Code

git.git
18 years agogit-commit: finishing touches.
Junio C Hamano [Mon, 6 Feb 2006 00:08:01 +0000 (16:08 -0800)]
git-commit: finishing touches.

Introduce --only flag to allow the new "partial commit"
semantics when paths are specified.  The default is still the
traditional --include semantics.  Once peoples' fingers and
scripts that want the traditional behaviour are updated to
explicitly say --include, we could change it to either default
to --only, or refuse to operate without either --only/--include
when paths are specified.

This also fixes a couple of bugs in the previous round.  Namely:

 - forgot to save/restore index in some cases.

 - forgot to use the temporary index to show status when '--only
   paths...' semantics was used.

 - --author did not take precedence when reusing an existing
   commit.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agogit-commit: revamp the git-commit semantics.
Junio C Hamano [Sun, 5 Feb 2006 08:07:44 +0000 (00:07 -0800)]
git-commit: revamp the git-commit semantics.

 - "git commit" without _any_ parameter keeps the traditional
   behaviour.  It commits the current index.

   We commit the whole index even when this form is run from a
   subdirectory.

 - "git commit --include paths..." (or "git commit -i paths...")
   is equivalent to:

    git update-index --remove paths...
        git commit

 - "git commit paths..." acquires a new semantics.  This is an
   incompatible change that needs user training, which I am
   still a bit reluctant to swallow, but enough people seem to
   have complained that it is confusing to them.  It

   1. refuses to run if $GIT_DIR/MERGE_HEAD exists, and reminds
      trained git users that the traditional semantics now needs
      -i flag.

   2. refuses to run if named paths... are different in HEAD and
      the index (ditto about reminding).  Added paths are OK.

   3. reads HEAD commit into a temporary index file.

   4. updates named paths... from the working tree in this
      temporary index.

   5. does the same updates of the paths... from the working
      tree to the real index.

   6. makes a commit using the temporary index that has the
      current HEAD as the parent, and updates the HEAD with this
      new commit.

 - "git commit --all" can run from a subdirectory, but it updates
   the index with all the modified files and does a whole tree
   commit.

 - In all cases, when the command decides not to create a new
   commit, the index is left as it was before the command is
   run.  This means that the two "git diff" in the following
   sequence:

       $ git diff
       $ git commit -a
       $ git diff

   would show the same diff if you abort the commit process by
   making the commit log message empty.

This commit also introduces much requested --author option.

$ git commit --author 'A U Thor <author@example.com>'

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agogit-rerere: reuse recorded resolve.
Junio C Hamano [Sun, 29 Jan 2006 07:15:24 +0000 (23:15 -0800)]
git-rerere: reuse recorded resolve.

In a workflow that employs relatively long lived topic branches,
the developer sometimes needs to resolve the same conflict over
and over again until the topic branches are done (either merged
to the "release" branch, or sent out and accepted upstream).

This commit introduces a new command, "git rerere", to help this
process by recording the conflicted automerge results and
corresponding hand-resolve results on the initial manual merge,
and later by noticing the same conflicted automerge and applying
the previously recorded hand resolution using three-way merge.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agofmt-merge-msg: show summary of what is merged.
Junio C Hamano [Sat, 4 Feb 2006 03:20:16 +0000 (19:20 -0800)]
fmt-merge-msg: show summary of what is merged.

In addition to the branch names, populate the log message with
one-line description from actual commits that are being merged.

This was prompted by Len's 12-way octopus.  You need to have
'merge.summary' in the configuration file to enable it:

$ git repo-config merge.summary yes

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoread-tree --aggressive
Junio C Hamano [Sat, 4 Feb 2006 06:04:14 +0000 (22:04 -0800)]
read-tree --aggressive

A new flag --aggressive resolves what we traditionally resolved
with external git-merge-one-file inside index while read-tree
3-way merge works.

git-merge-octopus and git-merge-resolve use this flag before
running git-merge-index with git-merge-one-file.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years ago[PATCH] mailinfo: reset CTE after each multipart
Junio C Hamano [Tue, 7 Feb 2006 05:35:25 +0000 (21:35 -0800)]
[PATCH] mailinfo: reset CTE after each multipart

If the first part uses quoted-printable to protect iso8859-1
name in the commit log, and the second part was plain ascii text
patchfile without even Content-Transfer-Encoding subheader, we
incorrectly tried to decode the patch as quoted printable.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoDocs: minor git-push copyediting
J. Bruce Fields [Mon, 6 Feb 2006 00:42:27 +0000 (19:42 -0500)]
Docs: minor git-push copyediting

Minor git-push copyediting

Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoDocs: move git url and remotes text to separate sections
J. Bruce Fields [Sun, 5 Feb 2006 23:29:49 +0000 (18:29 -0500)]
Docs: move git url and remotes text to separate sections

The sections on git urls and remotes files in the git-fetch,
git-pull, and git-push manpages seem long enough to be worth a
manpage section of their own.

Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoDocs: split up pull-fetch-param.txt
J. Bruce Fields [Sun, 5 Feb 2006 22:43:47 +0000 (17:43 -0500)]
Docs: split up pull-fetch-param.txt

The push and pull man pages include a bunch of shared text from
pull-fetch-param.txt.  This simplifies maintenance somewhat, but
there's actually quite a bit of text that applies only to one or the
other.

