From: Johannes Schindelin <Johannes.Schindelin@gmx.de>
To: Junio C Hamano <gitster@pobox.com>
Cc: Paul-Sebastian Ungureanu <ungureanupaulsebastian@gmail.com>,
git@vger.kernel.org, t.gummerer@gmail.com
Subject: Re: [PATCH v12 04/26] ident: add the ability to provide a "fallback identity"
Date: Thu, 27 Dec 2018 22:24:51 +0100 (STD) [thread overview]
Message-ID: <nycvar.QRO.7.76.6.1812272221380.45@tvgsbejvaqbjf.bet> (raw)
In-Reply-To: <xmqqpntoq8zs.fsf@gitster-ct.c.googlers.com>
Hi Junio,
On Wed, 26 Dec 2018, Junio C Hamano wrote:
> Paul-Sebastian Ungureanu <ungureanupaulsebastian@gmail.com> writes:
>
> > +static void set_env_if(const char *key, const char *value, int *given, int bit)
> > +{
> > + if ((*given & bit) || getenv(key))
> > + return; /* nothing to do */
> > + setenv(key, value, 0);
> > + *given |= bit;
> > +}
>
> We call setenv(3) with overwrite=0 but we protect the call with a
> check for existing value with getenv(3), which feels a bit like an
> anti-pattern. Wouldn't the following be simpler to follow, I wonder?
>
> if (!(*given & bit)) {
> setenv(key, value, 1);
> *given |= bit;
> }
>
> The only case these two may behave differently is when '*given' does
> not have the 'bit' set but the environment 'key' already exists.
Indeed, this is the case where your version would actually do the wrong
thing. Imagine that GIT_AUTHOR_NAME is set already. Your code would
*override* it. But that is not what we want to do here. We want to *fall
back* if there is no already-configured value.
And of course we won't set the `given` bit if we don't fall back here;
that should be done somewhere else, where that environment variable (that
we *refuse* to overwrite) is *actually* used.
Ciao,
Dscho
> The proposed patch will leave 'bit' in '*given' unset, so when a
> later code says "let's see if author_ident is explicitly given, and
> complain otherwise", such a check will trigger and cause complaint.
>
> On the other hand, the simplified version does not allow the
> "explicitly-given" bits to be left unset, so it won't cause
> complaint.
>
> Isn't it a BUG() if *given lacks 'bit' when the corresponding
> environment variable 'key' is missing? IOW, I would understand
> an implementation that is more elaborate than the simplified one I
> just gave above were something like
>
> if (!(*given & bit)) {
> if (getenv(key))
> BUG("why does %s exist and no %x bit set???", key, bit);
> setenv(key, value, 0);
> *given |= bit;
> }
>
> but I do not quite understand the reasoning behind the "check either
> the bit, or the environment variable" in the proposed patch.
>
> > +void prepare_fallback_ident(const char *name, const char *email)
> > +{
> > + set_env_if("GIT_AUTHOR_NAME", name,
> > + &author_ident_explicitly_given, IDENT_NAME_GIVEN);
> > + set_env_if("GIT_AUTHOR_EMAIL", email,
> > + &author_ident_explicitly_given, IDENT_MAIL_GIVEN);
> > + set_env_if("GIT_COMMITTER_NAME", name,
> > + &committer_ident_explicitly_given, IDENT_NAME_GIVEN);
> > + set_env_if("GIT_COMMITTER_EMAIL", email,
> > + &committer_ident_explicitly_given, IDENT_MAIL_GIVEN);
> > +}
>
> Introducing this function alone without a caller and without
> function doc is a bit unfriendly to future callers, who must be
> careful when to call it, I think. For example, they must know that
> it will be a disaster if they call this before they call
> git_ident_config(), right?
