Plug a leak in the index unmerged entries vector

Config bugfix

Add git_repository_is_bare() accessor

A TREE extension with an entry count of -1 means that it was
invalidated and we should ignore it. Do so instead of returning an
error.

This fixes issue #202

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

There are two reasons why read_tree_internal might return a NULL
tree. The first one is a corrupt index, but the second one is an
invalidated TREE extension. Up to now, its only way to communicate
with its caller was through the return value being NULL or not.

Allow read_tree_internal to report its exit status independently from
the tree pointer.

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

The `git_repository_index` call now returns a brand new index that must
be manually free'd.

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

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

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

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

Nothing is allowed betwen the closing quotation mark and the ] so
return an error if there is.

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

Also free the subsection if we find too many quotes

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

If parse_section_header{,_ext} return an error, current_section
doesn't get allocated. Set it to NULL after freeing so we don't try to
free it again.

This fixes part 2-2 of Issue #210.

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

Simplify cfg_readline and at the same time fix it so that it does
really ignore empty lines.

This fixes point 2-1 of Issue #210

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

On error, it would free the configuration object even though it didn't
own that memory, which would cause a double-free.

This fixes the first part of Issue #210

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

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

Fix compilation warnings in MSVC

This allows to successfully build libgit2 with waf on Windows.