When three-way merging indexes, we previously changed each path
as we read them, which would lead to us adding an index entry for
'foo', then removing an index entry for 'foo/file'.  With the new
index requirements, this is not allowed.  Removing entries in the
merged index, then adding them, resolves this.  In the previous
example, we now remove 'foo/file' before adding 'foo'.

In case insensitive index mode, we would stop at a prefixed entry,
treating the provided search key length as a substring, not the
length of the string to match.

Don't strcmp a git_buf, strcmp its char *

Add `git_commit_descendant_of`.

Add built in diff drivers

Writing a sample Javascript driver pointed out some extra
whitespace handling that needed to be done in the diff driver.
This adds some tests with some sample javascript code that I
pulled off of GitHub just to see what would happen.  Also, to
clean up the userdiff test data, I did a "git gc" and packed
up the test objects.

Leftover OID -> ID changes

Buff up returning strings

MSVC doesn't like modern code neither

Make blame example compile on MSVC

A lot of the tests were checking for overflow, which we don't have
anymore, so we can remove them.

Again, we already did this internally, so simply remove the conversions.

Internally we already did everything with git_bufs, so this is just
exposing those functions with public names.

Since we now export that type, we can avoid making the user guess a
size.

Fix git_remote_dup & add missing const

Fix write_object.

Fix local push to file:// URL.

Following the rest of the series, use 'id' when refering to the value.

Same as the other commits in the series, we use 'id' when talking about
thing rather than the datatype.

In the same vein as the previous commits in this series.

Rename git_tree_entry_byoid() to _byid() as per the convention.

This was not converted when we converted the rest, so do it now.

Since I don't have permission yet on the code from Git, I decided
I'd take a stab at writing patterns for PHP and Javascript myself.
I think these are pretty weak, but probably better than the
default behavior without them.

This moves the expected and actual test data along with the source
data for the userdiff tests into the tests/resources/userdiff test
repo and updates the test to use that.