Commit Graph

36496 Commits

Author SHA1 Message Date
e7880fcd41 t4209: factor out helper function test_log_icase()
Reduce code duplication by introducing test_log_icase() that runs the
same test with both --regexp-ignore-case and -i.  The specification of
the four basic test scenarios (matching/nomatching combined with case
sensitive/insensitive) becomes easier to read and write.

Signed-off-by: Rene Scharfe <l.s.r@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-24 15:12:42 -07:00
57b6dc76f2 t4209: factor out helper function test_log()
Twelve tests in t4209 follow the same simple pattern for description,
git log call and checking.  Extract that shared logic into a helper
function named test_log.  Test specifications become a lot more
compact, new tests can be added more easily.

Signed-off-by: Rene Scharfe <l.s.r@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-24 15:11:44 -07:00
9fe0cf3a5e branch.c: install_branch_config: simplify if chain
Simplify if chain in install_branch_config().

Signed-off-by: Adam <Adam@sigterm.info>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-24 12:43:07 -07:00
b0f7c7cf86 t4209: set up expectations up front
Instead of creating an expect file for each test, build three files with
the possible valid values during setup and use them in the tests.  This
shortens the test code and saves nine calls to git rev-parse.

Signed-off-by: Rene Scharfe <l.s.r@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-24 10:58:11 -07:00
b6c2a0d45d parse-options: make sure argh string does not have SP or _
We encourage to spell an argument hint that consists of multiple
words as a single-token separated with dashes.  In order to help
catching violations added by new callers of parse-options, make sure
argh does not contain SP or _ when the code validates the option
definitions.

Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-24 10:43:35 -07:00
ec160ae12b update-index: teach --cacheinfo a new syntax "mode,sha1,path"
The "--cacheinfo" option is unusual in that it takes three option
parameters.  An option with an optional parameter is bad enough.  An
option with multiple parameters is simply insane.

Introduce a new syntax that takes these three things concatenated
together with a comma, which makes the command line syntax more
uniform across subcommands, while retaining the traditional syntax
for backward compatiblity.

If we were designing the "update-index" subcommand from scratch
today, it may probably have made sense to make this option (and
possibly others) a command mode option that does not take any option
parameter (hence no need for arg-help).  But we do not live in such
an ideal world, and as far as I can tell, the command still supports
(and must support) mixed command modes in a single invocation, e.g.

    $ git update-index path1 --add path2 \
        --cacheinfo 100644 $(git hash-object --stdin -w <path3) path3 \
	path4

must make sure path1 is already in the index and update all of these
four paths.  So this is probably as far as we can go to fix this issue
without risking to break people's existing scripts.

Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-24 10:43:35 -07:00
e703d7118c parse-options: multi-word argh should use dash to separate words
"When you need to use space, use dash" is a strange way to say that
you must not use a space.  Because it is more common for the command
line descriptions to use dashed-multi-words, you do not even want to
use spaces in these places.  Rephrase the documentation to avoid
this strangeness.

Fix a few existing multi-word argument help strings, i.e.

 - GPG key-ids given to -S/--gpg-sign are "key-id";
 - Refs used for storing notes are "notes-ref"; and
 - Expiry timestamps given to --expire are "expiry-date".

and update the corresponding documentation pages.

Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-24 10:43:34 -07:00
ce7f8745aa t1502: protect runs of SPs used in the indentation
The expected output from the argument help use runs of SPs to align
the description of each option; a careless use of --whitespace=fix
can turn leading parts of them into appropriate number of HTs.
Prevent such a breakage by prefixing all the expected lines with
leading vertical bars in the original and stripping them with a
small sed script.

Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-23 17:28:03 -07:00
9bab5b6061 rev-parse --parseopt: option argument name hints
Built-in commands can specify names for option arguments when usage text
is generated for a command.  sh based commands should be able to do the
same.

Option argument name hint is any text that comes after [*=?!] after the
argument name up to the first whitespace.

Signed-off-by: Ilya Bobyr <ilya.bobyr@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-23 17:28:03 -07:00
3064b13053 Makefile: describe CHARSET_LIB better
The original explanation was not even grammatically correct or
readable.

Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-23 13:41:38 -07:00
8a2e8da367 userdiff: have 'cpp' hunk header pattern catch more C++ anchor points
The hunk header pattern 'cpp' is intended for C and C++ source code, but
it is actually not particularly useful for the latter, and even misses
some use-cases for the former.

