Like we want to separate libgit2 and utility source code, we want to
separate libgit2 and utility tests.  Start by moving all the tests into
libgit2.

The `git2internal` target is actually the git library; call it such so
that IDE users have visibility into it.

Instead of simply including the utility files directly, make them a
cmake object library for easy reusability between other projects within
libgit2.

Now the top-level `src` is responsible for platform selection, while the
next-level `libgit2` and `util` configurations are responsible for
identifying what objects they include.

The `git_error_set` function is useful for callers who implement
backends and advanced callbacks.  Expose it.

Newer gcc is complaining about `object` being potentially not
initialized; initialize it.

win32: `find_system_dirs` does not return `GIT_ENOTFOUND`

Allow for no Git for Windows installation.  When there is no GfW found
in the path or registry, `git_win32__find_system_dirs` would return a
`GIT_ENOTFOUND`.  Callers were not expecting this.  Since this is no
error, we simply return `0` so that callers can move on with their
lives.

remote: do store the update_tips callback error value

We use `git_error_set_after_callback_function` to determine whether
`update_tips` returned an error but do not store its return value making us
think it always returns 0.

Fix it by adding the common patter of storing it inside the `if` when calling it.

meta: update version number to v1.5.0-alpha

xdiff: use xdl_free not free

Free parent and ref in lg2_commit before returning.

cmake: Fix package name for system http-parser

We've added a lovely abstraction layer in xdiff so that it can call our
allocation functions.  But it also needs to call our free functions.  We
missed some `free` calls in `xmerge.c`.  Update them to use `xdl_free`.
Without this, we will pass a pointer allocated with a custom allocator
to the system free function.  💣

Update the version number in main to v1.5.0-alpha.  This helps people
understand that the main builds are not part of the v1.4.0 release
train.

We use "alpha" to indicate builds out of main (or nightlies) as semver
v2 requires the prerelease component is compared lexicographically.
Thus, our "beta" and "rc" releases should follow.

Fix building against system http-parser library by fixing
the find_package() argument.  It seems to have been accidentally changed
from HTTPParser to HTTP_Parser in de178d36, effectively making
the build against system library fail to find it:

```
CMake Warning at cmake/SelectHTTPParser.cmake:3 (find_package):
  By not providing "FindHTTP_Parser.cmake" in CMAKE_MODULE_PATH this project
  has asked CMake to find a package configuration file provided by
  "HTTP_Parser", but CMake did not find one.

  Could not find a package configuration file provided by "HTTP_Parser" with
  any of the following names:

    HTTP_ParserConfig.cmake
    http_parser-config.cmake

  Add the installation prefix of "HTTP_Parser" to CMAKE_PREFIX_PATH or set
  "HTTP_Parser_DIR" to a directory containing one of the above files.  If
  "HTTP_Parser" provides a separate development package or SDK, be sure it
  has been installed.
Call Stack (most recent call first):
  src/CMakeLists.txt:97 (include)

CMake Error at cmake/SelectHTTPParser.cmake:11 (message):
  http-parser support was requested but not found
Call Stack (most recent call first):
  src/CMakeLists.txt:97 (include)
```

meta: show build status for v1.3 and v1.4 branches

libgit2 v1.4

Fix crashes in example programs on Windows (sprintf_s not compatible with snprintf)

merge: fix overlap between GIT_MERGE_FILE_FAVOR__CONFLICTED and GIT_MERGE_FILE_SIMPLIFY_ALNUM

meta: add a release configuration file

diff_file: fix crash if size of diffed file changes in workdir

Let's let GitHub handle our release notes as much as possible:
https://docs.github.com/en/repositories/releasing-projects-on-github/automatically-generated-release-notes

mktmp: improve our temp file creation

We made the flags that enable recursive merge internal, on the
assumption that nobody would want them and they're hard to reason about.
(Giving people an option that nobody wants is just extra noise.)

However, it made it hard for _us_ to reason about.  There's no good
reason to keep it private, let's just make it public and push that
cognitive load onto our poor users.  But they should expect it, they're
dealing with git, after all.

When we know that we know a file's size, and the file's size changes,
fail.

When we know the file size (because we're producing it from a working
directory iterator, or an index with an up-to-date cache) then set a
flag indicating as such.  This removes the ambiguity about a 0 file
size, which could indicate that a file exists and is 0 bytes, or that we
haven't read it yet.

Move the empty tree ID into a global space so that it can be generally
used.