rebase__interactive() if we keep this option there)

Show 41 quoted lines
> +	if (options.committer_date_is_author_date ||
> +	    options.ignore_date)
>   		options.flags |= REBASE_FORCE;
>   
>   	for (i = 0; i < options.git_am_opts.argc; i++) {
>   		const char *option = options.git_am_opts.argv[i], *p;
> -		if (!strcmp(option, "--ignore-date") ||
> -		    !strcmp(option, "--whitespace=fix") ||
> +		if (!strcmp(option, "--whitespace=fix") ||
>   		    !strcmp(option, "--whitespace=strip"))
>   			options.flags |= REBASE_FORCE;
>   		else if (skip_prefix(option, "-C", &p)) {
> diff --git a/sequencer.c b/sequencer.c
> index e186136ccc..aecd9b4ad8 100644
> --- a/sequencer.c
> +++ b/sequencer.c
> @@ -148,6 +148,7 @@ static GIT_PATH_FUNC(rebase_path_refs_to_delete, "rebase-merge/refs-to-delete")
>    */
>   static GIT_PATH_FUNC(rebase_path_gpg_sign_opt, "rebase-merge/gpg_sign_opt")
>   static GIT_PATH_FUNC(rebase_path_cdate_is_adate, "rebase-merge/cdate_is_adate")
> +static GIT_PATH_FUNC(rebase_path_ignore_date, "rebase-merge/ignore_date")
>   static GIT_PATH_FUNC(rebase_path_orig_head, "rebase-merge/orig-head")
>   static GIT_PATH_FUNC(rebase_path_verbose, "rebase-merge/verbose")
>   static GIT_PATH_FUNC(rebase_path_quiet, "rebase-merge/quiet")
> @@ -919,6 +920,31 @@ static const char *read_author_ident(struct strbuf *buf)
>   	return buf->buf;
>   }
>   
> +static void ignore_author_date(const char **author)
> +{
> +	int len = strlen(*author);
> +	struct ident_split ident;
> +	struct strbuf new_author = STRBUF_INIT;
> +
> +	split_ident_line(&ident, *author, len);
> +	len = ident.mail_end - ident.name_begin + 1;
> +
> +	strbuf_addf(&new_author, "%.*s", len, *author);
> +	datestamp(&new_author);
> +	*author = strbuf_detach(&new_author, NULL);
> +}

It's good to see this using the indet api. I think this would be nicer if it took a char* returned the new author rather than changing the function argument, particularly as it does not free the string that is passed in so the ownership is unclear.

Show 8 quoted lines
> +
> +static void push_dates(struct child_process *child)
> +{
> +	time_t now = time(NULL);
> +	struct strbuf date = STRBUF_INIT;
> +
> +	strbuf_addf(&date, "@%"PRIuMAX, (uintmax_t)now);
> +	argv_array_pushf(&child->args, "--date=%s", date.buf);

it doesn't matter but it might have been nicer to set both dates the same way in the environment.

Show 19 quoted lines
> +	argv_array_pushf(&child->env_array, "GIT_COMMITTER_DATE=%s", date.buf);
> +	strbuf_release(&date);
> +}
> +
>   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"
> @@ -1020,10 +1046,18 @@ 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, which must set the author "
> +					    "before committing the tree");
> +				ignore_author_date(&author);
Is this leaking the old author? I'd rather see
	tmp_author = ignore_author_date(author);
	free(author);
	author = tmp_author;
Show 24 quoted lines
> +			}
>   			res = commit_tree(msg.buf, msg.len, cache_tree_oid,
>   					  NULL, &root_commit, author,
>   					  opts->gpg_sign);
> +		}
>   
>   		strbuf_release(&msg);
>   		strbuf_release(&script);
> @@ -1053,6 +1087,8 @@ 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)
> +		push_dates(&cmd);
>   	if (defmsg)
>   		argv_array_pushl(&cmd.args, "-F", defmsg, NULL);
>   	else if (!(flags & EDIT_MSG))
> @@ -1515,6 +1551,11 @@ static int try_to_commit(struct repository *r,
>   
>   	reset_ident_date();
>   
> +	if (opts->ignore_date) {
> +		ignore_author_date(&author);
> +		free(author_to_free);

Where is author_to_free set? We should always free the old author, see above.

