Signed-off-by: NJohannes Sixt <j6t@kdbg.org>

The effects of this patch can be tested on Linux by commenting out

  #define _FILE_OFFSET_BITS 64

in git-compat-util.h.
Signed-off-by: NJohannes Sixt <j6t@kdbg.org>

Signed-off-by: NJohannes Sixt <j6t@kdbg.org>

There are two prerequisites:

- The filesystem supports names with tabs or new-lines.

- Files cannot be removed if their containing directory is read-only.

Previously, whether these preconditions are satisified was tested inside
test_expect_success. We move these tests outside because, strictly
speaking, they are not part of the tests.
Signed-off-by: NJohannes Sixt <j6t@kdbg.org>

Some tests can be run only if a particular prerequisite is available. For
example, some tests require that an UTF-8 locale is available. Here we
introduce functions that are used in this way:

1. Insert code that checks whether the prerequisite is available. If it is,
   call test_set_prereq with an arbitrary tag name that subsequently can be
   used to check for the prerequisite:

      case $LANG in
      *.utf-8)
            test_set_prereq UTF8
            ;;
      esac

2. In the calls to test_expect_success pass the tag name:

      test_expect_success UTF8 '...description...' '...tests...'

3. There is an auxiliary predicate that can be used anywhere to test for
   a prerequisite explicitly:

      if test_have_prereq UTF8
      then
            ...code to be skipped if prerequisite is not available...
      fi
Signed-off-by: NJohannes Sixt <j6t@kdbg.org>

Signed-off-by: NJohannes Sixt <j6t@kdbg.org>

Many tests pass $(pwd) in some form to git and later test that the output
of git contains the correct value of $(pwd). For example, the test of
'git remote show' sets up a remote that contains $(pwd) and then the
expected result must contain $(pwd).

Again, MSYS-bash's path mangling kicks in: Plain $(pwd) uses the MSYS style
absolute path /c/path/to/git. The test case would write this name into
the 'expect' file. But when git is invoked, MSYS-bash converts this name to
the Windows style path c:/path/to/git, and git would produce this form in
the result; the test would fail.

We fix this by passing -W to bash's pwd that produces the Windows-style
path.

There are a two cases that need an accompanying change:

- In t1504 the value of $(pwd) becomes part of a path list. In this case,
  the lone 'c' in something like /foo:c:/path/to/git:/bar inhibits
  MSYS-bashes path mangling; IOW in this case we want the /c/path/to/git
  form to allow path mangling. We use $PWD instead of $(pwd), which always
  has the latter form.

- In t6200, $(pwd) - the Windows style path - must be used to construct the
  expected result because that is the path form that git sees. (The change
  in the test itself is just for consistency: 'git fetch' always sees the
  Windows-style path, with or without the change.)
Signed-off-by: NJohannes Sixt <j6t@kdbg.org>

t1002-read-tree-m-u-2way.sh uses 'sum', but it does not rely on the exact
form of the sum, only that it is a hash digest. Therefore, we can sneak
in 'md5sum' under the name 'sum'.
Signed-off-by: NJohannes Sixt <j6t@kdbg.org>

If the PATH lists the Windows system directories before the MSYS
directories, Windows's own incompatible sort and find commands would be
picked up. We implement these commands as functions and call the real
tools by absolute path.
Signed-off-by: NJohannes Sixt <j6t@kdbg.org>

MSYS's bash rewrites /something/bin/... into a Windows path that looks like
c:/msysgit/something/bin/... before git sees it. But later the test case
verifies that the path was used and compares it to the unmangled version.
This fails, of course. This make the path relative so that the path
mangling is not triggered.
Signed-off-by: NJohannes Sixt <j6t@kdbg.org>

We do not have /dev/zero on Windows. This replaces it by data generated
with printf, perl, or echo. Most of the cases do not depend on that the
data is a stream of zero bytes, so we use something printable; nor is an
unlimited stream of data needed, so we produce only as many bytes as the
test cases need.
Signed-off-by: NJohannes Sixt <j6t@kdbg.org>

On Windows, there is an unfortunate interaction between the MSYS bash and
git's command line processing:

- Since Windows's CMD does not do the wildcard expansion, but passes
  arguments like path* through to the programs, the programs must do the
  expansion themselves. This happens in the startup code before main() is
  entered.

- bash, however, passes the argument "path*" to git, assuming that git will
  see the unquoted word unchanged as a single argument.

But actually git expands the unquoted word before main() is entered.

In t2200, not all names that the test case is interested in exist as files
at the time when 'git ls-files' is invoked. git expands "path?" to only
the subset of files the exist, and only that subset was listed, so that the
test failed.  We now list all interesting paths explicitly.

