Skip to content

G
git

李少辉-开发者 / git
与 Fork 源项目一致

从无法访问的项目Fork

通知 2
Star 1
Fork 0
G
git

体验新版 GitCode，发现更多精彩内容 >>

切换分支/标签
  1. 17 2月, 2012 7 次提交
    • J
      config: stop using config_exclusive_filename · 270a3443
      由 Jeff King 提交于 
      The git-config command sometimes operates on the default set
of config files (either reading from all, or writing to repo
config), and sometimes operates on a specific file. In the
latter case, we set the magic global config_exclusive_filename,
and the code in config.c does the right thing.

Instead, let's have git-config use the "advanced" variants
of config.c's functions which let it specify an individual
filename (or NULL for the default). This makes the code a
lot more obvious, and fixes two small bugs:

  1. A relative path specified by GIT_CONFIG=foo will look
     in the wrong directory if we have to chdir as part of
     repository setup. We already handle this properly for
     "git config -f foo", but the GIT_CONFIG lookup used
     config_exclusive_filename directly. By dropping to a
     single magic variable, the GIT_CONFIG case now just
     works.

  2. Calling "git config -f foo --edit" would not respect
     core.editor. This is because just before editing, we
     called git_config, which would respect the
     config_exclusive_filename setting, even though this
     particular git_config call was not about looking in the
     user's specified file, but rather about loading actual
     git config, just as any other git program would.
Signed-off-by: NJeff King <peff@peff.net>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>
      270a3443
    • J
      config: provide a version of git_config with more options · c9b5e2a5
      由 Jeff King 提交于 
      Callers may want to provide a specific version of a file in which to look
for config. Right now this can be done by setting the magic global
config_exclusive_filename variable.  By providing a version of git_config
that takes a filename, we can take a step towards making this magic global
go away.

Furthermore, by providing a more "advanced" interface, we now have a a
natural place to add new options for callers like git-config, which care
about tweaking the specifics of config lookup, without disturbing the
large number of "simple" users (i.e., every other part of git).

The astute reader of this patch may notice that the logic for handling
config_exclusive_filename was taken out of git_config_early, but added
into git_config. This means that git_config_early will no longer respect
config_exclusive_filename.  That's OK, because the only other caller of
git_config_early is check_repository_format_gently, but the only function
which sets config_exclusive_filename is cmd_config, which does not call
check_repository_format_gently (and if it did, it would have been a bug,
anyway, as we would be checking the repository format in the wrong file).
Signed-off-by: NJeff King <peff@peff.net>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>
      c9b5e2a5
    • J
      config: teach git_config_rename_section a file argument · 42bd39b5
      由 Jeff King 提交于 
      The other config-writing functions (git_config_set and
git_config_set_multivar) each have an -"in_file" version to
write a specific file. Let's add one for rename_section,
with the eventual goal of moving away from the magic
config_exclusive_filename global.
Signed-off-by: NJeff King <peff@peff.net>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>
      42bd39b5
    • J
      config: teach git_config_set_multivar_in_file a default path · 0a5f5759
      由 Jeff King 提交于 
      The git_config_set_multivar_in_file function takes a
filename argument to specify the file into which the values
should be written. Currently, this value must be non-NULL.
Callers which want to write to the default location must use
the regular, non-"in_file" version, which will either write
to config_exclusive_filename, or to the repo config if the
exclusive filename is NULL.

Let's migrate the "default to using repo config" logic into
the "in_file" form. That will let callers get the same
default-if-NULL behavior as one gets with
config_exclusive_filename, but without having to use the
global variable.
Signed-off-by: NJeff King <peff@peff.net>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>
      0a5f5759
    • J
      config: copy the return value of prefix_filename · 839de252
      由 Jeff King 提交于 
      The prefix_filename function returns a pointer to a static
buffer which may be overwritten by subsequent calls. Since
we are going to keep the result around for a while, let's be
sure to duplicate it for safety.

I don't think this can be triggered as a bug in the current
code, but it's a good idea to be defensive, as any resulting
bug would be quite subtle.
Signed-off-by: NJeff King <peff@peff.net>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>
      839de252
    • J
      t1300: add missing &&-chaining · 27370b11
      由 Jeff King 提交于 
      Signed-off-by: NJeff King <peff@peff.net>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>
      27370b11
    • J
      docs/api-config: minor clarifications · d7be1f14
      由 Jeff King 提交于 
      The first change simply drops some parentheses to make a
statement more clear. The seconds clarifies that almost
nobody wants to call git_config_early.
Signed-off-by: NJeff King <peff@peff.net>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>
      d7be1f14
  3. 07 2月, 2012 1 次提交
  5. 28 1月, 2012 1 次提交
  7. 27 1月, 2012 3 次提交
  9. 24 1月, 2012 1 次提交
  11. 19 1月, 2012 12 次提交
  13. 17 1月, 2012 3 次提交
    • J
      credential-cache: ignore "connection refused" errors · 35a71f14
      由 Jeff King 提交于 
      The credential-cache helper will try to connect to its
daemon over a unix socket. Originally, a failure to do so
was silently ignored, and we would either give up (if
performing a "get" or "erase" operation), or spawn a new
daemon (for a "store" operation).

But since 8ec6c8d7, we try to report more errors. We detect a
missing daemon by checking for ENOENT on our connection
attempt.  If the daemon is missing, we continue as before
(giving up or spawning a new daemon). For any other error,
we die and report the problem.

However, checking for ENOENT is not sufficient for a missing
daemon. We might also get ECONNREFUSED if a dead daemon
process left a stale socket. This generally shouldn't
happen, as the daemon cleans up after itself, but the daemon
may not always be given a chance to do so (e.g., power loss,
"kill -9").

The resulting state is annoying not just because the helper
outputs an extra useless message, but because it actually
blocks the helper from spawning a new daemon to replace the
stale socket.

Fix it by checking for ECONNREFUSED.
Signed-off-by: NJeff King <peff@peff.net>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>
      35a71f14
    • J
      Merge branch 'jn/maint-gitweb-grep-fix' · b63103e9
      由 Junio C Hamano 提交于 
      * jn/maint-gitweb-grep-fix:
  gitweb: Harden "grep" search against filenames with ':'
  gitweb: Fix file links in "grep" search
      b63103e9
    • N
      diff-index: enable recursive pathspec matching in unpack_trees · 4838237c
      由 Nguyen Thai Ngoc Duy 提交于 
      The pathspec structure has a few bits of data to drive various operation
modes after we unified the pathspec matching logic in various codepaths.
For example, max_depth field is there so that "git grep" can limit the
output for files found in limited depth of tree traversal. Also in order
to show just the surface level differences in "git diff-tree", recursive
field stops us from descending into deeper level of the tree structure
when it is set to false, and this also affects pathspec matching when
we have wildcards in the pathspec.

The diff-index has always wanted the recursive behaviour, and wanted to
match pathspecs without any depth limit. But we forgot to do so when we
updated tree_entry_interesting() logic to unify the pathspec matching
logic.
Signed-off-by: NNguyễn Thái Ngọc Duy <pclouds@gmail.com>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>
      4838237c
  15. 15 1月, 2012 1 次提交
  17. 14 1月, 2012 2 次提交
  19. 13 1月, 2012 9 次提交