> +		author_to_free = (char *)author;
Why do we need the cast - is author const?
Best Wishes
Phillip
Show 82 quoted lines
> +	}
>   	if (commit_tree_extended(msg->buf, msg->len, &tree, parents,
>   				 oid, author, opts->gpg_sign, extra)) {
>   		res = error(_("failed to write commit object"));
> @@ -2592,6 +2633,11 @@ 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;
>   
> @@ -2676,6 +2722,8 @@ 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", "");
>   
> @@ -3601,6 +3649,8 @@ 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)
> +			push_dates(&cmd);
>   
>   		/* Add the tips to be merged */
>   		for (j = to_merge; j; j = j->next)
> @@ -3874,7 +3924,8 @@ 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 e3881e9275..bf5a79afdb 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 b2419a2b75..c060fcd10b 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.
> +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
>
Previous: Rohit AshiwalNext: Junio C Hamano
Message 34 of 96 in “[GSoC][PATCHl 0/6] rebase -i: support more options”
  1. Rohit AshiwalAug 6, 2019
  2. [GSoC][PATCHl 1/6] rebase -i: add --ignore-whitespace flagRohit Ashiwal, Aug 6, 2019
  3. Junio C HamanoAug 7, 2019
  4. Rohit AshiwalAug 7, 2019
  5. Phillip WoodAug 8, 2019
  6. Rohit AshiwalAug 12, 2019
  7. [GSoC][PATCHl 3/6] rebase -i: support --committer-date-is-author-dateRohit Ashiwal, Aug 6, 2019
  8. Phillip WoodAug 8, 2019
  9. Junio C HamanoAug 8, 2019
  10. [GSoC][PATCHl 2/6] sequencer: add NULL checks under read_author_scriptRohit Ashiwal, Aug 6, 2019
  11. [GSoC][PATCHl 4/6] sequencer: rename amend_author to author_to_renameRohit Ashiwal, Aug 6, 2019
  12. Phillip WoodAug 8, 2019
  13. [GSoC][PATCHl 5/6] rebase -i: support --ignore-dateRohit Ashiwal, Aug 6, 2019
  14. Johannes SchindelinAug 7, 2019
  15. Junio C HamanoAug 7, 2019
  16. Rohit AshiwalAug 7, 2019
  17. Phillip WoodAug 8, 2019
  18. Phillip WoodAug 8, 2019
  19. [GSoC][PATCHl 6/6] rebase: add --author-date-is-committer-dateRohit Ashiwal, Aug 6, 2019
  20. Phillip WoodAug 8, 2019
  21. [GSoC][PATCH v2 0/6] rebase -i: support more optionsRohit Ashiwal, Aug 12, 2019
  22. [GSoC][PATCH v2 1/6] rebase -i: add --ignore-whitespace flagRohit Ashiwal, Aug 12, 2019
  23. Phillip WoodAug 13, 2019
  24. [GSoC][PATCH v2 2/6] sequencer: add NULL checks under read_author_scriptRohit Ashiwal, Aug 12, 2019
  25. [GSoC][PATCH v2 3/6] rebase -i: support --committer-date-is-author-dateRohit Ashiwal, Aug 12, 2019
  26. Phillip WoodAug 13, 2019
  27. Phillip WoodAug 13, 2019
  28. Junio C HamanoAug 13, 2019
  29. Phillip WoodAug 14, 2019
  30. Phillip WoodAug 13, 2019
  31. [GSoC][PATCH v2 4/6] sequencer: rename amend_author to author_to_renameRohit Ashiwal, Aug 12, 2019
  32. Phillip WoodAug 13, 2019
  33. [GSoC][PATCH v2 5/6] rebase -i: support --ignore-dateRohit Ashiwal, Aug 12, 2019
  34. Phillip WoodAug 13, 2019
  35. Junio C HamanoAug 13, 2019
  36. Phillip WoodAug 14, 2019
  37. Junio C HamanoAug 13, 2019
  38. Phillip WoodAug 14, 2019
  39. Junio C HamanoAug 14, 2019
  40. Phillip WoodAug 17, 2019
  41. [GSoC][PATCH v2 6/6] rebase: add --author-date-is-committer-dateRohit Ashiwal, Aug 12, 2019
  42. Junio C HamanoAug 13, 2019