In t7004, git exanded the pattern "*a*" to "actual" (the file that stdout
was redirected to), which is not what the was tested for. We fix it by
renaming the output file (and removing any existing files matching *a*).
This was originally fixed by Johannes Schindelin.
Signed-off-by: NJohannes Sixt <j6t@kdbg.org>

On Windows, you cannot remove files that are in use, not even with
'rm -rf'.  So we need to run 'exec <foo/bar' inside a subshell lest
removing the whole test repository fail.
Signed-off-by: NJohannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: NJohannes Sixt <j6t@kdbg.org>

This function replaces sequences of 'chmod +x' and 'git update-index
--chmod=+x' in the test suite, whose purpose is to help filesystems
that need core.filemode=false. Two places where only 'chmod +x' was used
we also use this new function.

The function calls 'git update-index --chmod' without checking
core.filemode (unlike some of the call sites did). We do this because the
call sites *expect* that the executable bit ends up in the index (ie. it
is not the purpose of the call sites to *test* whether git treats
'chmod +x' and 'update-index --chmod=+x' correctly). Therefore, on
filesystems with core.filemode=true the 'git update-index --chmod' is a
no-op.

The function uses --add with update-index to help one call site in
t6031-merge-recursive. It makes no difference for the other callers.
Signed-off-by: NJohannes Sixt <j6t@kdbg.org>

Since the test case counter was incremented very late, there were a few
users of the counter had to do their own incrementing. Now we increment it
early and simplify these users.
Signed-off-by: NJohannes Sixt <j6t@kdbg.org>

In particular:

- Test case counting can be achieved by arithmetic expansion.

- The name of the test, e.g. t1234, can be computed with ${0%%} and ${0##}.
Signed-off-by: NJohannes Sixt <j6t@kdbg.org>

There were some uses of 'say' inside test_expect_success. But if the tests
were not run in verbose mode, this message went to /dev/null. Pull them out
of test_expect_success.
Signed-off-by: NJohannes Sixt <j6t@kdbg.org>

Some tests report that some tests will be skipped.  They used
'test_expect_success' with a trivially successful test.  Nowadays we have
the helper function 'say' for this purpose.

In on case, 'say_color skip' is replaced by 'say' because the former is
not intended as a public API.
Signed-off-by: NJohannes Sixt <j6t@kdbg.org>

The tests do not depend on that the clones are hard-linked, but used
--local only as an optimization: At the time that --local was used first
in t9400 hard-linked clones were not the default, yet.

By removing --local, we help filesystems that do not support hard-links.
Signed-off-by: NJohannes Sixt <j6t@kdbg.org>

Signed-off-by: NJunio C Hamano <gitster@pobox.com>

* jc/maint-1.6.0-read-tree-overlay:
  read-tree A B C: do not create a bogus index and do not segfault

* db/maint-missing-origin:
  Remove total confusion from git-fetch and git-push
  Give error when no remote is configured

* js/sideband-stderr:
  winansi: support ESC [ K (erase in line)
  recv_sideband: Bands #2 and #3 always go to stderr

* js/rsync-local:
  rsync transport: allow local paths, and fix tests

* rs/color-grep:
  grep: prefer builtin over external one when coloring results
  grep: cast printf %.*s "precision" argument explicitly to int
  grep: add support for coloring with external greps
  grep: color patterns in output
  grep: add pmatch and eflags arguments to match_one_pattern()
  grep: remove grep_opt argument from match_expr_eval()
  grep: micro-optimize hit collection for AND nodes

* js/remote-improvements: (23 commits)
  builtin-remote.c: no "commented out" code, please
  builtin-remote: new show output style for push refspecs
  builtin-remote: new show output style
  remote: make guess_remote_head() use exact HEAD lookup if it is available
  builtin-remote: add set-head subcommand
  builtin-remote: teach show to display remote HEAD
  builtin-remote: fix two inconsistencies in the output of "show <remote>"
  builtin-remote: make get_remote_ref_states() always populate states.tracked
  builtin-remote: rename variables and eliminate redundant function call
  builtin-remote: remove unused code in get_ref_states
  builtin-remote: refactor duplicated cleanup code
  string-list: new for_each_string_list() function
  remote: make match_refs() not short-circuit
  remote: make match_refs() copy src ref before assigning to peer_ref
  remote: let guess_remote_head() optionally return all matches
  remote: make copy_ref() perform a deep copy
  remote: simplify guess_remote_head()
  move locate_head() to remote.c
  move duplicated ref_newer() to remote.c
  move duplicated get_local_heads() to remote.c
  ...

