Drop SimpleAnnotationReader#199
Conversation
|
Question: do we want to enforce importing annotations (as in "stop looking for same-namespace annotations") to remove the ambiguity in resolved symbols? That was the biggest problem with this reader: too much implicit resolution. |
288062e to
6b2f430
Compare
6b2f430 to
9e43569
Compare
This may be a bit problematic since same-namespace Honestly I have little experience with this reader in real world since it was always too magical. What I hate most on this reader is the fact that you can register multiple namespaces in which to look for annotations. Then comes the WTF - colliding names are ignored, first one is used: annotations/lib/Doctrine/Common/Annotations/DocParser.php Lines 699 to 706 in b4fde48 Do we have any statistics for SimpleAnnotationReader usage? Would this even be reason to keep it? |
Agreed. I'm mostly talking from experience in dealing with bug reports, where people type in something like This is the scope of my original question. On the other side, this is consistent with PHP's import and name resolution semantics, so I may just be going nowhere with it.
No more: burn it with napalm.
We made the mistake of basing the ORM simplistic setup on it (including many of the docs). The replacement of the annotations can be automated (or at least reported), so we don't need to concern ourselves with it. Let's just kill it. Given a |
Ocramius
left a comment
There was a problem hiding this comment.
I think there's nothing else to be done here, and this is good to go from my perspective 👍
|
Good thing is that it's already gone from ORM dev-master. 🎉 Following PHP resolution rules is the way to go, the least confusing one.
Sounds good to me. 👍 |
|
Alright, let's move on. |
closes #87