Since the libgit2 has a default sorting method, which we also mention as
default in the documentation, there is no particular need to make the
user choose a sorting method when the order does not matter. We use
sorting NONE in that case.
We don't always need our own copy of a C string; sometimes we just need
to parse it. Create py_str_borrow_c_str() which returns a char pointer
to python's internal value string, with which we can avoid an extra copy.
The tree entries exist more or less independently of the index they were
retrieved from. The idea behind them is that they can be kept by a
binding without needing to refer to the original anymore, which may
disappear at any moment if the index is modified.
Keep a copy of the data in TreeEntry instead of pointing to the one
retrieved from the index, which is not safe to keep around.
The Reference, Branch, ConfigIter and Walker types were allowed to be created by the
user, but no constructor was set, and the default values are either
meaningless (in the case of Reference/Branch) or would cause a segfault
as soon as one tried to use them (ConfigIter and Walker).
Remove the constructor from these types, as they don't serve a purpose
and can only be used by mistake.
Fortunately pypy provides support for a lot of the CPython API, so the
changes are minimal.
The most important changes are:
- constructors always get a keyword argument dictionary, even if no
keyword arguments are passed
- trying to assign to a read-only attribute raises TypeError instead of
AttributeError
Apart from that, pypy does not provide MAXPATHLEN. There is a hack in
place currently, but there is only place that's using that macro, and
there shouldn't be a need for it much longer.
This fixes #209.
It seems I have been forgetting to update the documentation with the
last few changes, so adjust to the oid -> id renaming and add missing
attributes to the listings.
This is a lot more pythonic than two sets of getter-setter
functions. The old ones are left for backwards compatibility but they
should be removed in the next release.
Allow looking in subtrees as a convenience in the 'contains'
function. As slashes are not allowed to be part of the name of an entry,
there is no ambiguity in what they mean.
