Code

git.git
13 years agoapply: split quoted filename handling into new function
Jonathan Nieder [Thu, 19 Aug 2010 01:46:46 +0000 (20:46 -0500)]
apply: split quoted filename handling into new function

The new find_name_gnu() function handles new-style '--- "a/foo"'
patch header lines, leaving find_name() itself a bit less
daunting.

Functional change: do not clobber the p-value when there are not
enough path components in a quoted file name to honor it.

Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agot7403: add missing &&'s
Jens Lehmann [Wed, 18 Aug 2010 21:20:33 +0000 (23:20 +0200)]
t7403: add missing &&'s

Signed-off-by: Jens Lehmann <Jens.Lehmann@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agoTell ignore file about generate files in /gitweb/static
Mark Rada [Sat, 21 Aug 2010 16:18:37 +0000 (12:18 -0400)]
Tell ignore file about generate files in /gitweb/static

Signed-off-by: Mark Rada <marada@uwaterloo.ca>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agodocs: fix Makefile dependency for user manual
Jeff King [Sat, 21 Aug 2010 06:21:22 +0000 (02:21 -0400)]
docs: fix Makefile dependency for user manual

We use our custom xsl file to build the user manual, so make
sure we depend on it. We don't use it anywhere else, so we
can stick it straight in the rule.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agoDocumentation: remove stray backslash in show-branch discussion
Jonathan Nieder [Fri, 20 Aug 2010 10:37:06 +0000 (05:37 -0500)]
Documentation: remove stray backslash in show-branch discussion

Because there is no unescaped apostrophe to pair it with, asciidoc
does not consider this apostrophe a candidate for escaping and
the backslash passes through.

Reported-by: Frédéric Brière <fbriere@fbriere.net>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agoDocumentation: remove stray backslashes from "Fighting regressions" article
Jonathan Nieder [Fri, 20 Aug 2010 10:40:31 +0000 (05:40 -0500)]
Documentation: remove stray backslashes from "Fighting regressions" article

The intended text is "it's O(N * T) vs O(N * T * M)".  Asciidoc
notices the spaces around the asterisks so there is no need to
escape them (and if you try, it passes the backslashes through).

Cc: Christian Couder <chriscool@tuxfamily.org>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agoDocumentation: do not convert ... operator to ellipses
Jonathan Nieder [Fri, 20 Aug 2010 10:39:48 +0000 (05:39 -0500)]
Documentation: do not convert ... operator to ellipses

The symmetric difference or merge-base operator ... as used by
rev-list and diff is actually three period characters.  If it
gets replaced by an ellipsis glyph in the manual, that would
stop readers from copying and pasting it.

Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agoDocumentation: avoid stray backslash in user manual
Jonathan Nieder [Fri, 20 Aug 2010 10:38:14 +0000 (05:38 -0500)]
Documentation: avoid stray backslash in user manual

Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agoDocumentation: avoid stray backslashes in core tutorial
Jonathan Nieder [Fri, 20 Aug 2010 10:37:51 +0000 (05:37 -0500)]
Documentation: avoid stray backslashes in core tutorial

While at it:

 - remove some single-quotes that were being rendered as ’\n\';

 - do not escape ellipses (...) when they do not represent the
   literal three characters "...".  We may want to ensure the
   manpages render these as three ASCII periods to make the
   manual pages easier to search, but that would be a global
   output generation setting, not a context-specific thing;

Reported-by: Frédéric Brière <fbriere@fbriere.net>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agoDocumentation: remove stray backslashes in rev-parse manual
Jonathan Nieder [Fri, 20 Aug 2010 10:35:03 +0000 (05:35 -0500)]
Documentation: remove stray backslashes in rev-parse manual

Use the {asterisk} entity instead of \* or * to avoid both
stray backslashes in output and suppression of asterisks
misinterpreted as a bold-text delimiter.

Reported-by: Frédéric Brière <fbriere@fbriere.net>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agoDocumentation: remove backslash before ~ in fast-import manual
Jonathan Nieder [Fri, 20 Aug 2010 10:34:29 +0000 (05:34 -0500)]
Documentation: remove backslash before ~ in fast-import manual

Use the {tilde} entity to get a literal tilde without fuss.
With \~, asciidoc 8.5.2 (and probably earlier versions) keeps the
backslash in the output.

Reported-by: Frédéric Brière <fbriere@fbriere.net>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agoDocumentation: remove stray backslash from "git bundle" manual
Jonathan Nieder [Fri, 20 Aug 2010 10:33:48 +0000 (05:33 -0500)]
Documentation: remove stray backslash from "git bundle" manual

In v1.6.2.2~6^2~4 (Documentation: minor grammatical fixes
and rewording in git-bundle.txt, 2009-03-22), backslashes were
introduced before ~ to avoid introducing unintentional
superscripts.  In one paragraph there is only one ~, though,
making that not a candidate for quoting, and asciidoc 8.5.8
passes the backslash through so the man page says "\~10..master".

Maybe there is an asciidoc behavior change involved.

In any case, we should replace tildes with a {tilde} entity which
means the same thing regardless of where it is found.

Reported-by: Frédéric Brière <fbriere@fbriere.net>
Cc: David J. Mellor <dmellor@whistlingcat.com>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agoDocumentation/technical: avoid stray backslash in parse-options API docs
Jonathan Nieder [Fri, 20 Aug 2010 10:32:36 +0000 (05:32 -0500)]
Documentation/technical: avoid stray backslash in parse-options API docs

Due to some unpleasant interaction between the `quote', 'italics',
and `monospace` rules, a certain paragraph ends up rendered like so:

‘short` is a character for the short option
  (e.g. <tt>'e\’</tt> for <tt>-e</tt>, use <tt>0</tt> to omit),

Use the {apostrophe} to avoid this.

While at it, escape "->" strings: they are meant as a literal
two-character C operator, not a right-pointing arrow.

