racy-git, the missing link

Check the repository version

This is something we do on re-init but not when opening a
repository. This hasn't particularly mattered up to now as the version
has been 0 ever since the first release of git, but the times, they're
a-changing and we will soon see version 1 in the wild. We need to make
sure we don't open those.

git will assume the repository format version is 0 if the value is not
there. Do the same.

Explicitly handle GIT_DELTA_CONFLICTED in git_diff_merge()

This fixes a bug where if a file was in conflicted state in either diff,
it would not always remain in conflicted state in the merged diff.

CRLF

Perform LF->CRLF for core.autocrlf=true on non-Win32 because core
git does.

All platforms do terrible, horrible, no good, very bad translation
when core.autocrlf=true.  It's not just Windows!

Allow files to have mixed line endings instead of skipping processing
on them.

Use statistics (like core git) to control the behavior of the
to workdir CRLF filter.

Support hierarchical test resource data, such that you can have
`tests/resources/foo/bar` and move the `bar` directory in as
a fixture.

Calling `cl_fixture_sandbox` on a path that is not directly beneath
the test resources directory succeeds, placing that directory into
the test fixture.  (For example, `cl_fixture_sandbox("foo/bar")`
will sandbox the `foo/bar` directory as `bar`).

Add support for cleaning up directories created this way, by only
cleaning up the basename (in this example, `bar`) from the fixture
directory.

A corpus of files checked out with Git (Linux, 1.9.1) to ensure that
produce identical data when checking out using a CRLF filter.

A corpus of files checked out with Git for Windows (2.4.1.windows.1)
to ensure that we produce identical data when checking out using a
CRLF filter.

Given a variety of combinations of core.autocrlf settings and
attributes settings, test that we check out data into the working
directory the same as a known-good test resource created by git.git.

Include a shell script that will generate the expected CRLF data,
calling git.git to capture its output as a test resource for the
current platform.

Include the UTF8 and UTF8 BOM tests in the master crlf test
branch for completeness.

Include additional test data for CRLF tests: files with mixed
line endings and binary files.

commit: allow retrieving an arbitrary header field

Write modified index in git_stash_apply()

Even though the file is empty and thus the size in the entry matches, we
should be able to detect it as a difference.

They fit there much better, even though we often check by diffing, it's
about the behaviour of the index.

As we attempt to replicate a situation in which an older checkout has
put a file on disk with different filtering settings from us, set the
timestamp on the entry and file to a second before we're performing the
operation so the entry in the index counts as old.

This way we can test that we're not looking at the on-disk file when the
index has the entry and we detect it as clean.

This allows the user to look up fields which we don't parse in libgit2,
and allows them to access gpgsig or mergetag fields if they wish to
check the signature.

When an entry has a racy timestamp, we need to check whether the file
itself has changed since we put its entry in the index. Only then do we
smudge the size field to force a check the next time around.

When a file on the workdir has the same or a newer timestamp than the
index, we need to perform a full check of the contents, as the update of
the file may have happened just after we wrote the index.

The iterator changes are such that we can reach inside the workdir
iterator from the diff, though it may be better to have an accessor
instead of moving these structs into the header.

Same as with git_stash_save(), there's no reason not to write the index
to disk since it has been modified.

Don't propagate workdir's mode to the index during diff's update index

When updating the index during a diff, preserve the original mode,
which prevents us from dropping the mode to what we have interpreted
as on our system (eg, what the working directory claims it to be,
which may be a lie on some systems.)

Test to ensure that when status updates an index, it does not alter
the original mode for file types that are not supported (eg, symlinks
on Windows).

Fixed index being double-freed in stash tests

Use the checksum to check whether an index has been modified

This is used by the submodule in order to figure out if the index has
changed since it last read it. Using a timestamp is racy, so let's make
it use the checksum, just like we now do for reloading the index itself.

Quote LIBSSH2_LIBRARIES call

When ticking over one second, it can happen that the actual time ticks
over the same second between the time that we undermine our own race
protections and the time in which we perform the index update. Such
timing would make the time in the entries match the index' timestamp and
we have not gained anything.

Ticking over five seconds makes it so that if real-time rolls over that
second, our index is still ahead. This is still suboptimal as we're
dealing with timing, but five seconds should be long enough for any
reasonable test runner to finish the tests.

We currently use a timetamp to check whether an index file has been
modified since we last read it, but this is racy. If two updates happen
in the same second and we read after the first one, we won't detect the
second one.

Instead read the SHA-1 checksum of the file, which are its last 20 bytes which
gives us a sure-fire way to detect whether the file has changed since we
last read it.

As we're now keeping track of it, expose an accessor to this data.