Implement config parsing. Finally!

Take this opportunity to fix an instance of returning
GIT_EOBJCORRUPTED when malloc failed.

Signed-off-by: Carlos Martín Nieto <cmn@elego.de>

Now we use a simple closed-addressing cache. Cuckoo hashing was creating
too many issues with race conditions. Fuck that.

Let's see what happens performance wise, we may have to roll back or
come up with another way to implement an efficient multi-threaded cache.

Do not try to adquire the same node lock twice when the cuckoo hashing
resolves to the same node.

Fix a typo in README.md

This is not really an error, just a special return code to mark the end
of an iteration.

Updated fileops.c and repository.c to new error handling mechanism

fix git_otype typo when calling `git_odb_read_header`.

Replace errors

Add --with-redis flag to list of available build flags in README.

Signed-off-by: schu <schu-github@schulog.org>

Remove "redundant" check for v->_cmp in wrapper function
git_vector_bsearch().

Signed-off-by: schu <schu-github@schulog.org>

Signed-off-by: schu <schu-github@schulog.org>

Signed-off-by: schu <schu-github@schulog.org>

Hey, welcome to yet another minor libgit2 release. Sorry for the delay from
the last one. As you'll see the changelog is quite extensive -- hopefully from
now on we'll stick to more frequent minor releases.

Together with the usual bugfixes, here's a list of the new key features:

* Distfiles

	This version comes with proper distfiles as requested in #131. These are
	available in the Downloads section of the GitHub project.

* Error handling

	A new error handling API has been implemented that allows the library to
	return detailed error messages together with the generic error codes. We
	hope this will be a great when wrapping and integrating the library

	New external method to get the last detailed error message:

		+ git_lasterror(void)

	The old `git_strerror` still exists, but will be deprecated in the future
	as soon as every method in the library returns a valid error message.

	The task of writing error messages for every method is quite daunting.
	We appreciate pull requests with more error messages. Check the new error
	handling documentation in the following commit:

	https://github.com/libgit2/libgit2/commit/fa59f18d0ddbbb98d45e33934fb0efc3e2bf1557

* Redis backend

	We now have a Redis backend courtesy of Dmitry Kovega. Just like the
	SQLite backend, this allows the library to store Git objects in a Redis
	key-value store.

	The backend requires the `hiredis` library. Use `--with-redis` when
	building libgit2 to enable building the backend if `hiredis` is available.

* Commits

	New methods to access tree and parent data as a raw OID value
	instead of forcing a repository lookup

		+ git_commit_tree_oid(git_commit *commit)
		+ git_commit_parent_oid(git_commit *commit, unsigned int n)

* Index

	The `git_index_add` method has been split into 4 different calls
	which allow for appending and replacing in-memory entries and on-disk
	files to the index.

		+ git_index_add(git_index *index, const char *path, int stage)
		+ git_index_add2(git_index *index, const git_index_entry *source_entry)
		+ git_index_append(git_index *index, const char *path, int stage)
		+ git_index_append2(git_index *index, const git_index_entry *source_entry)

	Index entries can now also be efficiently removed from the index:

		+ git_index_remove(git_index *index, int position)

* References

	Methods to force the creation and renaming of references, even if those already
	exist on the repository.

		+ git_reference_create_symbolic_f(git_reference **ref_out, git_repository *repo,
				const char *name, const char *target)
		+ git_reference_create_oid_f(git_reference **ref_out, git_repository *repo,
				const char *name, const git_oid *id)
		+ git_reference_rename_f(git_reference *ref, const char *new_name)

* Repository

	New auxiliary methods with repository information

		+ git_repository_is_empty(git_repository *repo)
		+ git_repository_path(git_repository *repo)
		+ git_repository_workdir(git_repository *repo)

* Signatures

	New method to create a signature with the current date/time

		+ git_signature_now(const char *name, const char *email)

* Tags

	Several wrappers to automate tag creation.

		+ git_tag_create_frombuffer(git_oid *oid, git_repository *repo,
				const char *buffer)
		+ git_tag_create_f(git_oid *oid, git_repository *repo,
				const char *tag_name, const git_oid *target,
				git_otype target_type, const git_signature *tagger,
				const char *message);
		+ git_tag_create_fo(git_oid *oid, git_repository *repo,
				const char *tag_name, const git_object *target,
				const git_signature *tagger, const char *message)

	New functionality to delete and list tags in a repository without
	having to resort to the `references` API.

		+ git_tag_delete(git_repository *repo, const char *tag_name)

		+ git_tag_list(git_strarray *tag_names, git_repository *repo)

* Trees

	All instances of `git_tree_entry` are now returned and handled
	as constant, to remind the user that these opaque types are not
	supposed to be manually free'd.

	The `git_tree_entry_2object` method now takes a `git_repository`
	argument which defines in which repository the resolved object
	should be looked up. (It is expected to be the same repository
	that contains the parent `git_tree` for the entry).

