checkout: use cache when possible to determine if workdir item is dirty

Add `git_repository_open_bare`

If the on-disk file has been staged (it's stat data matches the stat data
in the cache) then we need not hash the file to determine whether it
differs from the checkout target; instead we can simply use the oid in
the index.

This prevents recomputing a file's hash unnecessarily, prevents loading
the file (when filtering) and prevents edge cases where filters suggest
that a file is dirty immediately after git writes the file.

Update link to Perl bindings

This is the last minor release before 1.0preview1.

Highlights of this release include:

- Branch API
- Checkout head, index and tree
- Finished clone support
- Abstracted reference API to use custom backends
- Full diff support
- New (faster) packbuilder
- Push support
- New Remotes API
- Revparse support (single and range commits)
- Stash support
- Submodules support

As always, the full changelog is available at:

	http://libgit2.github.com/libgit2/#p/changelog

Yeah, it's a huge release. Releasing stuff sucks.
Expect 1.0 and API freeze in less than a month.

Your faithful maintainer,
	 vmg

Signed-off-by: Vicent Marti <tanoku@gmail.com>

Unified rev-parse, with a revision object

Fetch fixes

Don't try to update anything if there are no heads to update. This
saves us from trying to look into a fetch refspec when there is none.

A better fix for compatibility with git when using remotes without
refspecs is still needed, but this stops us from segfaulting.

Keep the data around until free, as expected by our own fetch example

This will probably prevent many lookup/free
operations in calling code.

OpenBSD & sparc64 fixes

Mixing int and size_t through pointers leads to problems
in big-endian machines.

commit: correctly detect the start of the commit message

The end of the header is signaled by to consecutive LFs and the commit
message starts immediately after. Jumping over LFs at the start of the
message is a bug and leads to creating different commits if
when rebuilding history.

This also fixes an empty commit message being returned as "\n".

dec refcount on refdb instead of always freeing

Branch upstream configuration

Add functions to set and unset the upstream configuration to
complement the getter we already have.

Correctly return matched pathspec when passing "*" or "."

I also moved all tests related to notifying in their own file.

Reduce the number of unnecessary objects in pushed packs

Submodule diff tests and fixes

This adds tests for diffs with submodules in them and (perhaps
unsurprisingly) requires further fixes to be made.  Specifically,
this fixes:

- when considering if a submodule is dirty in the workdir, it was
  being treated as dirty even if only the index was dirty.
- git_diff_patch_to_str (and git_diff_patch_print) were "printing"
  the headers for files (and submodules) that were unmodified or
  had no meaningful content.
- added comment to previous fix and removed unneeded parens.