The parts of the pattern have the following flaws:

- The first part matches an identifier followed immediately by a colon
  and arbitrary text and is intended to reject goto labels and C++
  access specifiers (public, private, protected). But this pattern also
  rejects C++ constructs, which look like this:

    MyClass::MyClass()
    MyClass::~MyClass()
    MyClass::Item MyClass::Find(...

- The second part matches an identifier followed by a list of qualified
  names (i.e. identifiers separated by the C++ scope operator '::')
  separated by space or '*' followed by an opening parenthesis (with
  space between the tokens). It matches function declarations like

    struct item* get_head(...
    int Outer::Inner::Func(...

  Since the pattern requires at least two identifiers, GNU-style
  function definitions are ignored:

    void
    func(...

  Moreover, since the pattern does not allow punctuation other than '*',
  the following C++ constructs are not recognized:

  . template definitions:
      template<class T> int func(T arg)

  . functions returning references:
      const string& get_message()

  . functions returning templated types:
      vector<int> foo()

  . operator definitions:
      Value operator+(Value l, Value r)

- The third part of the pattern finally matches compound definitions.
  But it forgets about unions and namespaces, and also skips single-line
  definitions

    struct random_iterator_tag {};

  because no semicolon can occur on the line.

Change the first pattern to require a colon at the end of the line
(except for trailing space and comments), so that it does not reject
constructor or destructor definitions.

Notice that all interesting anchor points begin with an identifier or
keyword. But since there is a large variety of syntactical constructs
after the first "word", the simplest is to require only this word and
accept everything else. Therefore, this boils down to a line that begins
with a letter or underscore (optionally preceded by the C++ scope
operator '::' to accept functions returning a type anchored at the
global namespace). Replace the second and third part by a single pattern
that picks such a line.

This has the following desirable consequence:

- All constructs mentioned above are recognized.

and the following likely desirable consequences:

- Definitions of global variables and typedefs are recognized:

    int num_entries = 0;
    extern const char* help_text;
    typedef basic_string<wchar_t> wstring;

- Commonly used marco-ized boilerplate code is recognized:

    BEGIN_MESSAGE_MAP(CCanvas,CWnd)
    Q_DECLARE_METATYPE(MyStruct)
    PATTERNS("tex",...)

  (The last one is from this very patch.)

but also the following possibly undesirable consequence:

- When a label is not on a line by itself (except for a comment) it is
  no longer rejected, but can appear as a hunk header if it occurs at
  the beginning of a line:

    next:;

IMO, the benefits of the change outweigh the (possible) regressions by a
large margin.

Signed-off-by: Johannes Sixt <j6t@kdbg.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-21 15:03:32 -07:00
9cc444f057 t4018: test cases showing that the cpp pattern misses many anchor points
Most of the tests show C++ code, but there is also a union definition and
a GNU style function definition that are not recognized.

Signed-off-by: Johannes Sixt <j6t@kdbg.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-21 15:03:31 -07:00
02907a08cc t4018: test cases for the built-in cpp pattern
A later patch changes the built-in cpp pattern. These test cases
demonstrate aspects of the pattern that we do not want to change.

Signed-off-by: Johannes Sixt <j6t@kdbg.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-21 15:03:29 -07:00
ad5070fb36 t4018: reduce test files for pattern compilation tests
All test cases that need a file with specific text patterns have been
converted to utilize texts in the t4018/ directory. The remaining tests
in the test script deal only with the validity of the regular
expressions. These tests do not depend on the contents of files that
'git diff' is invoked on. Remove the largish here-document and use only
tiny files.

While we are touching these tests, convert grep to test_i18ngrep as the
texts checked for may undergo translation in the future.

Signed-off-by: Johannes Sixt <j6t@kdbg.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-21 15:03:28 -07:00
f1b75fbaf1 t4018: convert custom pattern test to the new infrastructure
For the test case "matches to end of line", extend the pattern by a few
wildcards so that the pattern captures the "RIGHT" token, which is needed
for verification, without mentioning it in the pattern.

Signed-off-by: Johannes Sixt <j6t@kdbg.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-21 15:02:57 -07:00
dd4dc5c574 t4018: convert java pattern test to the new infrastructure
Signed-off-by: Johannes Sixt <j6t@kdbg.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-21 15:02:36 -07:00
2d08413ba1 t4018: convert perl pattern tests to the new infrastructure
There is one subtlety: The old test case 'perl pattern gets full line of
POD header' does not have its own new test case, but the feature is
tested nevertheless by placing the RIGHT tag at the end of the expected
hunk header in t4018/perl-skip-sub-in-pod.

