Handle templated aliases. #85
Conversation
emilio
force-pushed
the
template-alias
branch
from
2b4740c to
0a4a202
Compare
| ty.name().map(ToOwned::to_owned) | ||
| .unwrap_or_else(|| format!("_bindgen_ty{}", self.id())) | ||
| // We really codegen and use the inner type, so use an empty | ||
| // base name so codegen doesn't get confused. |
There was a problem hiding this comment.
The start of this sentence has some funky grammar that prevents me from understanding it.
Is it supposed to be
We really want codegen to
anduse the inner type, ...
?
There was a problem hiding this comment.
I meant something like "We call codegen on the inner type, and since we use the parent chain to generate the canonical name, and we don't want this to appear, and can't be referenced itself from another type (we point to the actual alias for that), there's no need to generate a name for it".
Does that make sense?
There was a problem hiding this comment.
Thanks, I think I understand much better now. I would phrase it like this:
We call codegen on the inner type, but we do not want this alias's name to appear in the canonical name just because it is in the inner type's parent chain, so we use an empty base name.
How is that?
(Thanks for painting this bike shed with me)
There was a problem hiding this comment.
Sounds good, I'll update the comment. And no problem, this is one of the trickiest things in the new bindgen rewrite I think :)
| @@ -0,0 +1,13 @@ | |||
| // bindgen-flags: --enable-cxx-namespaces -- -std=c++14 | |||
There was a problem hiding this comment.
What happens if we don't have --enable-cxx-namespaces here? Will the generated rust code be all in the same namespace, or will it fail, or something else?
There was a problem hiding this comment.
It's the exact same test that tests/headers/template_alias.hpp, just that flag changes, so you can see the output right away :)
There was a problem hiding this comment.
D'oh this was exactly what I was looking for :P
Not sure how I missed that...
|
☔ The latest upstream changes (presumably #82) made this pull request unmergeable. Please resolve the merge conflicts. |
|
@emilio r=me once it is rebased. |
…ly applicable before discarding them.
emilio
force-pushed
the
template-alias
branch
from
6484e3e to
2a3f930
Compare
|
@bors-servo: r=nox |
|
📌 Commit 2a3f930 has been approved by |
|
☀️ Test successful - status-travis |
Fixes #83.
cc @fitzgen
r? @nox