Reported-by: Frédéric Brière <fbriere@fbriere.net>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agoDocumentation: remove backslashes in manpage synopses
Jonathan Nieder [Fri, 20 Aug 2010 10:31:58 +0000 (05:31 -0500)]
Documentation: remove backslashes in manpage synopses

For some reason, various manual pages have an asterisk escaped
with \ in the synopsis.  Since there is no other asterisk to pair it
with, Asciidoc does not consider this asterisk escapable, so it passes
the backslash through.

Each page either uses [verse] or has only one asterisk, so it
is safe to drop the backslashes (checked with asciidoc 8.5.2).

Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agoDocumentation: clarify quoting in gitignore docs
Jonathan Nieder [Fri, 20 Aug 2010 10:30:56 +0000 (05:30 -0500)]
Documentation: clarify quoting in gitignore docs

An asterisk in "Documentation/*.txt" quoted with \ to avoid bold text
is being output as \* because asciidoc does not consider it a
candidate for escaping (there is no matching * to pair it with).

So the manual looks like it is saying that one should write
"Documentation/\*.txt" in the .gitignore file.

Reported-by: Frédéric Brière <fbriere@fbriere.net>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agoDocumentation: clarify quoting in "git rm" example
Jonathan Nieder [Fri, 20 Aug 2010 10:30:24 +0000 (05:30 -0500)]
Documentation: clarify quoting in "git rm" example

Intended output:

git rm Documentation/\*.txt
Removes all *.txt files from the index that are under
the Documentation directory and any of its
subdirectories.

Note that the asterisk * is quoted from the shell in
this example; this lets git, and not the shell, expand
the pathnames of files and subdirectories under the
Documentation/ directory.

Without this change, there are too many backslashes output.
Tested with asciidoc 8.5.2.

Reported-by: Frédéric Brière <fbriere@fbriere.net>
Cc: Carl Worth <cworth@cworth.org>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agoDocumentation: add missing quotes to "git grep" examples
Jonathan Nieder [Fri, 20 Aug 2010 10:28:53 +0000 (05:28 -0500)]
Documentation: add missing quotes to "git grep" examples

Without an indication to the contrary, Asciidoc puts 'quoted
text' in italics, making the output look like this:

git grep time_t -- *.[ch]
    Looks for time_t in all tracked .c and .h
    files in the working directory and its subdirectories.

git grep -e '#define\' --and \( -e MAX_PATH -e PATH_MAX \)
    Looks for a line that has #define and either MAX_PATH or
    PATH_MAX.

In the first example, the *.[ch] argument needs to be protected from
the shell, or else it will only match files in the current directory.
The second example has a stray backslash.

Reported-by: Frédéric Brière <fbriere@fbriere.net>
Cc: Mark Lodato <lodatom@gmail.com>
Cc: Junio C Hamano <gitster@pobox.com>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agoDocumentation: clarify quoting in "git add" example
Jonathan Nieder [Fri, 20 Aug 2010 10:26:13 +0000 (05:26 -0500)]
Documentation: clarify quoting in "git add" example

The intended text looks like this:

· Adds content from all *.txt files under Documentation
  directory and its subdirectories:

$ git add Documentation/\*.txt

  Note that the asterisk * is quoted from the shell in this
  example; this lets the command include the files from
  subdirectories of Documentation/ directory.

The current asciidoc 8.5.2 output has a backslash before _every_
asterisk, which is more confusing than it needs to be.

Reported-by: Frédéric Brière <fbriere@fbriere.net>
Cc: Junio C Hamano <gitster@pobox.com>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agoDocumentation: unbreak regex in show-ref manual
Jonathan Nieder [Fri, 20 Aug 2010 10:23:54 +0000 (05:23 -0500)]
Documentation: unbreak regex in show-ref manual

I am not sure why, but the regular expression "(?:\^\{\})" gets
rendered by asciidoc as "(?:\{})".  The intent seems to be a regex
matching the literal string "^{}", so this rewrites the markup to
produce "(?:\^{})" as output.

Cc: Julian Phillips <julian@quantumfyre.co.uk>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agoDocumentation: quoting trouble in "git rm" discussion
Jonathan Nieder [Fri, 20 Aug 2010 10:21:53 +0000 (05:21 -0500)]
Documentation: quoting trouble in "git rm" discussion

The current output (with Asciidoc 8.5.2) seems a bit broken:

given two directories ‘d` and d2, there is a difference
between using git rm 'd*’ and ‘git rm 'd/\*\’`, as the
former will also remove all of directory d2.

In other words, the markup parses as

given two directories << d` and _d2_, there is a difference
between using _git rm 'd* >>_ and << git rm 'd/\*\ >> `.

I suspect there is an asciidoc bug involved (why is ' a candidate
closing-quote mark when it is preceded by a backslash?) but with
all the meanings of ` and ' involved I do not want to track it
down.  Better to use unambiguous {asterisk} and {apostrophe}
entities.

Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agoDocumentation: tweak description of log.date
Jonathan Nieder [Fri, 20 Aug 2010 10:20:36 +0000 (05:20 -0500)]
Documentation: tweak description of log.date

The markup "'git log'\'s" produces a stray backslash in the
produced man page.  Removing the backslash fixes it.

While at it, tweak the surrounding description for readability.

Reported-by: Frédéric Brière <fbriere@fbriere.net>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agomergetool: Remove explicit references to /dev/tty
Charles Bailey [Fri, 20 Aug 2010 15:25:09 +0000 (16:25 +0100)]
mergetool: Remove explicit references to /dev/tty

mergetool used /dev/tty to switch back to receiving input from the user
via inside a block with a redirected stdin.

This harms testability, so change mergetool to save its original stdin
to an alternative fd in this block and restore it for those sub-commands
that need the original stdin.

Includes additional compatibility fix from Jonathan Nieder.