Signed-off-by: Johannes Sixt <j6t@kdbg.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-21 15:02:19 -07:00
bfa7d01413 t4018: an infrastructure to test hunk headers
Add an infrastructure that simplifies adding new tests of the hunk
header regular expressions.

To add new tests, a file with the syntax to test can be dropped in the
directory t4018. The README file explains how a test file must contain;
the README itself tests the default behavior.

Signed-off-by: Johannes Sixt <j6t@kdbg.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-21 15:00:51 -07:00
abf8f98602 userdiff: support unsigned and long long suffixes of integer constants
Do not split constants such as 123U, 456ll, 789UL at the first U or
second L.

Signed-off-by: Johannes Sixt <j6t@kdbg.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-21 14:48:07 -07:00
407e07f2a6 userdiff: support C++ ->* and .* operators in the word regexp
The character sequences ->* and .* are valid C++ operators. Keep them
together in --word-diff mode.

Signed-off-by: Johannes Sixt <j6t@kdbg.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-21 14:47:50 -07:00
410c3428ed t0001: drop subshells just for "cd"
Many tests do something like:

  (
	mkdir foo &&
	cd foo &&
	git init
  )

You can do the same these days with "git init foo", which
makes the tests shorter and simpler to read.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-21 14:35:13 -07:00
99e1c7367f t0001: drop useless subshells
Many tests use subshells, but don't actually change the
shell environment. They were probably cargo-culted from
earlier tests which did need subshells. Drop the useless
ones.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-21 14:35:08 -07:00
0981140fcc t0001: use test_must_fail
We've hand-rolled several "if" statements looking for
failures. We can use test_must_fail here, which is shorter
and more robust.

Note that we modify the commands slightly (to use "git init
foo" rather than "cd foo && git init") to avoid dealing with
a subshell, but this should not affect the outcome.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-21 14:28:05 -07:00
2a472410cb t0001: use test_config_global
We hand-set several config options using :

  git config -f $HOME/.gitconfig ...

Instead, we can use "test_config_global". Not only is this
more readable, but it cleans up for us so that subsequent
tests aren't polluted by our settings.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-21 14:28:03 -07:00
633734d4a1 t0001: use test_path_is_*
t0001 predates the test_path_is_* helpers, and uses "test
-f" and "test -d" directly. Using the helpers provides
better debugging output, and are a little more robust.
As opposed to "! test -d", test_path_is_missing will
actually makes sure the path does not exist at all.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-21 14:27:58 -07:00
3d06c5f19d t0001: make symlink reinit test more careful
In the final test of t0001, we have a repo whose .git is a
symlink to a directory "here", and we use
"--separate-git-dir" to migrate that to a .git file pointing
to a different directory. We check that the data is migrated
to the new directory and that .git looks like a git-file.

We also check that "here" is not a directory, which is
slightly misleading. It should not be a directory, but
neither should it be gone. It is the actual resting place of
the git-file, and .git remains a symlink to it.

Let's check that more explicitly, both to make our test more
robust, and to make further cleanups in this area more
obvious.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-21 14:27:52 -07:00
f7e8714101 t: prefer "git config --file" to GIT_CONFIG
Doing:

  GIT_CONFIG=foo git config ...

is equivalent to:

  git config --file=foo ...

The latter is easier to read and slightly less error-prone,
because of issues with one-shot variables and shell
functions (e.g., you cannot use the former with
test_must_fail).

Note that we explicitly leave one case in t1300 which checks
the same operation on both GIT_CONFIG and "git config
--file". They are equivalent in the code these days, but
this will make sure it remains so.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-21 14:26:55 -07:00
551a3e60d1 t: prefer "git config --file" to GIT_CONFIG with test_must_fail
This lets us get rid of an extra "env" invocation in the
middle, and is slightly more readable.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-21 14:26:22 -07:00
3cc6a6f0f7 t: stop using GIT_CONFIG to cross repo boundaries
Some tests want to check or set config in another
repository. E.g., t1000 creates repositories and makes sure
that their core.bare and core.worktree settings are what we
expect. We can do this with:

  GIT_CONFIG=$repo/.git/config git config ...

but it better shows the intent to just enter the repository
and let "git config" do the normal lookups:

  (cd $repo && git config ...)

In theory, this would cause us to use an extra subshell, but
in all such cases, we are actually already in a subshell.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-21 14:24:40 -07:00
221bf98506 t: drop useless sane_unset GIT_* calls
Several test scripts manually unset GIT_CONFIG and other
GIT_* variables. These are generally taken care of for us by
test-lib.sh already.

