The workdir iterator has always tried to ignore .git files, but
it turns out there were some bugs.  This makes it more robust at
ignoring .git files.

This also makes iterators always check ".git" case insensitively
regardless of the properties of the system.  This will make libgit2
skip ".GIT" and the like.  This is different from core git, but on
systems with case insensitive but case preserving file systems,
allowing ".GIT" to be added is problematic.

This adds "." ".." and ".git" to the internal ignores list by
default - asking about paths with these files will always say
that they are ignored.

This checks for a leading '.' before looking for the invalid
tree entry names.  Even on pretty high levels of optimization,
this seems to make a measurable improvement.

I accidentally used && in the check initially instead of || and
while debugging ended up improving the error reporting of issues
with adding tree entries.  I thought I'd leave those changes, too.

Filename validation

This prevents the index api from calling write_tree() with a
bogus tree.

As per core git.

Indexer: Avoid a possible double-deletion in error case

Itsy bitsy tiny winy minor api related fixes

Passing 0 as the index now retrieves the most recent entry instead
of the oldest one.

odb: recursively load alternates

The maximum depth is 5, like in git

Win32 fixes for diff/checkout/reset

Add option to ignore file mode in diffs

Diff API signatures

A number of diff APIs and the `git_checkout_index` API take a
`git_repository` object an operate on the index.  This updates
them to take a `git_index` pointer explicitly and only fall back
on the `git_repository` index if the index input is NULL.  This
makes it easier to operate on a temporary index.

The index iterator could previously only be created from a repo
object, but this allows creating an iterator from a `git_index`
object instead (while keeping, though renaming, the old function).

The diff API is not in the parameter order one would expect from
other libgit2 APIs.  This fixes that.

The reset hard tests had hardcoded expected file content and was
not correctly compensating for CRLF filtering when a file needed
to be reverted by the reset hard.  This fixes that.

The existing p_lstat implementation on win32 is not quite POSIX
compliant when setting errno to ENOTDIR.  This adds an option to
make is be compliant so that code (such as checkout) that cares
to have separate behavior for ENOTDIR can use it portably.

This also contains a couple of other minor cleanups in the
posix_w32.c implementations to avoid unnecessary work.

Using the builtin strcmp and strcasecmp as function pointers is
problematic on win32.  This adds internal implementations and
divorces us from the platform linkage.

Explain a few CMake options in the README

Expose STDCALL and explain a few useful CMake options, as well as
explain the quirks building on Windows.

Allocate with calloc rather than conditionally memsetting a specific
part of the struct later on.