So, separate out the push- and pull/fetch-specific text into
pull-fetch-param.txt and git-push.txt, then include the largest chunk
of common stuff (the description of protocols and url's) from
urls.txt.  That cuts some irrelevant stuff from the man pages without
making us duplicate too much.

Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agocombine-diff: do not punt on removed or added files.
Junio C Hamano [Tue, 7 Feb 2006 02:54:08 +0000 (18:54 -0800)]
combine-diff: do not punt on removed or added files.

When we remove a file, the parents' contents are all removed so
it is not that interesting to show all of them, but the fact it
was removed when all parents had it *is* unusual.  When we add a
file, similarly the fact it was added when no parent wanted it
*is* unusual, and in addition the result matters, so show it.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agocombine-diff: show mode changes as well.
Junio C Hamano [Mon, 6 Feb 2006 20:53:07 +0000 (12:53 -0800)]
combine-diff: show mode changes as well.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agocombine-diff: do not send NULL to printf
Junio C Hamano [Mon, 6 Feb 2006 20:30:00 +0000 (12:30 -0800)]
combine-diff: do not send NULL to printf

When we run combined diff from working tree (diff-files --cc),
we sent NULL to printf that is returned by find_unique_abbrev().

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agocore-tutorial: adjust to recent reality.
Junio C Hamano [Mon, 6 Feb 2006 20:27:33 +0000 (12:27 -0800)]
core-tutorial: adjust to recent reality.

We still talked about HEAD symlinks but these days we use
symrefs by default.

Also 'failed/prevented' message is now gone from the merge
output.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agogit-diff: do not fall back on --cc when -[123], --ours etc. are given.
Junio C Hamano [Mon, 6 Feb 2006 19:17:56 +0000 (11:17 -0800)]
git-diff: do not fall back on --cc when -[123], --ours etc. are given.

These flags ask diff with a specific unmerged stage, so it
should fall back on -p instead.  Also when -c is given, we
should not do --cc.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoMerge branch 'jc/diff'
Junio C Hamano [Mon, 6 Feb 2006 07:55:45 +0000 (23:55 -0800)]
Merge branch 'jc/diff'

* jc/diff:
  git-diff-tree --stdin: show all parents.
  combine-diff: remove misguided --show-empty hack.

18 years agoMerge branches 'lt/show' and 'lt/revlist'
Junio C Hamano [Mon, 6 Feb 2006 07:55:09 +0000 (23:55 -0800)]
Merge branches 'lt/show' and 'lt/revlist'

* lt/show:
  git-show

* lt/revlist:
  rev-parse lstat() workaround cleanup.

18 years agoMerge branches 'jc/daemon' and 'mw/http'
Junio C Hamano [Mon, 6 Feb 2006 07:54:14 +0000 (23:54 -0800)]
Merge branches 'jc/daemon' and 'mw/http'

* jc/daemon:
  daemon: extend user-relative path notation.
  daemon: Set SO_REUSEADDR on listening sockets.
  daemon: do not forbid user relative paths unconditionally under --base-path

* mw/http:
  http-fetch: Tidy control flow in process_alternate_response
  http: Turn on verbose Curl messages if GIT_CURL_VERBOSE set in environment
  http-fetch: Fix message reporting rename of object file.
  http-fetch: Fix object list corruption in fill_active_slots().

18 years agogit-diff-tree --stdin: show all parents.
Junio C Hamano [Mon, 6 Feb 2006 07:00:41 +0000 (23:00 -0800)]
git-diff-tree --stdin: show all parents.

git-diff-tree --stdin ignored second and subsequent parents when
fed git-rev-list --parents output.  Update diff_tree_commit()
function to take a commit object, and pass a fabricated commit
object after grafting the fake parents from diff_tree_stdin().

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agocombine-diff: remove misguided --show-empty hack.
Junio C Hamano [Mon, 6 Feb 2006 06:25:00 +0000 (22:25 -0800)]
combine-diff: remove misguided --show-empty hack.

Now --always flag is available in diff-tree, there is no reason
to have that hack in the diffcore side.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agorev-parse lstat() workaround cleanup.
Linus Torvalds [Mon, 6 Feb 2006 05:41:47 +0000 (21:41 -0800)]
rev-parse lstat() workaround cleanup.

Earlier we had a workaround to avoid misspelled revision name to
be taken as a filename when "--no-revs --no-flags" are in
effect.  This cleans up the logic.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agogit-send-email: Fully implement --quiet and document it.
Ryan Anderson [Mon, 6 Feb 2006 01:13:52 +0000 (20:13 -0500)]
git-send-email: Fully implement --quiet and document it.

Also reorganizes the man page to list options alphabetically.

Signed-off-by: Ryan Anderson <ryan@michonline.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agodaemon: extend user-relative path notation.
Junio C Hamano [Sun, 5 Feb 2006 06:27:29 +0000 (22:27 -0800)]
daemon: extend user-relative path notation.

Earlier, we made --base-path to automatically forbid
user-relative paths, which was probably a mistake.  This
introduces --user-path (or --user-path=path) option to control
the use of user-relative paths independently.  The latter form
of the option can be used to restrict accesses to a part of each
user's home directory, similar to "public_html" some webservers
supports.

If we're invoked with --user-path=FOO option, then a URL of the
form git://~USER/PATH/... resolves to the path HOME/FOO/PATH/...,
where HOME is USER's home directory.

[jc: This is much reworked by me so bugs are mine, but the
 original patch was done by Mark Wooding.]