Unsetting these is not only useless, but can be confusing to
a reader, who may wonder why some tests in a script unset
them and others do not (t0001 is particularly guilty of this
inconsistency, probably because many of its tests predate
the test-lib.sh environment-cleansing).

Note that we cannot always get rid of such unsetting. For
example, t9130 can drop the GIT_CONFIG unset, but not the
GIT_DIR one, because lib-git-svn.sh sets the latter. And in
t1000, we unset GIT_TEMPLATE_DIR, which is explicitly
initialized by test-lib.sh.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-21 14:11:11 -07:00
35de3ac1db t/test-lib: drop redundant unset of GIT_CONFIG
This is already handled by the mass GIT_* unsetting added by
95a1d12 (tests: scrub environment of GIT_* variables,
2011-03-15).

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-21 14:11:09 -07:00
a6ca9dfa5c t/Makefile: stop setting GIT_CONFIG
Once upon a time, the setting of GIT_CONFIG in the
environment could affect how tests ran. Commit 9c3796f (Fix
setting config variables with an alternative GIT_CONFIG,
2006-06-20) unconditionally set GIT_CONFIG in the Makefile
when running tests to give us a known starting point.

This is insufficient for running the tests outside of the
Makefile, however, and 8565d2d (Make tests independent of
global config files, 2007-02-15) later set GIT_CONFIG
directly in test-lib.sh. At that point the Makefile setting
was redundant, but we never removed it. Let's do so now.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-21 14:10:30 -07:00
3f09db07b3 Update draft release notes to 2.0 2014-03-21 13:41:27 -07:00
fe3623c635 Merge branch 'lt/request-pull'
Discard the accumulated "heuristics" to guess from which branch the
result wants to be pulled from and make sure what the end user
specified is not second-guessed by "git request-pull", to avoid
mistakes.

* lt/request-pull:
  request-pull: documentation updates
  request-pull: resurrect "pretty refname" feature
  request-pull: test updates
  request-pull: pick up tag message as before
  request-pull: allow "local:remote" to specify names on both ends
  request-pull: more strictly match local/remote branches
2014-03-21 12:50:44 -07:00
53d7d1b129 Merge branch 'es/sh-i18n-envsubst'
* es/sh-i18n-envsubst:
  sh-i18n--envsubst: retire unused string_list_member()
2014-03-21 12:50:39 -07:00
1ddb4d7e5e Merge branch 'nd/upload-pack-shallow'
Serving objects from a shallow repository needs to write a
temporary file to be used, but the serving upload-pack may not have
write access to the repository which is meant to be read-only.

Instead feed these temporary shallow bounds from the standard input
of pack-objects so that we do not have to use a temporary file.

* nd/upload-pack-shallow:
  upload-pack: send shallow info over stdin to pack-objects
2014-03-21 12:49:08 -07:00
6dada01b95 Merge branch 'jn/wt-status'
Unify the codepaths that format new/modified/changed sections and
conflicted paths in the "git status" output and make it possible to
properly internationalize their output.

* jn/wt-status:
  wt-status: lift the artificual "at least 20 columns" floor
  wt-status: i18n of section labels
  wt-status: extract the code to compute width for labels
  wt-status: make full label string to be subject to l10n
2014-03-21 12:48:59 -07:00
10bdb20d6a Merge branch 'jc/stash-pop-not-popped'
"stash pop", upon failing to apply the stash, refrains from
discarding the stash to avoid information loss.  Be more explicit
in the error message.

The wording may want to get a bit more bikeshedding.

* jc/stash-pop-not-popped:
  stash pop: mention we did not drop the stash upon failing to apply
2014-03-21 12:48:51 -07:00
1be645c0b1 Merge branch 'dk/skip-prefix-scan-only-once'
Update implementation of skip_prefix() to scan only once; given
that most "prefix" arguments to the inline function are constant
strings whose strlen() can be determined at the compile time, this
might actually make things worse with a compiler with sufficient
intelligence.

* dk/skip-prefix-scan-only-once:
  skip_prefix(): scan prefix only once
2014-03-21 12:47:41 -07:00
b6de0c633e Merge branch 'nd/tag-version-sort'
Allow v1.9.0 sorted before v1.10.0 in "git tag --list" output.

* nd/tag-version-sort:
  tag: support --sort=<spec>
2014-03-21 12:47:39 -07:00
3e14384b12 Merge branch 'jk/shallow-update-fix'
Serving objects from a shallow repository needs to write a
new file to hold the temporary shallow boundaries but it was not
cleaned when we exit due to die() or a signal.

