diff options
| author | Giovanni Campagna <gcampagn@cs.stanford.edu> | 2015-09-26 15:23:35 -0700 |
|---|---|---|
| committer | Colin Walters <walters@verbum.org> | 2015-09-27 14:32:00 -0400 |
| commit | 4d9453f218074d03a5c44dbd44eeadb8e9e89f6c (patch) | |
| tree | 33f397cf60e89bcecfc72e621b58585e5abfdd68 /giscanner/codegen.py | |
| parent | 232f3c831260f596e36159112292897962a505b4 (diff) | |
| download | gobject-introspection-4d9453f218074d03a5c44dbd44eeadb8e9e89f6c.tar.gz | |
giscanner: fix alias resolution
Atk has a "typedef GSList AtkAttributeSet", which causes
an ast.Alias() targeting a ast.List(). The latter has a
target_fundamental of '<list>', which cannot be looked
up in ast.type_names (because it's not a real fundamental
type), and that breaks the build.
It turns out that we don't need to find the fundamental
type and then the fundamental type node - the target
field of the ast.Alias already points to the desired
type node.
https://bugzilla.gnome.org/show_bug.cgi?id=755681
Diffstat (limited to 'giscanner/codegen.py')
0 files changed, 0 insertions, 0 deletions