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agodaemon: Set SO_REUSEADDR on listening sockets.
Mark Wooding [Fri, 3 Feb 2006 20:27:04 +0000 (20:27 +0000)]
daemon: Set SO_REUSEADDR on listening sockets.

Without this, you can silently lose the ability to receive IPv4
connections if you stop and restart the daemon.

[jc: tweaked code organization a bit and made this controllable
 from a command line option.]

Signed-off-by: Mark Wooding <mdw@distorted.org.uk>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agodaemon: do not forbid user relative paths unconditionally under --base-path
Junio C Hamano [Sat, 4 Feb 2006 07:50:55 +0000 (23:50 -0800)]
daemon: do not forbid user relative paths unconditionally under --base-path

Using base-path to relocate the server public space does not
have anything to do with allowing or forbidding user relative
paths.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agohttp-fetch: Tidy control flow in process_alternate_response
Mark Wooding [Wed, 1 Feb 2006 11:44:39 +0000 (11:44 +0000)]
http-fetch: Tidy control flow in process_alternate_response

It's a bit convoluted.  Tidy it up.

Signed-off-by: Mark Wooding <mdw@distorted.org.uk>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agohttp: Turn on verbose Curl messages if GIT_CURL_VERBOSE set in environment
Mark Wooding [Wed, 1 Feb 2006 11:44:37 +0000 (11:44 +0000)]
http: Turn on verbose Curl messages if GIT_CURL_VERBOSE set in environment

Signed-off-by: Mark Wooding <mdw@distorted.org.uk>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agohttp-fetch: Fix message reporting rename of object file.
Mark Wooding [Wed, 1 Feb 2006 11:44:35 +0000 (11:44 +0000)]
http-fetch: Fix message reporting rename of object file.

move_temp_to_file returns 0 or -1.  This is not a good thing to pass to
strerror(3).  Fortunately, someone already reported the error, so don't
worry too much.

Signed-off-by: Mark Wooding <mdw@distorted.org.uk>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agohttp-fetch: Fix object list corruption in fill_active_slots().
Mark Wooding [Wed, 1 Feb 2006 11:44:28 +0000 (11:44 +0000)]
http-fetch: Fix object list corruption in fill_active_slots().

In fill_active_slots() -- if we find an object which has already arrived,
say as part of a pack, /don't/ remove it from the list.  It's already been
prefetched and someone will ask for it later.  Just label it as done and
carry blithely on.  (As it was, the code would dereference a freed object
to continue through the list anyway.)

Signed-off-by: Mark Wooding <mdw@distorted.org.uk>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agogit-show
Junio C Hamano [Mon, 6 Feb 2006 00:42:49 +0000 (16:42 -0800)]
git-show

This is essentially 'git whatchanged -n1 --always --cc "$@"'.
Just like whatchanged takes default flags from
whatchanged.difftree configuration, this uses show.difftree
configuration.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agogit-diff: use --cc instead of -p.
Junio C Hamano [Mon, 6 Feb 2006 00:37:18 +0000 (16:37 -0800)]
git-diff: use --cc instead of -p.

The --cc output is much nicer when dealing with merges, so use
it by default.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agodiff-index: make --cc a synonym for -p for now.
Junio C Hamano [Mon, 6 Feb 2006 00:36:12 +0000 (16:36 -0800)]
diff-index: make --cc a synonym for -p for now.

It could be made later to show unmerged state nicer than the
default as we did for diff-files later, but this would suffice
for now.  We would like to make --cc the default for 'git diff'.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agodiff-tree --always flag
Linus Torvalds [Sun, 5 Feb 2006 19:42:09 +0000 (11:42 -0800)]
diff-tree --always flag

It _might_ make sense for certain users like gitk and gitview if
we had a single tool that gives --pretty and its diff even if
the diff is empty.  Having said that, the flag --cc -m is too
specific.  If some uses want to see the commit log even for an
empty diff, that flag should not be something only --cc honors.

Here's an "--always" flag that does that.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoUse adler32() from zlib instead of defining our own.
Peter Eriksen [Sat, 4 Feb 2006 22:54:01 +0000 (23:54 +0100)]
Use adler32() from zlib instead of defining our own.

Since we already depend on zlib, we don't need to define our
own adler32().  Spotted by oprofile.

Signed-off-by: Peter Eriksen <s022018@student.dtu.dk>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoFix git-rev-parse over-eager errors
Linus Torvalds [Sun, 5 Feb 2006 19:58:34 +0000 (11:58 -0800)]
Fix git-rev-parse over-eager errors

Using "--verify" together with "--no-flags" makes perfect sense, but
git-rev-parse would complain about it when it saw a flag, even though it
would never actually use/output that flag.

This fixes it.

Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoDo not fall back on vi on dumb terminals.
Junio C Hamano [Sun, 5 Feb 2006 06:10:32 +0000 (22:10 -0800)]
Do not fall back on vi on dumb terminals.

When TERM is set to 'dumb', do not start vi to edit the commit log
message.

Suggested by Amos Waterland.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoMerge branches 'jc/sha1', 'jc/diff' and 'jc/ws'
Junio C Hamano [Sat, 4 Feb 2006 07:52:20 +0000 (23:52 -0800)]
Merge branches 'jc/sha1', 'jc/diff' and 'jc/ws'

* jc/sha1:
  get_sha1_1: allow octopus^12 to be properly parsed.