>
> > +
> > static int buf_cmp(const char *a_begin, const char *a_end,
> > const char *b_begin, const char *b_end)
> > {
>
>
>
next prev parent reply other threads:[~2018-12-27 21:25 UTC|newest]
Thread overview: 106+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <https://public-inbox.org/git/cover.1542925164.git.ungureanupaulsebastian@gmail.com/>
2018-12-20 19:44 ` [PATCH v12 00/26] Convert "git stash" to C builtin Paul-Sebastian Ungureanu
2018-12-20 19:44 ` [PATCH v12 01/26] sha1-name.c: add `get_oidf()` which acts like `get_oid()` Paul-Sebastian Ungureanu
2018-12-20 19:44 ` [PATCH v12 02/26] strbuf.c: add `strbuf_join_argv()` Paul-Sebastian Ungureanu
2018-12-20 19:44 ` [PATCH v12 03/26] strbuf.c: add `strbuf_insertf()` and `strbuf_vinsertf()` Paul-Sebastian Ungureanu
2018-12-20 19:44 ` [PATCH v12 04/26] ident: add the ability to provide a "fallback identity" Paul-Sebastian Ungureanu
2018-12-26 21:21 ` Junio C Hamano
2018-12-27 21:24 ` Johannes Schindelin [this message]
2018-12-28 19:40 ` Junio C Hamano
2018-12-20 19:44 ` [PATCH v12 05/26] stash: improve option parsing test coverage Paul-Sebastian Ungureanu
2018-12-20 19:44 ` [PATCH v12 06/26] t3903: modernize style Paul-Sebastian Ungureanu
2018-12-20 19:44 ` [PATCH v12 07/26] stash: rename test cases to be more descriptive Paul-Sebastian Ungureanu
2018-12-20 19:44 ` [PATCH v12 08/26] stash: add tests for `git stash show` config Paul-Sebastian Ungureanu
2018-12-20 19:44 ` [PATCH v12 09/26] stash: mention options in `show` synopsis Paul-Sebastian Ungureanu
2018-12-20 19:44 ` [PATCH v12 10/26] stash: convert apply to builtin Paul-Sebastian Ungureanu
2018-12-20 19:44 ` [PATCH v12 11/26] stash: convert drop and clear " Paul-Sebastian Ungureanu
2018-12-20 19:44 ` [PATCH v12 12/26] stash: convert branch " Paul-Sebastian Ungureanu
2018-12-20 19:44 ` [PATCH v12 13/26] stash: convert pop " Paul-Sebastian Ungureanu
2018-12-20 19:44 ` [PATCH v12 14/26] stash: convert list " Paul-Sebastian Ungureanu
2018-12-20 19:44 ` [PATCH v12 15/26] stash: convert show " Paul-Sebastian Ungureanu
2018-12-20 19:44 ` [PATCH v12 16/26] stash: convert store " Paul-Sebastian Ungureanu
2018-12-20 19:44 ` [PATCH v12 17/26] stash: convert create " Paul-Sebastian Ungureanu
2018-12-20 19:44 ` [PATCH v12 18/26] stash: convert push " Paul-Sebastian Ungureanu
2019-02-08 11:30 ` SZEDER Gábor
2019-02-10 22:17 ` Thomas Gummerer
2019-02-11 1:13 ` SZEDER Gábor
2019-02-12 23:18 ` Thomas Gummerer
2019-02-19 0:23 ` SZEDER Gábor
2019-02-19 10:47 ` Johannes Schindelin
2019-02-19 19:59 ` Junio C Hamano
2019-02-20 21:01 ` Johannes Schindelin
2019-02-19 23:59 ` Thomas Gummerer
2019-02-20 4:37 ` Junio C Hamano
2019-02-20 21:10 ` Johannes Schindelin
2019-02-20 22:30 ` Thomas Gummerer
2019-02-25 23:16 ` [PATCH v13 00/27] Convert "git stash" to C builtin Thomas Gummerer
2019-02-25 23:16 ` [PATCH v13 01/27] sha1-name.c: add `get_oidf()` which acts like `get_oid()` Thomas Gummerer
2019-02-25 23:16 ` [PATCH v13 02/27] strbuf.c: add `strbuf_join_argv()` Thomas Gummerer
2019-02-25 23:16 ` [PATCH v13 03/27] strbuf.c: add `strbuf_insertf()` and `strbuf_vinsertf()` Thomas Gummerer
2019-02-25 23:16 ` [PATCH v13 04/27] ident: add the ability to provide a "fallback identity" Thomas Gummerer
2019-02-25 23:16 ` [PATCH v13 05/27] stash: improve option parsing test coverage Thomas Gummerer
2019-02-25 23:16 ` [PATCH v13 06/27] t3903: modernize style Thomas Gummerer
2019-02-25 23:16 ` [PATCH v13 07/27] t3903: add test for --intent-to-add file Thomas Gummerer
2019-02-25 23:16 ` [PATCH v13 08/27] stash: rename test cases to be more descriptive Thomas Gummerer
2019-02-25 23:16 ` [PATCH v13 09/27] stash: add tests for `git stash show` config Thomas Gummerer
2019-02-25 23:16 ` [PATCH v13 10/27] stash: mention options in `show` synopsis Thomas Gummerer
2019-02-25 23:16 ` [PATCH v13 11/27] stash: convert apply to builtin Thomas Gummerer
2019-03-14 13:19 ` regression in new built-in stash + fsmonitor (was Re: [PATCH v13 11/27] stash: convert apply to builtin) Ævar Arnfjörð Bjarmason
2019-03-14 15:20 ` Johannes Schindelin
2019-03-14 15:40 ` Ævar Arnfjörð Bjarmason
2019-03-14 22:45 ` Johannes Schindelin