Conflicts:
	builtin-clone.c

* tr/maint-1.6.0-send-email-irt:
  send-email: test --no-thread --in-reply-to combination
  send-email: respect in-reply-to regardless of threading

Conflicts:
	t/t9001-send-email.sh

* kb/checkout-optim:
  Revert "lstat_cache(): print a warning if doing ping-pong between cache types"
  checkout bugfix: use stat.mtime instead of stat.ctime in two places
  Makefile: Set compiler switch for USE_NSEC
  Create USE_ST_TIMESPEC and turn it on for Darwin
  Not all systems use st_[cm]tim field for ns resolution file timestamp
  Record ns-timestamps if possible, but do not use it without USE_NSEC
  write_index(): update index_state->timestamp after flushing to disk
  verify_uptodate(): add ce_uptodate(ce) test
  make USE_NSEC work as expected
  fix compile error when USE_NSEC is defined
  check_updates(): effective removal of cache entries marked CE_REMOVE
  lstat_cache(): print a warning if doing ping-pong between cache types
  show_patch_diff(): remove a call to fstat()
  write_entry(): use fstat() instead of lstat() when file is open
  write_entry(): cleanup of some duplicated code
  create_directories(): remove some memcpy() and strchr() calls
  unlink_entry(): introduce schedule_dir_for_removal()
  lstat_cache(): swap func(length, string) into func(string, length)
  lstat_cache(): generalise longest_match_lstat_cache()
  lstat_cache(): small cleanup and optimisation

As far as I know, not all grep programs support coloring, so we should
rely on builtin grep. If you want external grep, set
color.grep.external to empty string.
Signed-off-by: NNguyễn Thái Ngọc Duy <pclouds@gmail.com>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

Signed-off-by: NEmil Sit <sit@emilsit.net>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

The code will end up calling lstat() to check whether the
file still exists; obviously this doesn't work if we're not
in the worktree.
Signed-off-by: NJeff King <peff@peff.net>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

'git branch -f a b' resets a to b when a exists, rather then deleting a.
Say so in the documentation.
Signed-off-by: NMichael J Gruber <git@drmicha.warpmail.net>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

When --pickaxe-regex is used, forward past the end of matches instead of
advancing to the byte after their start.  This way matches count only
once, even if the regular expression matches their tail -- like in the
fixed-string fork of the code.

E.g.: /.*/ used to count the number of bytes instead of the number of
lines.  /aa/ resulted in a count of two in "aaa" instead of one.

Also document the fact that regexec() needs a NUL-terminated string as
its second argument by adding an assert().
Signed-off-by: NRene Scharfe <rene.scharfe@lsrfire.ath.cx>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

Signed-off-by: NStephen Boyd <bebarino@gmail.com>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

Several old tests were written before test_cmp was introduced, convert
these to test_cmp.

If were are at it, fix the order of the arguments where necessary to
make expected come first, so the command shows how the test result
deviates from the correct output.
Signed-off-by: NMiklos Vajna <vmiklos@frugalware.org>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

If literal text (asciidoc `...`) can be rendered in a differently from
normal text for each output format (man, HTML), then we do not need
extra quotes or other wrapping around inline literal text segments.

config.txt

  Change '`...`' to `...`. In asciidoc, the single quotes provide
  emphasis, literal text should be distintive enough.

  Change "`...`" to `...`. These double quotes do not work if present
  in the described config value, so drop them.

git-checkout.txt

  Change "`...`" to `...` or `"..."`. All instances are command line
  argument examples. One "`-`" becomes `-`. Two others are involve
  curly braces, so move the double quotes inside the literal region to
  indicate that they might need to be quoted on the command line of
  certain shells (tcsh).

git-merge.txt

  Change "`...`" to `...`. All instances are used to describe merge
  conflict markers. The quotes should are not important.

git-rev-parse.txt

  Change "`...`" to `...`. All instances are around command line
  arguments where no in-shell quoting should be necessary.

gitcli.txt

  Change `"..."` to `...`. All instances are around command line
  examples or single command arguments. They do not semanticly belong
  inside the literal text, and they are not needed outside it.

glossary-content.txt
user-manual.txt

  Change "`...`" to `...`. All instances were around command lines.
Signed-off-by: NChris Johnsen <chris_johnsen@pobox.com>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

Signed-off-by: NChris Johnsen <chris_johnsen@pobox.com>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

It is broken because of the tricks we have to play with
lstat to get the bearable perfomance out of the call.
Sadly, it disables access to Cygwin's executable attribute,
which Windows filesystems do not have at all.
Signed-off-by: NAlex Riesen <raa.lkml@gmail.com>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

This reverts commit 7734f048.

I guess that the reverted commit, 7734f048, has been in test long
enough, and should now be reverted.  I have not received any info
regarding any debug output of the reverted commit, so lets hope that
the lstat_cache() function do not cause any ping-pong.
Signed-off-by: NKjetil Barvik <barvik@broadpark.no>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

Signed-off-by: NJohannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: NJohannes Sixt <j6t@kdbg.org>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>