* jk/shallow-update-fix:
  shallow: verify shallow file after taking lock
  shallow: automatically clean up shallow tempfiles
  shallow: use stat_validity to check for up-to-date file
2014-03-21 12:33:29 -07:00
4291cc10e6 Merge branch 'tc/commit-dry-run-exit-status-tests'
* tc/commit-dry-run-exit-status-tests:
  demonstrate git-commit --dry-run exit code behaviour
2014-03-21 12:33:25 -07:00
93728b23ad config.txt: third-party tools may and do use their own variables
Signed-off-by: Chris Angelico <rosuav@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-21 11:55:07 -07:00
22d55aee20 doc: status, remove leftover statement about '#' prefix
This hasn't been true since 2556b996 (status: disable display of '#'
comment prefix by default, 2013-09-06).

Signed-off-by: Dirk Wallenstein <halsmit@t-online.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-21 11:00:46 -07:00
1a27a15452 tree-diff: simplify tree_entry_pathcmp
Since an earlier "Finally switch over tree descriptors to contain a
pre-parsed entry", we can safely access all tree_desc->entry fields
directly instead of first "extracting" them through
tree_entry_extract.

Use it. The code generated stays the same - only it now visually looks
cleaner.

Signed-off-by: Kirill Smelkov <kirr@mns.spb.ru>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-20 15:04:32 -07:00
5acabd84a6 tree-diff: show_path prototype is not needed anymore
We moved all action-taking code below show_path() in recent HEAD~~
(tree-diff: move all action-taking code out of compare_tree_entry).

Signed-off-by: Kirill Smelkov <kirr@mns.spb.ru>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-20 15:04:32 -07:00
9bc0619655 tree-diff: rename compare_tree_entry -> tree_entry_pathcmp
Since previous commit, this function does not compare entry hashes, and
mode are compared fully outside of it. So what it does is compare entry
names and DIR bit in modes. Reflect this in its name.

Add documentation stating the semantics, and move the note about
files/dirs comparison to it.

Signed-off-by: Kirill Smelkov <kirr@mns.spb.ru>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-20 15:04:32 -07:00
903bba68ab tree-diff: move all action-taking code out of compare_tree_entry()
- let it do only comparison.

This way the code is cleaner and more structured - cmp function only
compares, and the driver takes action based on comparison result.

There should be no change in performance, as effectively, we just move
if series from on place into another, and merge it to was-already-there
same switch/if, so the result is maybe a little bit faster.

Signed-off-by: Kirill Smelkov <kirr@mns.spb.ru>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-20 15:04:31 -07:00
5dfb2bbd8d tree-diff: don't assume compare_tree_entry() returns -1,0,1
It does, but we'll be reworking it in the next patch after it won't, and
besides it is better to stick to standard
strcmp/memcmp/base_name_compare/etc... convention, where comparison
function returns <0, =0, >0

Regarding performance, comparing for <0, =0, >0 should be a little bit
faster, than switch, because it is just 1 test-without-immediate
instruction and then up to 3 conditional branches, and in switch you
have up to 3 tests with immediate and up to 3 conditional branches.

No worry, that update_tree_entry(t2) is duplicated for =0 and >0 - it
will be good after we'll be adding support for multiparent walker and
will stay that way.

=0 case goes first, because it happens more often in real diffs - i.e.
paths are the same.

Signed-off-by: Kirill Smelkov <kirr@mns.spb.ru>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-20 15:04:31 -07:00
d00e980c22 tree-diff: consolidate code for emitting diffs and recursion in one place
Currently both compare_tree_entry() and show_entry() invoke opt diff
callbacks (opt->add_remove() and opt->change()), and also they both have
code which decides whether to recurse into sub-tree, and whether to emit
a tree as separate entry if DIFF_OPT_TREE_IN_RECURSIVE is set.

I.e. we have code duplication and logic scattered on two places.

Let's consolidate it - all diff emiting code and recurion logic moves
to show_entry, which is now named as show_path, because it shows diff
for a path, based on up to two tree entries, with actual diff emitting
code being kept in new helper emit_diff() for clarity.

What we have as the result, is that compare_tree_entry is now free from
code with logic for diff generation, and also performance is not
affected as timings for

    `git log --raw --no-abbrev --no-renames`

for navy.git and `linux.git v3.10..v3.11`, just like in previous patch,
stay the same.

Signed-off-by: Kirill Smelkov <kirr@mns.spb.ru>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2014-03-20 15:03:24 -07:00