Null byte in hunk header

Make the git_signature `const` in the stash API.

The diff hunk context string that is returned to xdiff need not
be NUL terminated because the xdiff code just copies the number of
bytes that you report directly into the output.  There was an off
by one in the diff driver code when the header context was longer
than the output buffer size, the output buffer length included
the NUL byte which was copied into the hunk header.

Fixes #1710

Segmentation fault on git_clone

Fix segfault in git_status_foreach_ext()

Add tests for the `GIT_STATUS_SHOW_XXX` flags.

[trivial] Fix small typo in docs for git_repository_message.

[trivial] Fix docs to use proper enum names

Fix Win32/64 compilation warnings & memory leak

Test submodules with empty index or orphaned head

In both of these cases, the submodule data should still be loaded
just (obviously) without the data that comes from either the index
or the HEAD.

This fixes a bug in the orphaned head case.

Dirty submodules do not correctly appear in diffs

There was a bug where submodules whose HEAD had not been moved
were being marked as having an UNMODIFIED delta record instead
of being left MODIFIED.  This fixes that and fixes the tests to
notice if a submodule has been incorrectly marked as UNMODIFIED.

Prefer younger merge bases over older ones.

Unbreak git_remote_ls on a local transport after disconnecting.

Correctly handle win32 junctions

git-core prefers younger merge bases over older ones in case that multiple valid merge bases exists.

A junction has S_IFDIR | S_IFLNK set, however, only one makes sense.

Signed-off-by: Sven Strickroth <email@cs-ware.de>

This reverts commit 32c12ea6.

Minor point release! We got a lot of rather large features that we
wanted to get settled in:

- New (threadsafe) cache for objects
- Iterator for Status
- New Merge APIs
- SSH support on *NIX
- Function context on diff
- Namespaces support
- Index add/update/remove with wildcard support
- Iterator for References
- Fetch and push refspecs for Remotes
- Rename support in Status
- New 'sys/` namespace for external headers with low-level APIs

As always, this comes with hundreds of bug fixes and performance
improvements. We're faster and better than ever. And we haven't broken
many APIs this time!

Build stuff.

Fix a leak in the local transport code.

Fix checkout tests on Windows

Fixed most documentation header bugs

Add O_CLOEXEC to open calls

In loose objects backend, constrain mkdir calls to avoid extra mkdirs

In theory, p_stat should never return an S_ISLNK result, but due
to the current implementation on Windows with mount points it is
possible that it will.  For now, work around that by allowing a
link in the path to a directory being created.  If it is really a
problem, then the issue will be caught on the next iteration of
the loop, but typically this will be the right thing to do.