Tested-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Charles Bailey <charles@hashpling.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agoMerge branch 'maint'
Junio C Hamano [Fri, 20 Aug 2010 19:55:41 +0000 (12:55 -0700)]
Merge branch 'maint'

* maint:
  xmalloc: include size in the failure message

13 years agoMerge branch 'jc/maint-follow-rename-fix' into maint
Junio C Hamano [Fri, 20 Aug 2010 19:53:09 +0000 (12:53 -0700)]
Merge branch 'jc/maint-follow-rename-fix' into maint

* jc/maint-follow-rename-fix:
  log: test for regression introduced in v1.7.2-rc0~103^2~2
  diff --follow: do call diffcore_std() as necessary
  diff --follow: do not waste cycles while recursing

13 years agoMerge branch 'jn/maint-plug-leak' into maint
Junio C Hamano [Fri, 20 Aug 2010 19:53:09 +0000 (12:53 -0700)]
Merge branch 'jn/maint-plug-leak' into maint

* jn/maint-plug-leak:
  write-tree: Avoid leak when index refers to an invalid object
  read-tree: stop leaking tree objects
  core: Stop leaking ondisk_cache_entrys

13 years agoMerge branch 'jn/fix-abbrev' into maint
Junio C Hamano [Fri, 20 Aug 2010 19:53:09 +0000 (12:53 -0700)]
Merge branch 'jn/fix-abbrev' into maint

* jn/fix-abbrev:
  examples/commit: use --abbrev for commit summary
  checkout, commit: remove confusing assignments to rev.abbrev
  archive: abbreviate substituted commit ids again

13 years agoMerge branch 'vs/doc-spell' into maint
Junio C Hamano [Fri, 20 Aug 2010 19:53:08 +0000 (12:53 -0700)]
Merge branch 'vs/doc-spell' into maint

* vs/doc-spell:
  Documentation: spelling fixes

13 years agoMerge branch 'jn/rebase-rename-am' into maint
Junio C Hamano [Fri, 20 Aug 2010 19:53:08 +0000 (12:53 -0700)]
Merge branch 'jn/rebase-rename-am' into maint

* jn/rebase-rename-am:
  rebase: protect against diff.renames configuration
  t3400 (rebase): whitespace cleanup
  Teach "apply --index-info" to handle rename patches
  t4150 (am): futureproof against failing tests
  t4150 (am): style fix

13 years agoMerge branch 'jn/doc-pull' into maint
Junio C Hamano [Fri, 20 Aug 2010 19:53:08 +0000 (12:53 -0700)]
Merge branch 'jn/doc-pull' into maint

* jn/doc-pull:
  Documentation: flesh out “git pull” description

13 years agoMerge branch 'bc/use-more-hardlinks-in-install' into maint
Junio C Hamano [Fri, 20 Aug 2010 19:53:07 +0000 (12:53 -0700)]
Merge branch 'bc/use-more-hardlinks-in-install' into maint

* bc/use-more-hardlinks-in-install:
  Makefile: make hard/symbolic links for non-builtins too
  Makefile: link builtins residing in bin directory to main git binary too

13 years agoMerge branch 'tr/rfc-reset-doc' into maint
Junio C Hamano [Fri, 20 Aug 2010 19:53:07 +0000 (12:53 -0700)]
Merge branch 'tr/rfc-reset-doc' into maint

* tr/rfc-reset-doc:
  Documentation/reset: move "undo permanently" example behind "make topic"
  Documentation/reset: reorder examples to match description
  Documentation/reset: promote 'examples' one section up
  Documentation/reset: separate options by mode
  Documentation/git-reset: reorder modes for soft-mixed-hard progression

13 years agobuiltin/checkout: Fix message when switching to an existing branch
Ramkumar Ramachandra [Fri, 20 Aug 2010 17:41:47 +0000 (23:11 +0530)]
builtin/checkout: Fix message when switching to an existing branch

Fix "Switched to a new branch <name>" to read "Switched to branch
<name>" when <name> corresponds to an existing branch. This bug was
introduced in 02ac983 while introducing the `-B` switch.

Cc: Tay Ray Chuan <rctay89@gmail.com>
Signed-off-by: Ramkumar Ramachandra <artagnon@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agoDocumentation: set a !DOCTYPE for user manual
Jonathan Nieder [Fri, 20 Aug 2010 05:04:02 +0000 (00:04 -0500)]
Documentation: set a !DOCTYPE for user manual

asciidoc already takes care of including a doctype for most of the
HTML documentation, but the user manual which is processed with
docbook-xsl directly lacks one (at least with Debian docbook-xsl
1.75.2+dfsg-5).  This makes it harder to automatically validate the
HTML.

Reported-by: 積丹尼 <jidanni@jidanni.org>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agoxmalloc: include size in the failure message
Matthieu Moy [Fri, 20 Aug 2010 15:09:51 +0000 (17:09 +0200)]
xmalloc: include size in the failure message

Out-of-memory errors can either be actual lack of memory, or bugs (like
code trying to call xmalloc(-1) by mistake). A little more information
may help tracking bugs reported by users.

Signed-off-by: Matthieu Moy <Matthieu.Moy@imag.fr>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agomerge-recursive: Avoid excessive output for and reprocessing of renames
Elijah Newren [Fri, 20 Aug 2010 12:55:40 +0000 (06:55 -0600)]
merge-recursive: Avoid excessive output for and reprocessing of renames

In 5a2580d (merge_recursive: Fix renames across paths below D/F conflicts
2010-07-09) and ae74548 (merge-recursive: Fix multiple file rename across
D/F conflict 2010-08-17), renames across D/F conflicts were fixed by
making process_renames() consider as unprocessed renames whose dst_entry
"still" had higher stage entries.  The assumption was that those higher
stage entries would have been cleared out of dst_entry by that point in
cases where the conflict could be resolved (normal renames with no D/F
conflicts).  That is not the case -- higher stage entries will remain in
all cases.