		+ git_tree_entry_2object(git_object **object_out, git_repository *repo,
				const git_tree_entry *entry)

	New opaque type `git_treebuilder` with functionality to create and
	write trees on memory

		+ git_treebuilder_create(git_treebuilder **builder_p, const git_tree *source)
		+ git_treebuilder_clear(git_treebuilder *bld)
		+ git_treebuilder_free(git_treebuilder *bld)
		+ git_treebuilder_get(git_treebuilder *bld, const char *filename)
		+ git_treebuilder_insert(git_tree_entry **entry_out, git_treebuilder *bld,
				const char *filename, const git_oid *id, unsigned int attributes)
		+ git_treebuilder_remove(git_treebuilder *bld, const char *filename)
		+ git_treebuilder_filter(git_treebuilder *bld,
				int (*filter)(const git_tree_entry *, void *), void *payload)
		+ git_treebuilder_write(git_oid *oid, git_repository *repo, git_treebuilder *bld)

	New method to write an index file as a tree to the ODB.

		+ git_tree_create_fromindex(git_oid *oid, git_index *index)

Thanks to the usual guility parties that make this this happen, to
all the new contributors who are starting to submit pull requests, and
to the bindings developers who have to keep up with our shit.

Feedback and questions welcome on libgit2@librelist.org

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

We cannot totally deprecate this until the new error handling mechanisms
are all in place.

Fix issue #79 - git_lasterror() isn't appearing in git2.dll in windows.

Move signature.c and tag.c to the new error handling

Signed-off-by: Carlos Martín Nieto <cmn@elego.de>

Signed-off-by: Carlos Martín Nieto <cmn@elego.de>

Configuration options can come from different sources. Currently,
there is only support for reading them from a flat file, but it might
make sense to read it from a database at some point.

Move the parsing code into src/config_file.c and create an include
file include/git2/config_backend.h to allow for other backends to be
developed.

Signed-off-by: Carlos Martín Nieto <cmn@elego.de>

The GIT_EXPORT macro is used to declare a function to be externally
accessible to other libraries.  This commit uses GIT_EXPORT to declare
the git_lasterror() function as externally exported.  I verified with
depends.exe that the function is available to external callers (i.e.
in the exports table of the PE file).

Ok, this is the real deal. Hopefully. Here's how it's going to work:

- One main method, called `git__throw`, that sets the error
	code and error message when an error happens.

	This method must be called in every single place where an error
	code was being returned previously, setting an error message
	instead.

	Example, instead of:

		return GIT_EOBJCORRUPTED;

	Use:

		return git__throw(GIT_EOBJCORRUPTED,
			"The object is missing a finalizing line feed");

	And instead of:

		[...] {
			error = GIT_EOBJCORRUPTED;
			goto cleanup;
		}

	Use:

		[...] {
			error = git__throw(GIT_EOBJCORRUPTED, "What an error!");
			goto cleanup;
		}

	The **only** exception to this are the allocation methods, which
	return NULL on failure but already set the message manually.

		/* only place where an error code can be returned directly,
		   because the error message has already been set by the wrapper */
		if (foo == NULL)
			return GIT_ENOMEM;

- One secondary method, called `git__rethrow`, which can be used to
fine-grain an error message and build an error stack.

	Example, instead of:

		if ((error = foobar(baz)) < GIT_SUCCESS)
			return error;

	You can now do:

		if ((error = foobar(baz)) < GIT_SUCCESS)
			return git__rethrow(error, "Failed to do a major operation");

	The return of the `git_lasterror` method will be a string in the
	shape of:

		"Failed to do a major operation. (Failed to do an internal
		operation)"

	E.g.

		"Failed to open the index. (Not enough permissions to access
		'/path/to/index')."

	NOTE: do not abuse this method. Try to write all `git__throw`
	messages in a descriptive manner, to avoid having to rethrow them to
	clarify their meaning.

	This method should only be used in the places where the original
	error message set by a subroutine is not specific enough.

	It is encouraged to continue using this style as much possible to
	enforce error propagation:

		if ((error = foobar(baz)) < GIT_SUCCESS)
			return error; /* `foobar` has set an error message, and
							 we are just propagating it */

The error handling revamp will take place in two phases:

	- Phase 1: Replace all pieces of code that return direct error codes
	with calls to `git__throw`. This can be done semi-automatically
	using `ack` to locate all the error codes that must be replaced.

	- Phase 2: Add some `git__rethrow` calls in those cases where the
	original error messages are not specific enough.

Phase 1 is the main goal. A minor libgit2 release will be shipped once
Phase 1 is ready, and the work will start on gradually improving the
error handling mechanism by refining specific error messages.

OTHER NOTES:

	- When writing error messages, please refrain from using weasel
	words. They add verbosity to the message without giving any real
	information. (<3 Emeric)

	E.g.

		"The reference file appears to be missing a carriage return"
			Nope.

		"The reference file is missing a carriage return"
			Yes.

	- When calling `git__throw`, please try to use more generic error
	codes so we can eventually reduce the list of error codes to
	something more reasonable. Feel free to add new, more generic error
	codes if these are going to replace several of the old ones.

	E.g.

		return GIT_EREFCORRUPTED;

	Can be turned into:

		return git__throw(GIT_EOBJCORRUPTED,
			"The reference is corrupted");