* jc/diff:
  combine-diff: finishing touches to git-diff-tree --cc

* jc/ws:
  whitespace cleanup.

18 years agoUse sha1_file.c's mkdir-like routine in apply.c.
Jason Riedy [Sat, 4 Feb 2006 06:50:57 +0000 (22:50 -0800)]
Use sha1_file.c's mkdir-like routine in apply.c.

As far as I can see, create_subdirectories() in apply.c just
duplicates the functionality of safe_create_leading_directories() from
sha1_file.c.  The former has a warm, fuzzy const parameter, but that's
not important.

The potential problem with EEXIST and creating directories should
never occur here, but will be removed by future
safe_create_leading_directories() changes.  Other uses of EEXIST in
apply.c should be fine barring intentionally malicious behavior.

Signed-off-by: Jason Riedy <ejr@cs.berkeley.edu>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agodaemon: Provide missing argument for logerror() call.
Mark Wooding [Fri, 3 Feb 2006 20:27:02 +0000 (20:27 +0000)]
daemon: Provide missing argument for logerror() call.

Could cause a crash if --base-path set.  Unlikely to be a security the
concern: message doesn't go to the client, so we can't leak anything
(except by dumping core), and we've already forked, so it's not a denial
of service.

Signed-off-by: Mark Wooding <mdw@distorted.org.uk>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agogit-merge: Properly quote $merge_msg variable.
Junio C Hamano [Sat, 4 Feb 2006 03:41:33 +0000 (19:41 -0800)]
git-merge: Properly quote $merge_msg variable.

Otherwise it would go though shell expansion...

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoget_sha1_1: allow octopus^12 to be properly parsed.
Junio C Hamano [Fri, 3 Feb 2006 07:48:36 +0000 (23:48 -0800)]
get_sha1_1: allow octopus^12 to be properly parsed.

We probably thought anybody who does more than 9 parents in an
Octopus is insane when this was initially done, but there is no
inherent reason to limit the number of independent topic
branches that happen to mature at the same time.

Our commit-tree allows up to 16 already, so at least we should
prepare to handle what we can produce, if only to be consistent.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agowhitespace cleanup.
Junio C Hamano [Wed, 1 Feb 2006 02:15:51 +0000 (18:15 -0800)]
whitespace cleanup.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agocombine-diff: finishing touches to git-diff-tree --cc
Junio C Hamano [Fri, 3 Feb 2006 01:43:47 +0000 (17:43 -0800)]
combine-diff: finishing touches to git-diff-tree --cc

This updates the output format to make administrative lines more
consistent with the traditional diffs.

The "index" line shows blob object names from each parents
(separated by commas), double dots and the object name of the
resulting blob.

The hunk header line begins with N+1 '@' characters for N-way
diff, the line number L of the first line in the hunk and line
count C from the parent in "-L,C" format for each parents and
then the line number of the first line in the hunk and line
count from the resulting file in "+L,C" format, and finally
N+1 '@' characters (earlier versions had the line numbers from
the resulting file at the beginning).

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agocombine-diff: cleanup.
Junio C Hamano [Thu, 2 Feb 2006 23:17:42 +0000 (15:17 -0800)]
combine-diff: cleanup.

The flag on the surviving lines meant "this parent is not
different" while the parent_map flag on the lost lines meant
"this parent is different", which was confusing.  So swap the
meaning of on-bit in the flag.  Also more heavily comment the
code.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agocombine-diff: show parent line numbers as well.
Junio C Hamano [Tue, 31 Jan 2006 06:33:15 +0000 (22:33 -0800)]
combine-diff: show parent line numbers as well.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agocombine-diff: add a bit more comments.
Junio C Hamano [Tue, 31 Jan 2006 00:34:29 +0000 (16:34 -0800)]
combine-diff: add a bit more comments.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agogit-send-email: Add --quiet to reduce some of the chatter when sending emails.
Ryan Anderson [Thu, 2 Feb 2006 16:56:06 +0000 (11:56 -0500)]
git-send-email: Add --quiet to reduce some of the chatter when sending emails.

Signed-off-by: Ryan Anderson <ryan@michonline.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoProvide a more meaningful initial "From " line when using --compose in git-send-email.
Ryan Anderson [Thu, 2 Feb 2006 16:56:06 +0000 (11:56 -0500)]
Provide a more meaningful initial "From " line when using --compose in git-send-email.

git-send-email, when used with --compose, provided the user with a mbox-format
file to edit.  Some users, however, were confused by the leading, blank, "From
" line, so this change puts the value that will appear on the From: line of the
actual email on this line, along with a note that the line is ignored.

Signed-off-by: Ryan Anderson <ryan@michonline.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agocommit.c: "Merge" fix in pretty_print_commit.
Junio C Hamano [Fri, 3 Feb 2006 01:52:19 +0000 (17:52 -0800)]
commit.c: "Merge" fix in pretty_print_commit.

Earlier, f2d4227530499db3e273ae84f30adfd4b70791c6 commit broke Merge:
lines for unabbreviated case.  Do not emit extra dots if we do not
abbreviate.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agomerge-recursive: Speed up commit graph construction
Fredrik Kuivinen [Thu, 2 Feb 2006 11:43:35 +0000 (12:43 +0100)]
merge-recursive: Speed up commit graph construction

Use __slots__ to speed up construction and decrease memory consumption
of the Commit objects.

