From: Giuseppe Bilotta <giuseppe.bilotta@gmail.com>
To: Junio C Hamano <gitster@pobox.com>
Cc: git@vger.kernel.org, Heiko Voigt <hvoigt@hvoigt.net>,
msysGit Mailinglist <msysgit@googlegroups.com>,
Johannes Sixt <j.sixt@viscovery.net>,
Jens Lehmann <Jens.Lehmann@web.de>,
Johannes Schindelin <Johannes.Schindelin@gmx.de>
Subject: Re: [PATCHv4 1/6] environment: static list of repo-local env vars
Date: Wed, 24 Feb 2010 20:20:14 +0100 [thread overview]
Message-ID: <cb7bb73a1002241120i4bbbce0bl50460c76db0c6529@mail.gmail.com> (raw)
In-Reply-To: <7vhbp64odi.fsf@alter.siamese.dyndns.org>
On Wed, Feb 24, 2010 at 4:58 PM, Junio C Hamano <gitster@pobox.com> wrote:
> Giuseppe Bilotta <giuseppe.bilotta@gmail.com> writes:
>
>> +/* Repository-local GIT_* environment variables */
>> +const char *const local_repo_env[] = {
>> + ALTERNATE_DB_ENVIRONMENT,
>> + CONFIG_ENVIRONMENT,
>> + DB_ENVIRONMENT,
>> + GIT_DIR_ENVIRONMENT,
>> + GIT_WORK_TREE_ENVIRONMENT,
>> + GRAFT_ENVIRONMENT,
>> + INDEX_ENVIRONMENT,
>> + NO_REPLACE_OBJECTS_ENVIRONMENT,
>> + NULL
>> +};
>> +
>> +const unsigned int local_repo_env_size = ARRAY_SIZE(local_repo_env);
>
> This does not look very useful for two reasons.
>
> - It counts the NULL at the tail, so the number is one-off; you cannot
> say
>
> for (i = 0; i < local_repo_env_size; i++)
> do_something_to(local_repo_env[i]);
This is obviously very easy to fix by decrementing the size by 1 in
the definition.
> - local_repo_env_size is not a compile time constant, so you cannot do:
>
> const char *env[local_repo_env_size + 20];
> memcpy(env, local_repo_env, sizeof(env[0]) * local_repo_env_size);
> for (i = local_repo_env_size; i < ARRAY_SIZE(env); i++)
> add_more_to(env + i, ...);
Indeed, I was just discussing this on the other thread. I personally
have no objection to using this C99 feature.
I guess #defining local_repo_env_size in cache.h, and keeping it
up-to-date when local_repo_env is updated is the best alternative.
This can be done in such a way that if it the array needs expansion,
forgetting to update its size is catched at compile time, but not for
contraction. Of course we don't expect that it will contract. The
disadvantage of this is that a change in cache.h requires an almost
complete recompile because almost everything depends on it. It's a
minor disadvantage, when compared to other approaches (double-walking
the list or other stuff).
(Of course, I still wonder what's wrong with C99 VLAs, but anyway).
--
Giuseppe "Oblomov" Bilotta
next prev parent reply other threads:[~2010-02-24 19:20 UTC|newest]
Thread overview: 15+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-02-24 7:23 [PATCHv4 0/6] repo-local env vars handling Giuseppe Bilotta
2010-02-24 7:23 ` [PATCHv4 1/6] environment: static list of repo-local env vars Giuseppe Bilotta
2010-02-24 15:58 ` Junio C Hamano
2010-02-24 19:20 ` Giuseppe Bilotta [this message]
2010-02-24 19:36 ` Junio C Hamano
2010-02-24 23:11 ` Giuseppe Bilotta
2010-02-24 7:23 ` [PATCHv4 2/6] connect: use " Giuseppe Bilotta
2010-02-24 7:23 ` [PATCHv4 3/6] rev-parse: --local-env-vars option Giuseppe Bilotta
2010-02-24 7:23 ` [PATCHv4 4/6] shell setup: clear_local_git_env() function Giuseppe Bilotta
2010-02-24 7:23 ` [PATCHv4 5/6] submodules: ensure clean environment when operating in a submodule Giuseppe Bilotta
2010-02-24 7:23 ` [PATCHv4 6/6] is_submodule_modified(): clear environment properly Giuseppe Bilotta
2010-02-24 8:06 ` Johannes Sixt
2010-02-24 8:55 ` Giuseppe Bilotta
2010-02-24 9:18 ` Johannes Sixt
2010-02-24 10:07 ` Giuseppe Bilotta
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=cb7bb73a1002241120i4bbbce0bl50460c76db0c6529@mail.gmail.com \
--to=giuseppe.bilotta@gmail.com \
--cc=Jens.Lehmann@web.de \
--cc=Johannes.Schindelin@gmx.de \
--cc=git@vger.kernel.org \
--cc=gitster@pobox.com \
--cc=hvoigt@hvoigt.net \
--cc=j.sixt@viscovery.net \
--cc=msysgit@googlegroups.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 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 NNTP newsgroup(s).