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

Re: [PATCH v3] Expand ~ and ~user in core.excludesfile, commit.template

From
Junio C Hamano <gitster@pobox.com>
Date
Aug 29, 2008, 04:08 UTC
Message-ID
<7vod3ca2ey.fsf@gitster.siamese.dyndns.org>
In-Reply-To
<20080829032630.GA7024@coredump.intra.peff.net>
Jeff King <peff@peff.net> writes:
Show 15 quoted lines
> On Thu, Aug 28, 2008 at 02:09:38AM -0700, Karl Chen wrote:
>
>>  builtin-commit.c |    2 +-
>>  cache.h          |    2 +
>>  config.c         |   13 +++++++-
>>  path.c           |   88 +++++++++++++++++++++++++++++++++--------------------
>>  4 files changed, 69 insertions(+), 36 deletions(-)
>
> Documentation?
>
>>  	if (!strcmp(k, "commit.template"))
>> -		return git_config_string(&template_file, k, v);
>> +		return git_config_userdir(&template_file, k, v);
>
> I like this.

Likewise, except for the name. It is more like "pathname"; "userdir" is stressing only one aspect of the magic we would do to a value that is a pathname compared to a value that is a string without any magic.

Show 11 quoted lines
>> +int git_config_userdir(const char **dest, const char *var, const char *value) {
>> +	if (!value)
>> +		return config_error_nonbool(var);
>> +	*dest = expand_user_path(NULL, value, 0);
>> +	if (!*dest || !**dest) die("Failed to expand user dir in: '%s'", value);
>> +	return 0;
>> +}
>
> I am not sure about !**dest here. This precludes somebody from using "".
> While it might not matter here, if there are other users of
> git_config_userdir(), they might want to allow a blank entry.
True again.
Show 5 quoted lines
>> +	if (begin_username == end_username) {
>> +		return getpwuid(getuid());
>> +	} else {
>
> Style: omit braces on one-liner conditionals:

... except in cases like this where "else" side is a multi-statement block, in which case the above is fine. But as you pointed out, the early return from here makes the else block unnecessary so you do not need the braces around "if" side.

>> +		char *username = alloca(username_len + 1);
>
> I don't think we use alloca() anywhere else. I don't know if there are
> portability issues.

Avoidance of alloca() and c99 dynamic array on stack is deliberate in the current codebase. Portable use of alloca() is quite hard to get right.

Show 7 quoted lines
>> +static inline const char *strchr_or_end(const char *str, char c)
>> +{
>> +	while (*str && *str != c) ++str;
>> +	return str;
>> +}
>
> This really seems like premature optimization to me.
So is overuse of inline, it seems.
Previous: Jeff KingNext: Karl Chen
Message 35 of 45 in “Support "core.excludesfile = ~/.gitignore"”
  1. Support "core.excludesfile = ~/.gitignore"Karl Chen, Aug 22, 2008
  2. Eric RaibleAug 22, 2008
  3. Bert WesargAug 22, 2008
  4. Junio C HamanoAug 22, 2008
  5. Karl ChenAug 24, 2008
  6. Junio C HamanoAug 24, 2008
  7. Jeff KingAug 24, 2008
  8. Junio C HamanoAug 24, 2008
  9. Jeff KingAug 24, 2008
  10. Junio C HamanoAug 24, 2008
  11. limiting relationship of git dir and worktree (was Re: [PATCH] Support "core.excludesfile = ~/.gitignore")Jeff King, Aug 24, 2008
  12. Dropping core.worktree and GIT_WORK_TREE support (was Re: limiting relationship of git dir and worktree)Junio C Hamano, Aug 25, 2008
  13. Miklos VajnaAug 25, 2008
  14. Junio C HamanoAug 25, 2008
  15. Miklos VajnaAug 25, 2008
  16. Nguyen Thai Ngoc DuyAug 25, 2008
  17. git diff/diff-index/diff-files: call setup_work_tree()Miklos Vajna, Aug 25, 2008
  18. Nguyen Thai Ngoc DuyAug 25, 2008
  19. Miklos VajnaAug 25, 2008
  20. git diff/diff-index/diff-files: call setup_work_tree()Miklos Vajna, Aug 25, 2008
  21. Nguyen Thai Ngoc DuyAug 25, 2008
  22. Junio C HamanoAug 26, 2008
  23. diff*: fix worktree setupNguyễn Thái Ngọc Duy, Aug 28, 2008
  24. Junio C HamanoAug 25, 2008
  25. Miklos VajnaAug 25, 2008
  26. Michael J GruberAug 26, 2008
  27. Jeff KingAug 27, 2008
  28. Support "core.excludesfile = ~/.gitignore"Karl Chen, Aug 25, 2008
  29. Johannes SixtAug 26, 2008
  30. Jeff KingAug 27, 2008
  31. Karl ChenAug 27, 2008
  32. Junio C HamanoAug 27, 2008
  33. Expand ~ and ~user in core.excludesfile, commit.templateKarl Chen, Aug 28, 2008
  34. Jeff KingAug 29, 2008
  35. Junio C HamanoAug 29, 2008
  36. Expand ~ and ~user in core.excludesfile, commit.templateKarl Chen, Aug 29, 2008
  37. Junio C HamanoAug 29, 2008
  38. Karl ChenAug 29, 2008
  39. Junio C HamanoAug 29, 2008
  40. Karl ChenAug 29, 2008
  41. Junio C HamanoAug 30, 2008
  42. Jeff KingAug 30, 2008
  43. Johannes SixtAug 29, 2008
  44. Karl ChenAug 27, 2008
  45. Junio C HamanoAug 27, 2008

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.