Signed-off-by: Fredrik Kuivinen <freku045@student.liu.se>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agomerge-recursive: Make use of provided bases
Fredrik Kuivinen [Thu, 2 Feb 2006 11:43:20 +0000 (12:43 +0100)]
merge-recursive: Make use of provided bases

This makes some cases faster as we don't have to build the commit graph.

Signed-off-by: Fredrik Kuivinen <freku045@student.liu.se>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoDocumentation: git-diff-tree --cc also omits empty commits
Junio C Hamano [Thu, 2 Feb 2006 19:28:38 +0000 (11:28 -0800)]
Documentation: git-diff-tree --cc also omits empty commits

A misguided attempt to show logs at all time was inserted only to
the documentation of this flag.  Worse yet, it was not even implemented,
causing more confusion.  Drop it.

We might want to have an option to show --pretty even when there is no
diff output, but that is applicable to all forms of diff, not just --cc.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agocombine-diff: fix placement of deletion.
Junio C Hamano [Thu, 2 Feb 2006 13:21:14 +0000 (05:21 -0800)]
combine-diff: fix placement of deletion.

The code misplaced a raw hunk that consists of solely deleted
lines by one line.  This showed e.g. Len's 12-way octopus
(9fdb62af in the linux-2.6), kernel/power/disk.c, hunk starting
at line 95, incorrectly.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agocombine-diff: add safety check to --cc.
Junio C Hamano [Thu, 2 Feb 2006 09:28:08 +0000 (01:28 -0800)]
combine-diff: add safety check to --cc.

The earlier change implemented "only two version" check but
without checking if the change rewrites from all the parents.
This implements a check to make sure that a change introduced
by the merge from all the parents is caught to be interesting.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agocombine-diff: update --cc "uninteresting hunks" logic.
Junio C Hamano [Thu, 2 Feb 2006 08:12:55 +0000 (00:12 -0800)]
combine-diff: update --cc "uninteresting hunks" logic.

Earlier logic was discarding hunks that has difference from only
one parent or the same difference from all but one parent.  This
changes it to check if the differences on all lines are from the
same sets of parents.  This discards more uninteresting hunks
and seems to match expectations more naturally.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agocombine-diff: reuse diff from the same blob.
Junio C Hamano [Thu, 2 Feb 2006 07:29:03 +0000 (23:29 -0800)]
combine-diff: reuse diff from the same blob.

When dealing with an insanely large Octopus, it is possible to
optimize by noticing that more than one parents have the same
blob and avoid running diff between a parent and the merge
result by reusing an earlier result.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoAllow diff and index commands to be interrupted
Petr Baudis [Tue, 31 Jan 2006 23:40:33 +0000 (00:40 +0100)]
Allow diff and index commands to be interrupted

So far, e.g. git-update-index --refresh was basically uninterruptable
by ctrl-c, since it hooked the SIGINT handler, but that handler would
only unlink the lockfile but not actually quit. This makes it propagate
the signal to the default handler.

Note that I expected it to work without resetting the signal handler to
SIG_DFL, but without that it ended in an infinite loop of tgkill()s -
is my glibc violating SUS or what?

Signed-off-by: Petr Baudis <pasky@suse.cz>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agorev-list: omit duplicated parents.
Junio C Hamano [Sun, 29 Jan 2006 23:24:42 +0000 (15:24 -0800)]
rev-list: omit duplicated parents.

Showing the same parent more than once for a commit does not
make much sense downstream, so stop it.

This can happen with an incorrectly made merge commit that
merges the same parent twice, but can happen in an otherwise
sane development history while squishing the history by taking
into account only commits that touch specified paths.

For example,

$ git rev-list --max-count=1 --parents addafaf -- rev-list.c

would have to show this commit ancestry graph:

                  .---o---.
                 /         \
                .---*---o---.
               /    93b74bc  \
   ---*---o---o-----o---o-----o addafaf
      d8f6b34  \             /
                .---o---o---.
                 \         /
                  .---*---.
                      3815f42

where 5 independent development tracks, only two of which have
changes in the specified paths since they forked.  The last
change for the other three development tracks was done by the
same commit before they forked, and we were showing that three
times.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoupdate-index --index-info: allow stage 0 entries.
Junio C Hamano [Wed, 1 Feb 2006 02:03:37 +0000 (18:03 -0800)]
update-index --index-info: allow stage 0 entries.

Somehow we did not allow stuffing the index with stage 0 entries
through --index-info interface.  I do not think of a reason to
forbid it offhand.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agogit-svnimport.perl: fix for 'arg list too long...'
Sasha Khapyorsky [Wed, 1 Feb 2006 15:53:31 +0000 (17:53 +0200)]
git-svnimport.perl: fix for 'arg list too long...'

This fixes 'arg list too long..' problem with git-ls-files.

Note that second arg list separation loop (with 'git-update-index') is
needed since git-ls-files arguments can be directories.

Signed-off-by: Sasha Khapyorsky <sashak@voltaire.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoUse local structs for HTTP slot callback data
Nick Hengeveld [Wed, 1 Feb 2006 02:00:37 +0000 (18:00 -0800)]
Use local structs for HTTP slot callback data

There's no need for these structures to be static, and it could potentially
cause problems down the road.

Signed-off-by: Nick Hengeveld <nickh@reactrix.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agorev-list: allow -<n> as shorthand for --max-count=<n>
Eric Wong [Mon, 30 Jan 2006 00:28:02 +0000 (16:28 -0800)]
rev-list: allow -<n> as shorthand for --max-count=<n>

This builds on top of the previous one.