  43. 0/6 rebase -i: support more optionsRohit Ashiwal, Aug 20, 2019
  44. 1/6 rebase -i: add --ignore-whitespace flagRohit Ashiwal, Aug 20, 2019
  45. Phillip WoodAug 20, 2019
  46. Rohit AshiwalAug 20, 2019
  47. 2/6 sequencer: add NULL checks under read_author_scriptRohit Ashiwal, Aug 20, 2019
  48. Junio C HamanoAug 23, 2019
  49. 3/6 rebase -i: support --committer-date-is-author-dateRohit Ashiwal, Aug 20, 2019
  50. Phillip WoodAug 20, 2019
  51. 4/6 sequencer: rename amend_author to author_to_renameRohit Ashiwal, Aug 20, 2019
  52. 5/6 rebase -i: support --ignore-dateRohit Ashiwal, Aug 20, 2019
  53. Phillip WoodAug 20, 2019
  54. Junio C HamanoAug 20, 2019
  55. Phillip WoodAug 20, 2019
  56. [GSoC][PATCH v2 6/6] rebase: add --author-date-is-committer-dateRohit Ashiwal, Aug 20, 2019
  57. Rohit AshiwalAug 20, 2019
  58. 6/6 rebase: add --reset-author-dateRohit Ashiwal, Aug 20, 2019
  59. Rohit AshiwalAug 20, 2019
  60. Phillip WoodAug 20, 2019
  61. Junio C HamanoAug 20, 2019
  62. Phillip WoodAug 20, 2019
  63. 0/6 rebase -i: support more optionsRohit Ashiwal, Sep 7, 2019
  64. 1/6 rebase -i: add --ignore-whitespace flagRohit Ashiwal, Sep 7, 2019
  65. Phillip WoodOct 4, 2019
  66. Elijah NewrenOct 5, 2019
  67. Rohit AshiwalOct 6, 2019
  68. 2/6 sequencer: allow callers of read_author_script() to ignore fieldsRohit Ashiwal, Sep 7, 2019
  69. 3/6 rebase -i: support --committer-date-is-author-dateRohit Ashiwal, Sep 7, 2019
  70. Phillip WoodOct 4, 2019
  71. Rohit AshiwalOct 6, 2019
  72. Phillip WoodOct 24, 2019
  73. 4/6 sequencer: rename amend_author to author_to_renameRohit Ashiwal, Sep 7, 2019
  74. 5/6 rebase -i: support --ignore-dateRohit Ashiwal, Sep 7, 2019
  75. Rohit AshiwalSep 7, 2019
  76. Phillip WoodSep 27, 2019
  77. Rohit AshiwalOct 6, 2019
  78. Phillip WoodOct 24, 2019
  79. 6/6 rebase: add --reset-author-dateRohit Ashiwal, Sep 7, 2019
  80. Junio C HamanoSep 9, 2019
  81. Phillip WoodSep 9, 2019
  82. Junio C HamanoSep 9, 2019
  83. 0/6 rebase -i: support more optionsRohit Ashiwal, Nov 1, 2019
  84. 1/6 rebase -i: add --ignore-whitespace flagRohit Ashiwal, Nov 1, 2019
  85. 2/6 sequencer: allow callers of read_author_script() to ignore fieldsRohit Ashiwal, Nov 1, 2019
  86. 3/6 rebase -i: support --committer-date-is-author-dateRohit Ashiwal, Nov 1, 2019
  87. 4/6 sequencer: rename amend_author to author_to_renameRohit Ashiwal, Nov 1, 2019
  88. 5/6 rebase -i: support --ignore-dateRohit Ashiwal, Nov 1, 2019
  89. Junio C HamanoNov 2, 2019
  90. Junio C HamanoNov 2, 2019
  91. 6/6 rebase: add --reset-author-dateRohit Ashiwal, Nov 1, 2019
  92. Junio C HamanoNov 2, 2019
  93. Junio C HamanoNov 21, 2019
  94. Alban GruinNov 21, 2019
  95. Junio C HamanoNov 22, 2019
  96. Phillip WoodNov 28, 2019

Read the whole thread, see it on lore, or plain text.

$ cat FOOTERMessages come from the public archive at lore.kernel.org/git, fetched every hour. The front page is chosen and written each morning by an AI editor and can be wrong; the threads themselves are the record. About and API. For agents: an MCP server at https://gitlist.dev/mcp, and any thread, story or person page as Markdown by adding .md to its URL (or sending Accept: text/markdown). Details in /llms.txt.