Fix this by checking for higher stage entries corresponding to D/F
conflicts, namely that stages 2 and 3 have exactly one nonzero mode between
them.  The nonzero mode stage corresponds to a file at the path, while the
stage with a zero mode will correspond to a directory at that path (since
rename/delete conflicts will have already been handled before this codepath
is reached.)

Signed-off-by: Elijah Newren <newren@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agoMerge branch 'maint' to sync with 1.7.2.2
Junio C Hamano [Thu, 19 Aug 2010 23:04:41 +0000 (16:04 -0700)]
Merge branch 'maint' to sync with 1.7.2.2

Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agoGit 1.7.2.2 v1.7.2.2
Junio C Hamano [Thu, 19 Aug 2010 23:03:18 +0000 (16:03 -0700)]
Git 1.7.2.2

Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agoMerge branch 'tr/xsize-bits' into maint
Junio C Hamano [Thu, 19 Aug 2010 22:49:54 +0000 (15:49 -0700)]
Merge branch 'tr/xsize-bits' into maint

* tr/xsize-bits:
  xsize_t: check whether we lose bits

13 years agoMerge branch 'jc/sha1-name-find-fix' into maint
Junio C Hamano [Thu, 19 Aug 2010 22:49:24 +0000 (15:49 -0700)]
Merge branch 'jc/sha1-name-find-fix' into maint

* jc/sha1-name-find-fix:
  sha1_name.c: fix parsing of ":/token" syntax

Conflicts:
sha1_name.c

13 years agoautoconf: don't use platform regex if it lacks REG_STARTEND
Jonathan Nieder [Tue, 17 Aug 2010 09:24:42 +0000 (09:24 +0000)]
autoconf: don't use platform regex if it lacks REG_STARTEND

If the platform regex cannot match null bytes, we might as well
use the glibc version instead.

Cc: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Cc: René Scharfe <rene.scharfe@lsrfire.ath.cx>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Tested-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agot/t7008-grep-binary.sh: un-TODO a test that needs REG_STARTEND
Ævar Arnfjörð Bjarmason [Tue, 17 Aug 2010 09:24:41 +0000 (09:24 +0000)]
t/t7008-grep-binary.sh: un-TODO a test that needs REG_STARTEND

Now that we have a regex engine that supports REG_STARTEND this test
should fail if "git grep" can't grep NULL characters.

Platforms that don't have a POSIX regex engine which supports
REG_STARTEND should always define NO_REGEX=YesPlease when compiling.

Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Acked-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agocompat/regex: get rid of old-style definition
Junio C Hamano [Thu, 19 Aug 2010 22:31:30 +0000 (15:31 -0700)]
compat/regex: get rid of old-style definition

These files mostly used ANSI style function definitions, but with small
number of old-style ones.  Convert them to consistently use ANSI style.

Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agot/t9010-svn-fe.sh: add an +x bit to this test
Ævar Arnfjörð Bjarmason [Thu, 19 Aug 2010 15:53:50 +0000 (15:53 +0000)]
t/t9010-svn-fe.sh: add an +x bit to this test

Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agocompat/regex: define out variables only used under RE_ENABLE_I18N
Ævar Arnfjörð Bjarmason [Thu, 19 Aug 2010 18:30:01 +0000 (18:30 +0000)]
compat/regex: define out variables only used under RE_ENABLE_I18N

Wrap variables that were only used RE_ENABLE_I18N in `#ifdef
RE_ENABLE_I18N`. This eliminates compiler warnings when compiling with
NO_REGEX=YesPlease.

Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agot9155: fix compatibility with older SVN
Eric Wong [Thu, 19 Aug 2010 19:03:18 +0000 (12:03 -0700)]
t9155: fix compatibility with older SVN

The "--parents" option did not appear until SVN 1.5.x
and is completely unnecessary in this case.

Reported-by: Jens Lehmann <Jens.Lehmann@web.de>
Signed-off-by: Eric Wong <normalperson@yhbt.net>
13 years agoMerge branch 'sp/fix-smart-http-deadlock-on-error' into maint
Junio C Hamano [Wed, 18 Aug 2010 23:30:11 +0000 (16:30 -0700)]
Merge branch 'sp/fix-smart-http-deadlock-on-error' into maint

* sp/fix-smart-http-deadlock-on-error:
  smart-http: Don't deadlock on server failure

13 years agogit-gui: ensure correct application termination in git-gui--askpass
Pat Thoyts [Wed, 18 Aug 2010 22:19:24 +0000 (23:19 +0100)]
git-gui: ensure correct application termination in git-gui--askpass

With Tk 8.5 the askpass utility can hang waiting for the wish shell
implicit event loop to exit. This patch uses an explicit event loop
to ensure correct application termination.

Reported-by: Anders Kaseorg <andersk@mit.edu>
Tested-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
13 years agoChange regerror() declaration from K&R style to ANSI C (C89)
Frank Li [Tue, 17 Aug 2010 09:24:40 +0000 (09:24 +0000)]
Change regerror() declaration from K&R style to ANSI C (C89)

The MSVC headers typedef errcode as int, and thus confused the compiler in
the K&R style definition. ANSI style deconfuses it.

This patch was originally applied as v1.6.5-rc2~23 but needs to be
re-applied since compat/regex was overwritten by Ævar Arnfjörð
Bjarmason with the gawk regex engine.

Signed-off-by: Frank Li <lznuaa@gmail.com>
Signed-off-by: Marius Storm-Olsen <mstormo@gmail.com>
Acked-by: Johannes Sixt <j6t@kdbg.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agocompat/regex: get the gawk regex engine to compile within git
Ævar Arnfjörð Bjarmason [Tue, 17 Aug 2010 09:24:39 +0000 (09:24 +0000)]
compat/regex: get the gawk regex engine to compile within git

We need to define -DGAWK -DNO_MBSUPPORT so that the gawk regex engine
will compile, and include stdio.h and stddef.h in regex.h. Gawk itself
includes these headers before it includes the regex.h header.

Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Acked-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agocompat/regex: use the regex engine from gawk for compat
Ævar Arnfjörð Bjarmason [Tue, 17 Aug 2010 09:24:38 +0000 (09:24 +0000)]
compat/regex: use the regex engine from gawk for compat

Change the regex engine in compat to use the gawk engine from the
gawk-devel module in gawk CVS. This engine supports the REG_STARTEND
flag, which was optionally available in Git since v1.7.2-rc0~77^2~1.

The source was grabbed from cvs.savannah.gnu.org:/sources/gawk, and
these are the upstream versions of the files being included:

    regcomp.c               1.4
    regex.h                 1.3
    regex.h                 1.3
    regex_internal.c        1.3
    regex_internal.h        1.3
    regexec.c               1.3

Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Acked-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agomerge script: learn --[no-]rerere-autoupdate
Jonathan Nieder [Tue, 17 Aug 2010 07:13:40 +0000 (02:13 -0500)]
merge script: learn --[no-]rerere-autoupdate

Port v1.7.0-rc0~83^2 (Teach --[no-]rerere-autoupdate option to
merge, revert and friends, 2009-12-04) to the example merge script.

After this change, all tests pass for me with the scripted
merge.

Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agomerge script: notice @{-1} shorthand
Jonathan Nieder [Tue, 17 Aug 2010 07:11:30 +0000 (02:11 -0500)]
merge script: notice @{-1} shorthand

Port v1.6.2-rc1~10^2 (Teach @{-1} to git merge, 2009-02-13) to
the old merge script.

Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agomerge script: handle --no-ff --no-commit correctly
Jonathan Nieder [Tue, 17 Aug 2010 07:11:03 +0000 (02:11 -0500)]
merge script: handle --no-ff --no-commit correctly

In a --no-ff merge with conflicts, "git commit" used to forget the
--no-ff when used to complete the merge.  That was fixed by
v1.6.1-rc1~134^2 (builtin-commit: use reduce_heads() only when
appropriate, 2008-10-03) for the builtin merge.  Port the change to
the merge script in contrib/examples.

Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agomerge script: --ff-only to disallow true merge
Jonathan Nieder [Tue, 17 Aug 2010 07:10:17 +0000 (02:10 -0500)]
merge script: --ff-only to disallow true merge

Port v1.6.6-rc0~62^2 (Teach 'git merge' and 'git pull' the option
--ff-only, 2009-10-29) to the old merge script.

Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agomerge script: handle many-way octopus
Jonathan Nieder [Tue, 17 Aug 2010 07:09:58 +0000 (02:09 -0500)]
merge script: handle many-way octopus

Based on v1.6.0-rc0~51^2~5 (Build in merge, 2008-07-07).

Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agomerge script: handle -m --log correctly
Jonathan Nieder [Tue, 17 Aug 2010 07:06:56 +0000 (02:06 -0500)]
merge script: handle -m --log correctly

Based on v1.7.1.1~23^2 (merge: --log appends shortlog to message if
specified, 2010-05-11).  Without this change, the scripted
(non-builtin) merge does not pass t7604.

Cc: Tay Ray Chuan <rctay89@gmail.com>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agomerge script: forbid merge -s index
Jonathan Nieder [Tue, 17 Aug 2010 07:06:06 +0000 (02:06 -0500)]
merge script: forbid merge -s index

Some git-merge-* commands are not merge strategies.  This is based on
v1.6.1-rc1~294^2~7 (builtin-merge: allow using a custom strategy,
2008-07-30) but it is less smart: we just use a hard-coded list of
forbidden strategy names.  It is okay if this falls out of date, since
the code is just an example.

Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agomerge script: allow custom strategies
Jonathan Nieder [Tue, 17 Aug 2010 07:05:32 +0000 (02:05 -0500)]
merge script: allow custom strategies

The idea comes from v1.6.1-rc1~294^2~7 (builtin-merge: allow using a
custom strategy, 2008-07-30).

Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agomerge script: merge -X<option>
Jonathan Nieder [Tue, 17 Aug 2010 07:05:10 +0000 (02:05 -0500)]
merge script: merge -X<option>

Without this support, the scripted merge cannot pass t6037.

Based on v1.7.0-rc0~55^2~5 (git merge -X<option>, 2009-11-25).

Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agomerge script: improve log message subject
Jonathan Nieder [Tue, 17 Aug 2010 07:04:43 +0000 (02:04 -0500)]
merge script: improve log message subject

 - point out remote-tracking branches as "remote branch
   'upstream/master'";
 - avoid misleading log messages when a tag and branch
   share a name.

This approximates the builtin merge command's behavior well
enough to pass the relevant tests.

Based roughly on v1.6.4.2~10^2 (merge: indicate remote tracking
branches in merge message, 2009-08-09) and v1.6.4.2~10^2~1 (merge: fix
incorrect merge message for ambiguous tag/branch, 2009-08-09).

Cc: Jeff King <peff@peff.net>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agomerge script: refuse to merge during merge
Jonathan Nieder [Tue, 17 Aug 2010 07:03:58 +0000 (02:03 -0500)]
merge script: refuse to merge during merge

Check MERGE_HEAD and bail out if it exists.  Based on v1.6.3.3~3^2
(refuse to merge during a merge, 2009-06-01).  Without this change,
the scripted merge does not pass t3030.

Cc: Clemens Buchacher <drizzd@aon.at>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agomerge script: tweak unmerged files message to match builtin
Jonathan Nieder [Tue, 17 Aug 2010 07:03:36 +0000 (02:03 -0500)]
merge script: tweak unmerged files message to match builtin

Before:

You are in the middle of a conflicted merge.

After:

Merge is not possible because you have unmerged files.

I prefer the old message, but the new one is more consistent with
other commands and tests expect it.  In particular, without this
change the scripted merge does not pass t3030.