2019-03-14 23:39 ` Ævar Arnfjörð Bjarmason
2019-03-15 2:23 ` Ben Peart
2019-02-25 23:16 ` [PATCH v13 12/27] stash: convert drop and clear to builtin Thomas Gummerer
2019-03-07 19:15 ` Jeff King
2019-03-09 18:30 ` Thomas Gummerer
2019-03-10 23:26 ` Jeff King
2019-03-11 1:40 ` Junio C Hamano
2019-03-11 21:40 ` Thomas Gummerer
2019-02-25 23:16 ` [PATCH v13 13/27] stash: convert branch " Thomas Gummerer
2019-02-25 23:16 ` [PATCH v13 14/27] stash: convert pop " Thomas Gummerer
2019-02-25 23:16 ` [PATCH v13 15/27] stash: convert list " Thomas Gummerer
2019-02-25 23:16 ` [PATCH v13 16/27] stash: convert show " Thomas Gummerer
2019-02-25 23:16 ` [PATCH v13 17/27] stash: convert store " Thomas Gummerer
2019-02-25 23:16 ` [PATCH v13 18/27] stash: convert create " Thomas Gummerer
2019-03-07 19:18 ` Jeff King
2019-03-08 15:30 ` Johannes Schindelin
2019-03-09 18:26 ` Thomas Gummerer
2019-03-11 1:47 ` Junio C Hamano
2019-03-11 7:30 ` Junio C Hamano
2019-03-11 21:42 ` Thomas Gummerer
2019-03-11 22:16 ` [PATCH v2] stash: pass pathspec as pointer Thomas Gummerer
2019-03-12 6:50 ` Junio C Hamano
2019-03-12 22:35 ` Johannes Schindelin
2019-03-12 23:40 ` Thomas Gummerer
2019-03-13 1:47 ` Junio C Hamano
2019-03-13 22:14 ` Johannes Schindelin
2019-03-15 22:33 ` Thomas Gummerer
2019-02-25 23:16 ` [PATCH v13 19/27] stash: convert push to builtin Thomas Gummerer
2019-02-25 23:16 ` [PATCH v13 20/27] stash: make push -q quiet Thomas Gummerer
2019-02-25 23:16 ` [PATCH v13 21/27] stash: convert save to builtin Thomas Gummerer
2019-02-25 23:16 ` [PATCH v13 22/27] stash: optimize `get_untracked_files()` and `check_changes()` Thomas Gummerer
2019-02-25 23:16 ` [PATCH v13 23/27] stash: replace all `write-tree` child processes with API calls Thomas Gummerer
2019-02-25 23:16 ` [PATCH v13 24/27] stash: convert `stash--helper.c` into `stash.c` Thomas Gummerer
2019-02-25 23:16 ` [PATCH v13 25/27] stash: add back the original, scripted `git stash` Thomas Gummerer
2019-02-25 23:16 ` [PATCH v13 26/27] stash: optionally use the scripted version again Thomas Gummerer
2019-02-25 23:16 ` [PATCH v13 27/27] tests: add a special setup where stash.useBuiltin is off Thomas Gummerer
2019-02-26 12:40 ` [PATCH v13 00/27] Convert "git stash" to C builtin Johannes Schindelin
2019-02-26 20:48 ` Thomas Gummerer
2019-02-26 21:45 ` Ævar Arnfjörð Bjarmason
2019-02-26 22:37 ` Johannes Schindelin
2019-03-03 1:24 ` Junio C Hamano
2019-03-03 1:25 ` Junio C Hamano
2019-03-03 10:03 ` Thomas Gummerer
2018-12-20 19:44 ` [PATCH v12 19/26] stash: make push -q quiet Paul-Sebastian Ungureanu
2018-12-20 19:44 ` [PATCH v12 20/26] stash: convert save to builtin Paul-Sebastian Ungureanu
2018-12-20 19:44 ` [PATCH v12 21/26] stash: optimize `get_untracked_files()` and `check_changes()` Paul-Sebastian Ungureanu
2019-01-06 22:47 ` Thomas Gummerer
2018-12-20 19:44 ` [PATCH v12 22/26] stash: replace all `write-tree` child processes with API calls Paul-Sebastian Ungureanu
2018-12-20 19:44 ` [PATCH v12 23/26] stash: convert `stash--helper.c` into `stash.c` Paul-Sebastian Ungureanu
2018-12-20 19:44 ` [PATCH v12 24/26] stash: add back the original, scripted `git stash` Paul-Sebastian Ungureanu
2018-12-20 19:44 ` [PATCH v12 25/26] stash: optionally use the scripted version again Paul-Sebastian Ungureanu
2019-01-06 22:59 ` Thomas Gummerer
2018-12-20 19:44 ` [PATCH v12 26/26] tests: add a special setup where stash.useBuiltin is off Paul-Sebastian Ungureanu
2019-01-03 23:39 ` [PATCH v12 00/26] Convert "git stash" to C builtin Junio C Hamano
2019-01-18 12:06 ` Johannes Schindelin
2019-01-18 17:49 ` Junio C Hamano
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
List information: http://vger.kernel.org/majordomo-info.html
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=nycvar.QRO.7.76.6.1812272221380.45@tvgsbejvaqbjf.bet \
--to=johannes.schindelin@gmx.de \
--cc=git@vger.kernel.org \
--cc=gitster@pobox.com \
--cc=t.gummerer@gmail.com \
--cc=ungureanupaulsebastian@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.
Code repositories for project(s) associated with this public inbox
https://80x24.org/mirrors/git.git
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).