Traditionally, head(1) and tail(1) allow their line limits to be
parsed this way.

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agorev-list: allow -n<n> as shorthand for --max-count=<n>
Eric Wong [Mon, 30 Jan 2006 00:26:40 +0000 (16:26 -0800)]
rev-list: allow -n<n> as shorthand for --max-count=<n>

Both -n<n> and -n <n> are supported.  POSIX versions of head(1) and
tail(1) allow their line limits to be parsed this way.  I find
--max-count to be a commonly used option, and also similar in spirit to
head/tail, so I decided to make life easier on my worn out (and lazy :)
fingers with this patch.

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoMake apply accept the -pNUM option like patch does.
Daniel Barkalow [Tue, 31 Jan 2006 05:36:24 +0000 (00:36 -0500)]
Make apply accept the -pNUM option like patch does.

This only applies to traditional diffs, not to git diffs.

Signed-off-by: Daniel Barkalow <barkalow@iabervon.org>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoFix HTTP request result processing after slot reuse
Nick Hengeveld [Tue, 31 Jan 2006 19:06:55 +0000 (11:06 -0800)]
Fix HTTP request result processing after slot reuse

Add a way to store the results of an HTTP request when a slot finishes
so the results can be processed after the slot has been reused.

Signed-off-by: Nick Hengeveld <nickh@reactrix.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoMake git-tar-tree use the tree_desc abstractions
Linus Torvalds [Tue, 31 Jan 2006 22:13:40 +0000 (14:13 -0800)]
Make git-tar-tree use the tree_desc abstractions

Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoMake the "struct tree_desc" operations available to others
Linus Torvalds [Tue, 31 Jan 2006 22:10:56 +0000 (14:10 -0800)]
Make the "struct tree_desc" operations available to others

We have operations to "extract" and "update" a "struct tree_desc", but we
only used them in tree-diff.c and they were static to that file.

But other tree traversal functions can use them to their advantage

Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoMerge branches 'bf/doc' and 'db/tartree'
Junio C Hamano [Tue, 31 Jan 2006 06:41:00 +0000 (22:41 -0800)]
Merge branches 'bf/doc' and 'db/tartree'

18 years agodocumentation: cvs migration - typofix.
Junio C Hamano [Tue, 31 Jan 2006 05:01:25 +0000 (21:01 -0800)]
documentation: cvs migration - typofix.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoUse struct commit in tar-tree
Daniel Barkalow [Sun, 29 Jan 2006 19:05:20 +0000 (14:05 -0500)]
Use struct commit in tar-tree

It was open-coding getting the commit date from a commit.

Signed-off-by: Daniel Barkalow <barkalow@iabervon>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoUse struct tree in tar-tree
Daniel Barkalow [Sun, 29 Jan 2006 19:04:59 +0000 (14:04 -0500)]
Use struct tree in tar-tree

It was using an open-coded tree parser; use a struct tree instead.

Signed-off-by: Daniel Barkalow <barkalow@iabervon.org>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agocvs-migration documentation update
J. Bruce Fields [Sun, 29 Jan 2006 04:31:47 +0000 (23:31 -0500)]
cvs-migration documentation update

Here's some changes to the cvs-migration.txt.  As usual, in my attempt
to make things clearer someone may have found I've made them less so, or
I may have just gotten something wrong; so any review is welcomed.

I can break up this sort of thing into smaller steps if preferred, the
monolothic patch is just a bit simpler for me for this sort of
thing.

I moved the material describing shared repository management from
core-tutorial.txt to cvs-migration.txt, where it seems more appropriate,
and combined two sections to eliminate some redundancy.

I also revised the earlier sections of cvs-migration.txt, mainly trying
to make it more concise.

I've left the last section of cvs-migration.txt (on CVS annotate
alternatives) alone for now.

Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agocvsexportcommit: add some examples to the documentation
Martin Langhoff [Mon, 30 Jan 2006 06:34:15 +0000 (19:34 +1300)]
cvsexportcommit: add some examples to the documentation

Updated with Randall Schwartz's suggestion.

Signed-off-by: Martin Langhoff <martin@catalyst.net.nz>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoexportcommit: replace backticks with safe_pipe_capture() or system() - initial pass
Martin Langhoff [Mon, 30 Jan 2006 06:12:12 +0000 (19:12 +1300)]
exportcommit: replace backticks with safe_pipe_capture() or system() - initial pass

Replaced backticks with potentially troublesome unescaped input with
safe_pipe_capture().

Signed-off-by: Martin Langhoff <martin@catalyst.net.nz>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agocvsimport: Add -S <skipfileregex> support and -v announces files retrieved
Martin Langhoff [Mon, 30 Jan 2006 06:12:41 +0000 (19:12 +1300)]
cvsimport: Add -S <skipfileregex> support and -v announces files retrieved

A couple of things that seem to help importing broken CVS repos...