Based on v1.7.0-rc0~66^2 (Be more user-friendly when refusing to do
something because of conflict., 2010-01-12).

Cc: Matthieu Moy <Matthieu.Moy@imag.fr>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agomerge script: --squash, --ff from unborn branch are errors
Jonathan Nieder [Tue, 17 Aug 2010 07:03:07 +0000 (02:03 -0500)]
merge script: --squash, --ff from unborn branch are errors

Port v1.6.1-rc1~319 (provide more errors for the "merge into empty
head" case, 2008-08-21) to the example merge script.

Noticed by comparison with builtin merge.

Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agofmt-merge-msg -m to override merge title
Jonathan Nieder [Tue, 17 Aug 2010 23:00:34 +0000 (18:00 -0500)]
fmt-merge-msg -m to override merge title

Since v1.7.1.1~23^2 (merge: --log appends shortlog to message if
specified, 2010-05-11), the fmt-merge-msg backend supports custom text
to override the merge title "Merge <foo> into <bar>".

Expose this functionality for scripted callers.  Example:

 git fmt-merge-msg --log -m \
"$(printf '%s\n' \
    "Merge branch 'api-cleanup' into feature" \
    '' \
    'This is to use a few functions refactored for this purpose.'
)" <.git/FETCH_HEAD

Cc: Tay Ray Chuan <rctay89@gmail.com>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agomerge-base --independent to print reduced parent list in a merge
Jonathan Nieder [Tue, 17 Aug 2010 07:01:54 +0000 (02:01 -0500)]
merge-base --independent to print reduced parent list in a merge

While show-branch --independent does not support more than MAX_REVS
revs, git internally supports more with a different algorithm.
Expose that functionality as "git merge-base --independent".

This should help scripts to catch up with builtin merge in supporting
dodecapus.

Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agomerge-base --octopus to mimic show-branch --merge-base
Jonathan Nieder [Tue, 17 Aug 2010 07:01:15 +0000 (02:01 -0500)]
merge-base --octopus to mimic show-branch --merge-base

While show-branch --merge-base does not support more than MAX_REVS
revs, git supports more with a different algorithm
(v1.6.0-rc0~51^2~13, Introduce get_octopus_merge_bases() in commit.c,
2008-06-27).  Expose that functionality.

This should help scripts to catch up with builtin merge in supporting
dodecapus.

Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agoDocumentation: add a SEE ALSO section for merge-base
Jonathan Nieder [Tue, 17 Aug 2010 06:59:55 +0000 (01:59 -0500)]
Documentation: add a SEE ALSO section for merge-base

For example, a person reading the merge-base man page might wonder
about the fastest way to check if one commit is an ancestor of
another (which would require rev-list).

Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agot6200 (fmt-merge-msg): style nitpicks
Jonathan Nieder [Tue, 17 Aug 2010 06:57:40 +0000 (01:57 -0500)]
t6200 (fmt-merge-msg): style nitpicks

Guard setup with test_expect_success.  Use test_might_fail
instead of ignoring the exit code from git config --unset.
Point out setup commands that are shared by multiple tests,
to make it easy to write GIT_SKIP_TESTS specifications that
work.

Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agot6010 (merge-base): modernize style
Jonathan Nieder [Tue, 17 Aug 2010 06:57:12 +0000 (01:57 -0500)]
t6010 (merge-base): modernize style

Guard setup with test_expect_success, put the opening quote
starting each test on the same line as the test_expect_* invocation,
and combine related actions into single tests.

While at it:

 - use test_cmp instead of expr or test $foo = $bar, for more helpful
   output with -v when tests fail;

 - use test_commit for brevity.

Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agot7600 (merge): test merge from branch yet to be born
Jonathan Nieder [Tue, 17 Aug 2010 06:56:46 +0000 (01:56 -0500)]
t7600 (merge): test merge from branch yet to be born

Some people like to "git fetch origin && merge origin/master" from
the unborn branch provided when first initializing a repository.

Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agot7600 (merge): check reflog entry
Jonathan Nieder [Tue, 17 Aug 2010 06:56:13 +0000 (01:56 -0500)]
t7600 (merge): check reflog entry

The details of the reflog message are not important, but
including something sane in the reflog is.

Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agot7600 (merge): do not launch gitk for --debug
Jonathan Nieder [Tue, 17 Aug 2010 06:54:59 +0000 (01:54 -0500)]
t7600 (merge): do not launch gitk for --debug

Probably as a development aid, this test script runs gitk --all
to allow the driver to inspect history between tests when run
with --debug.  As a result, running all tests with --debug
requires closing a long series of gitk displays, one at a time.

Use git log --graph --oneline instead.  This way, the history is
available for viewing with "git show" but the test script finishes
without interaction.

Longer term, it would be nice to have an option to run a
user-specified command between tests.  This patch does not do
that.

Cc: Lars Hjemli <hjemli@gmail.com>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agot7600 (merge): modernize style
Jonathan Nieder [Tue, 17 Aug 2010 06:53:56 +0000 (01:53 -0500)]
t7600 (merge): modernize style

Guard setup commands with test_expect_success, so they are easier
to visually skip over and get to the good part.  While at it:

 - use "printf '%s\n' a b ..." instead of "cat <<EOF" for test
   vectors with short lines;

 - use test_cmp instead of test foo = bar where possible, for
   better output with -v on failure;

 - do not go to extraordinary lengths to print a relevant message
   when test commands fail.  There is a patch in flight that could be
   used to restore the nice error messages in a cleaner way.

Cc: Lars Hjemli <hjemli@gmail.com>
Cc: Matthieu Moy <Matthieu.Moy@imag.fr>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agomerge: do not mistake (ancestor of) tag for branch
Jonathan Nieder [Tue, 17 Aug 2010 06:52:48 +0000 (01:52 -0500)]
merge: do not mistake (ancestor of) tag for branch

If no branch 'foo' exists but a tag 'foo' does, then
git merge foo^ results in

Merge branch 'foo' (early part)

as a commit message, because the relevant code path checks that
refs/heads/foo is a valid refname for writing rather than for
reading.

Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agosubmodule sync: Update "submodule.<name>.url"
David Aguilar [Wed, 18 Aug 2010 15:58:33 +0000 (08:58 -0700)]
submodule sync: Update "submodule.<name>.url"

When "git submodule sync" synchronizes the repository URLs
it only updates submodules' .git/config.  However, the old
URLs still exist in the super-project's .git/config.

Update the super-project's configuration so that commands
such as "git submodule update" use the URLs from .gitmodules.

Signed-off-by: David Aguilar <davvid@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agotests: fix syntax error in "Use advise() for hints" test
Ævar Arnfjörð Bjarmason [Wed, 18 Aug 2010 14:36:44 +0000 (14:36 +0000)]
tests: fix syntax error in "Use advise() for hints" test

Change the test introduced in the "Use advise() for hints" patch by
Jonathan Nieder not to use '' for quotes inside '' delimited code. It
ended up introducing a file called <paths> to the main git repository.

Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agoMerge branch 'jc/maint-follow-rename-fix'
Junio C Hamano [Wed, 18 Aug 2010 19:47:18 +0000 (12:47 -0700)]
Merge branch 'jc/maint-follow-rename-fix'

* jc/maint-follow-rename-fix:
  log: test for regression introduced in v1.7.2-rc0~103^2~2
  diff --follow: do call diffcore_std() as necessary
  diff --follow: do not waste cycles while recursing

13 years agoMerge branch 'tf/string-list-init'
Junio C Hamano [Wed, 18 Aug 2010 19:47:04 +0000 (12:47 -0700)]
Merge branch 'tf/string-list-init'

* tf/string-list-init:
  string_list: Add STRING_LIST_INIT macro and make use of it.

13 years agoMerge branch 'cc/find-commit-subject'
Junio C Hamano [Wed, 18 Aug 2010 19:46:55 +0000 (12:46 -0700)]
Merge branch 'cc/find-commit-subject'

* cc/find-commit-subject:
  blame: use find_commit_subject() instead of custom code
  merge-recursive: use find_commit_subject() instead of custom code
  bisect: use find_commit_subject() instead of custom code
  revert: rename variables related to subject in get_message()
  revert: refactor code to find commit subject in find_commit_subject()
  revert: fix off by one read when searching the end of a commit subject

13 years agocvs tests: do not touch test CVS repositories shipped with source
Junio C Hamano [Mon, 16 Aug 2010 16:25:01 +0000 (09:25 -0700)]
cvs tests: do not touch test CVS repositories shipped with source

Some tests in t96xx series (cvsimport) want to write into the control area
(CVSROOT) of their test CVS repositories, but this does not work well when
the source area is made read-only (test trash directories are moved via
--root=else/where option).

Copy the supplied test CVS repository to a scratch place at the beginning
of these tests.

Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agot/t9602-cvsimport-branches-tags.sh: Add a PERL prerequisite
Ævar Arnfjörð Bjarmason [Fri, 13 Aug 2010 20:40:13 +0000 (20:40 +0000)]
t/t9602-cvsimport-branches-tags.sh: Add a PERL prerequisite

Change this test to declare a PERL prerequisite. These tests use the
-p switch, so they implicitly depend on Perl code, but nothing was
declaring this.

Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agot/t9601-cvsimport-vendor-branch.sh: Add a PERL prerequisite
Ævar Arnfjörð Bjarmason [Fri, 13 Aug 2010 20:40:12 +0000 (20:40 +0000)]
t/t9601-cvsimport-vendor-branch.sh: Add a PERL prerequisite

Change this test to declare a PERL prerequisite. These tests use the
-p switch, so they implicitly depend on Perl code, but nothing was
declaring this.

Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agot/t7105-reset-patch.sh: Add a PERL prerequisite
Ævar Arnfjörð Bjarmason [Fri, 13 Aug 2010 20:40:11 +0000 (20:40 +0000)]
t/t7105-reset-patch.sh: Add a PERL prerequisite

Change this test to declare a PERL prerequisite. These tests use the
-p switch, so they implicitly depend on Perl code, but nothing was
declaring this.

Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agot/t9001-send-email.sh: convert setup code to tests
Ævar Arnfjörð Bjarmason [Fri, 13 Aug 2010 20:40:10 +0000 (20:40 +0000)]
t/t9001-send-email.sh: convert setup code to tests

Change the setup code in t/t9001-send-email.sh to use
test_expect_success. This way it isn't needlessly run in environments
where the test prerequisites aren't met.

Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agot/t9001-send-email.sh: change from skip_all=* to prereq skip
Ævar Arnfjörð Bjarmason [Fri, 13 Aug 2010 20:40:09 +0000 (20:40 +0000)]
t/t9001-send-email.sh: change from skip_all=* to prereq skip

Change this test to skip test with test prerequisites, and to do setup
work in tests. This improves the skipped statistics on platforms where
the test isn't run.

Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agot/t9001-send-email.sh: Remove needless PROG=* assignment
Ævar Arnfjörð Bjarmason [Fri, 13 Aug 2010 20:40:08 +0000 (20:40 +0000)]
t/t9001-send-email.sh: Remove needless PROG=* assignment

Remove the PROG=* assignment from t9001-send-email.sh. It's been there
since v1.4.0-rc1~30 when the test was originally added, but only tests
that source annotate-tests.sh need it, it was seemingly introduced to
this test via copy/paste coding.

Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agot/t9600-cvsimport.sh: change from skip_all=* to prereq skip
Ævar Arnfjörð Bjarmason [Fri, 13 Aug 2010 20:40:07 +0000 (20:40 +0000)]
t/t9600-cvsimport.sh: change from skip_all=* to prereq skip

Change this test to skip test with test prerequisites, and to do setup
work in tests. This improves the skipped statistics on platforms where
the test isn't run.

Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agolib-patch-mode tests: change from skip_all=* to prereq skip
Ævar Arnfjörð Bjarmason [Fri, 13 Aug 2010 20:40:06 +0000 (20:40 +0000)]
lib-patch-mode tests: change from skip_all=* to prereq skip

Change this test to skip test with test prerequisites, and to do setup
work in tests. This improves the skipped statistics on platforms where
the test isn't run.

Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agot/t3701-add-interactive.sh: change from skip_all=* to prereq skip
Ævar Arnfjörð Bjarmason [Fri, 13 Aug 2010 20:40:05 +0000 (20:40 +0000)]
t/t3701-add-interactive.sh: change from skip_all=* to prereq skip

Change this test to skip test with test prerequisites, and to do setup
work in tests. This improves the skipped statistics on platforms where
the test isn't run.

Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agotests: Move FILEMODE prerequisite to lib-prereq-FILEMODE.sh
Ævar Arnfjörð Bjarmason [Fri, 13 Aug 2010 20:40:04 +0000 (20:40 +0000)]
tests: Move FILEMODE prerequisite to lib-prereq-FILEMODE.sh

Change the five tests that were all checking "git config --bool
core.filemode" to use a new FILEMODE prerequisite in
lib-prereq-FILEMODE.sh.

Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agot/Makefile: Create test-results dir for smoke target
Ævar Arnfjörð Bjarmason [Sat, 14 Aug 2010 11:37:05 +0000 (11:37 +0000)]
t/Makefile: Create test-results dir for smoke target

Change the smoke target to create a test-results directory. This was
done implicitly by the test-lib before my "test-lib: Don't write
test-results when HARNESS_ACTIVE" patch, but after that smoking from
the pu branch hasn't worked.

Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agogit-notes: Run partial expensive test everywhere
Ævar Arnfjörð Bjarmason [Tue, 10 Aug 2010 23:37:48 +0000 (23:37 +0000)]
git-notes: Run partial expensive test everywhere

The git-notes expensive timing test is only expensive because it
either did 10,100,1k and 10k iterations or nothing.

Change it to do 10 by default, with an option to run the expensive
version with the old GIT_NOTES_TIMING_TESTS=ZomgYesPlease variable.

Since nobody was ostensibly running this test under TAP the code had
bitrotted so that it emitted invalid TAP. This change fixes that.

The old version would also mysteriously fail on systems without
/usr/bin/time, there's now a check for that using the multiple test
prerequisite facility.

Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agot/t3300-funny-names: change from skip_all=* to prereq skip
Ævar Arnfjörð Bjarmason [Wed, 11 Aug 2010 19:04:10 +0000 (19:04 +0000)]
t/t3300-funny-names: change from skip_all=* to prereq skip

Change this test to skip test with test prerequisites, and to do setup
work in tests. This improves the skipped statistics on platforms where
the test isn't run.

Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agot/t3902-quoted: change from skip_all=* to prereq skip
Ævar Arnfjörð Bjarmason [Wed, 11 Aug 2010 19:04:09 +0000 (19:04 +0000)]
t/t3902-quoted: change from skip_all=* to prereq skip

Change this test to skip test with test prerequisites, and to do setup
work in tests. This improves the skipped statistics on platforms where
the test isn't run.

Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agot/t4016-diff-quote: change from skip_all=* to prereq skip
Ævar Arnfjörð Bjarmason [Wed, 11 Aug 2010 19:04:08 +0000 (19:04 +0000)]
t/t4016-diff-quote: change from skip_all=* to prereq skip

Change this test to skip test with test prerequisites, and to do setup
work in tests. This improves the skipped statistics on platforms where
the test isn't run.

Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agot/t5503-tagfollow: change from skip_all=* to prereq skip
Ævar Arnfjörð Bjarmason [Wed, 11 Aug 2010 19:04:07 +0000 (19:04 +0000)]
t/t5503-tagfollow: change from skip_all=* to prereq skip

Change this test to skip test with test prerequisites, and to do setup
work in tests. This improves the skipped statistics on platforms where
the test isn't run.

Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agot/t7005-editor: change from skip_all=* to prereq skip
Ævar Arnfjörð Bjarmason [Wed, 11 Aug 2010 19:04:06 +0000 (19:04 +0000)]
t/t7005-editor: change from skip_all=* to prereq skip

Change this test to skip test with test prerequisites, and to do setup
work in tests. This improves the skipped statistics on platforms where
the test isn't run.

Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agot/t5705-clone-2gb: change from skip_all=* to prereq skip
Ævar Arnfjörð Bjarmason [Wed, 11 Aug 2010 19:04:05 +0000 (19:04 +0000)]
t/t5705-clone-2gb: change from skip_all=* to prereq skip

Change this test to skip test with test prerequisites, and to do setup
work in tests. This improves the skipped statistics on platforms where
the test isn't run.

Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agot/t1304-default-acl: change from skip_all=* to prereq skip
Ævar Arnfjörð Bjarmason [Wed, 11 Aug 2010 19:04:04 +0000 (19:04 +0000)]
t/t1304-default-acl: change from skip_all=* to prereq skip

Change this test to skip test with test prerequisites, and to do setup
work in tests. This improves the skipped statistics on platforms where
the test isn't run.

Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
13 years agot/README: Update "Skipping tests" to align with best practices
Ævar Arnfjörð Bjarmason [Wed, 28 Jul 2010 10:34:59 +0000 (10:34 +0000)]
t/README: Update "Skipping tests" to align with best practices

The example I initially added to "Skipping tests" wasn't very
good. We'd rather skip tests using the three-arg prereq form to the
test_* functions, not bail out with a skip message.

Change the documentation to reflect that, but retain the bailout
example under a disclaimer which explains that it's probably not a
good idea to use it.

Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>