 -S '<slash-delimited-regex>' skips files with a matching path
 -v prints file name and version before fetching from cvs

Signed-off-by: Martin Langhoff <martin@catalyst.net.nz>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoMerge fixes up to GIT 1.1.6
Junio C Hamano [Mon, 30 Jan 2006 01:56:45 +0000 (17:56 -0800)]
Merge fixes up to GIT 1.1.6

18 years agoGIT 1.1.6 v1.1.6
Junio C Hamano [Mon, 30 Jan 2006 01:55:20 +0000 (17:55 -0800)]
GIT 1.1.6

18 years agogit push -f documentation
J. Bruce Fields [Mon, 30 Jan 2006 01:40:50 +0000 (20:40 -0500)]
git push -f documentation

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agogit-branch: Documentation fixes
Fredrik Kuivinen [Sun, 29 Jan 2006 14:02:51 +0000 (15:02 +0100)]
git-branch: Documentation fixes

Signed-off-by: Fredrik Kuivinen <freku045@student.liu.se>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agomerge-recursive: Improve the error message printed when merge(1) isn't found.
Fredrik Kuivinen [Sun, 29 Jan 2006 12:16:08 +0000 (13:16 +0100)]
merge-recursive: Improve the error message printed when merge(1) isn't found.

Signed-off-by: Fredrik Kuivinen <freku045@student.liu.se>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years ago[PATCH] pre-commit sample hook: do not barf on the initial import
Junio C Hamano [Sun, 29 Jan 2006 22:52:16 +0000 (14:52 -0800)]
[PATCH] pre-commit sample hook: do not barf on the initial import

The example hook barfs on the initial import.  Ideally it should
produce a diff from an empty tree, but for now let's stop at
squelching the bogus error message.  Often an initial import
involves tons of badly formatted files from foreign SCM, so not
complaining about them like this patch does might actually be a
better idea than enforcing the "Perfect Patch" format on them.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoDocumentation: diff -c/--cc
Junio C Hamano [Sat, 28 Jan 2006 10:19:22 +0000 (02:19 -0800)]
Documentation: diff -c/--cc

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agodiff-files -c/--cc: combine only when both ours and theirs exist.
Junio C Hamano [Sat, 28 Jan 2006 10:09:08 +0000 (02:09 -0800)]
diff-files -c/--cc: combine only when both ours and theirs exist.

The previous round forgot to make sure there actually are two
versions to compare against the working tree version.  Otherwise
using -c/--cc would not make much sense.

Also plug a small memory leak.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoMerge lt/revlist,jc/diff,jc/revparse,jc/abbrev
Junio C Hamano [Sat, 28 Jan 2006 08:16:09 +0000 (00:16 -0800)]
Merge lt/revlist,jc/diff,jc/revparse,jc/abbrev

18 years agorev-parse: make "whatchanged -- git-fetch-script" work again.
Junio C Hamano [Fri, 27 Jan 2006 01:24:52 +0000 (17:24 -0800)]
rev-parse: make "whatchanged -- git-fetch-script" work again.

The latest update to avoid misspelled revs interfered when we
were not interested in parsing non flags or arguments not meant
for rev-list.  This makes these two forms work again:

git whatchanged -- git-fetch-script

We could enable "!def" in the part this change touches to make
the above work without '--', but then it would cause misspelled
v2.6.14..v2.6.16 to be given to diff-tree and defeats the whole
point of the previous fix.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agodiff --abbrev=<n> option fix.
Junio C Hamano [Fri, 27 Jan 2006 10:19:51 +0000 (02:19 -0800)]
diff --abbrev=<n> option fix.

Earier specifying an abbreviation shorter than minimum fell back
to full 40 letters, which was nonsense.  Make it to fall back to
the minimum number (currently 4).

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agopretty_print_commit: honor grafts.
Junio C Hamano [Fri, 27 Jan 2006 10:17:19 +0000 (02:17 -0800)]
pretty_print_commit: honor grafts.

When displaying Merge: lines, we used to take the real commit
parents from the commit objects.  Use the parsed parents from
the commit object instead, so that we honor fake parent
information from info/grafts.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agopretty_print_commit(): pass commit object instead of commit->buffer.
Junio C Hamano [Fri, 27 Jan 2006 09:54:59 +0000 (01:54 -0800)]
pretty_print_commit(): pass commit object instead of commit->buffer.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoRename rev-parse --abbrev to --short.
Junio C Hamano [Fri, 27 Jan 2006 01:02:07 +0000 (17:02 -0800)]
Rename rev-parse --abbrev to --short.

The usage of rev-parse to serve as a flag/option parser
for git-whatchanged and other commands have serious limitation
that the flags cannot be something that is supported by
rev-parse itself, and it cannot worked around easily.  Since
this is rarely used "poor-man's describe", rename the option for
now as an easier workaround.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agorev-parse --abbrev: do not try abbrev shorter than minimum.
Junio C Hamano [Thu, 26 Jan 2006 08:48:19 +0000 (00:48 -0800)]
rev-parse --abbrev: do not try abbrev shorter than minimum.

We do not allow abbreviation shorter than 4 letters in other
parts of the system so do not attempt to generate such.

Noticed by Uwe Zeisberger.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agodiff-tree: abbreviate merge parent object names with --abbrev --pretty.
Junio C Hamano [Wed, 25 Jan 2006 10:37:40 +0000 (02:37 -0800)]
diff-tree: abbreviate merge parent object names with --abbrev --pretty.

When --abbrev is in effect, abbreviate the merge parent names
in prettyprinted output.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agorev-parse: --abbrev option.
Junio C Hamano [Wed, 25 Jan 2006 09:35:38 +0000 (01:35 -0800)]
rev-parse: --abbrev option.

The new option behaves just like --verify, but outputs an
abbreviated object name that is unique within the repository.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoabbrev cleanup: use symbolic constants
Junio C Hamano [Wed, 25 Jan 2006 09:03:18 +0000 (01:03 -0800)]
abbrev cleanup: use symbolic constants

The minimum length of abbreviated object name was hardcoded in
different places to be 4, risking inconsistencies in the future.
Also there were three different "default abbreviation
precision".  Use two C preprocessor symbols to clean up this
mess.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agorev-list --remove-empty: add minimum help and doc entry.
Junio C Hamano [Fri, 27 Jan 2006 09:39:24 +0000 (01:39 -0800)]
rev-list --remove-empty: add minimum help and doc entry.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agorev-list: stop when the file disappears
Linus Torvalds [Wed, 18 Jan 2006 22:47:30 +0000 (14:47 -0800)]
rev-list: stop when the file disappears

The one thing I've considered doing (I really should) is to add a "stop
when you don't find the file" option to "git-rev-list". This patch does
some of the work towards that: it removes the "parent" thing when the
file disappears, so a "git annotate" could do do something like

git-rev-list --remove-empty --parents HEAD -- "$filename"

and it would get a good graph that stops when the filename disappears
(it's not perfect though: it won't remove all the unintersting commits).

It also simplifies the logic of finding tree differences a bit, at the
cost of making it a tad less efficient.

The old logic was two-phase: it would first simplify _only_ merges tree as
it traversed the tree, and then simplify the linear parts of the remainder
independently. That was pretty optimal from an efficiency standpoint
because it avoids doing any comparisons that we can see are unnecessary,
but it made it much harder to understand than it really needed to be.

The new logic is a lot more straightforward, and compares the trees as it
traverses the graph (ie everything is a single phase). That makes it much
easier to stop graph traversal at any point where a file disappears.

As an example, let's say that you have a git repository that has had a
file called "A" some time in the past. That file gets renamed to B, and
then gets renamed back again to A. The old "git-rev-list" would show two
commits: the commit that renames B to A (because it changes A) _and_ as
its parent the commit that renames A to B (because it changes A).

With the new --remove-empty flag, git-rev-list will show just the commit
that renames B to A as the "root" commit, and stop traversal there
(because that's what you want for "annotate" - you want to stop there, and
for every "root" commit you then separately see if it really is a new
file, or if the paths history disappeared because it was renamed from some
other file).

With this patch, you should be able to basically do a "poor mans 'git
annotate'" with a fairly simple loop:

push("HEAD", "$filename")
while (revision,filename = pop()) {
for each i in $(git-rev-list --parents --remove-empty $revision -- "$filename")

pseudo-parents($i) = git-rev-list parents for that line

if (pseudo-parents($i) is non-empty) {
show diff of $i against pseudo-parents
continue
}

/* See if the _real_ parents of $i had a rename */
parent($i) = real-parent($i)
if (find-rename in $parent($i)->$i)
push $parent($i), "old-name"
}

which should be doable in perl or something (doing stacks in shell is just
too painful to be worth it, so I'm not going to do this).

Anybody want to try?

Linus

18 years agodiff-files: -c and --cc options.
Junio C Hamano [Sat, 28 Jan 2006 08:03:38 +0000 (00:03 -0800)]
diff-files: -c and --cc options.

This ports the "combined diff" to diff-files so that differences
to the working tree files since stage 2 and stage 3 are shown
the same way as combined diff output from diff-tree for the
merge commit would be shown if the current working tree files
are committed.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agocombine-diff: better hunk splitting.
Junio C Hamano [Thu, 26 Jan 2006 11:53:01 +0000 (03:53 -0800)]
combine-diff: better hunk splitting.

It considered an otherwise unchanged line that had line removals
in front of it an interesting line, which caused hunks to have
one extra the trailing context line.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agodiff-tree --cc: squelch header generation on empty patch.
Junio C Hamano [Wed, 25 Jan 2006 22:26:22 +0000 (14:26 -0800)]
diff-tree --cc: squelch header generation on empty patch.

Earlier round showed the commit log header and "diff --combined"
header even for paths that had no interesting hunk under --cc
flag.  Move the header display logic around to squelch them.
With this, a merge that does not have any interesting merges
will not be shown with --cc option, unless -m is used at the
same time.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agocombine-diff: extend --cc logic to Octopus.
Junio C Hamano [Wed, 25 Jan 2006 21:11:38 +0000 (13:11 -0800)]
combine-diff: extend --cc logic to Octopus.

Santi Bejar points out that a hunk that changes from all the
same common parents except one is uninteresting.  The earlier
round marked changes from only one parent uninteresting, but
this also marks hunks that have the same change from all but one
parent uninteresting, which is a natural extension of the
original idea to Octopus merges.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agocombine-diff: minor output changes.
Junio C Hamano [Wed, 25 Jan 2006 19:55:40 +0000 (11:55 -0800)]
combine-diff: minor output changes.

Remove extra whitespace between the change indicators and the
body text.  That is more in line with the uncombined unified
diff output (pointed out by Santi Bejar).

When showing --cc, say so instead of saying just --combined.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agocombine-diff: fix appending at the tail of a list.
Junio C Hamano [Wed, 25 Jan 2006 11:34:10 +0000 (03:34 -0800)]
combine-diff: fix appending at the tail of a list.

... and use the established pattern of tail initialized to point
at the head pointer for an empty list, and updated to point at
the next pointer field of the item at the tail when appending.

Signed-off-by: Junio C Hamano <junkio@cox.net>