Fix arg=str() default argument #6193
Conversation
Fixes cython#6192. The basic issue is that the argument is deduced to be dynamic default argument, and then at a late stage is optimized into an empty string literal. This means that correct handling code for the literal is skipped because is has "is_dynamic" set, and the correct assignment code into the dynamic struct is skipped because it's a literal. I think the solution is to make the code that writes into the dynamic struct more tolerant of literals.
Isn't rather this the problem? Can we do this earlier somehow? It seems strange to do "oops, we're doing the wrong thing here so let's do something else" kind of operations at code generation time. |
That implies that either no optimisations should be made after the AnalyseExpressions stage. Or that annotations should be decided on much later (which may be true... Annotations are a mess and probably do need a clean-up). I don't think this PR is the most messy fix though. All it really does is just |
Fixes #6192.
The basic issue is that the argument is deduced to be dynamic default argument, and then at a late stage is optimized into an empty string literal.
This means that correct handling code for the literal is skipped because is has "is_dynamic" set, and the correct assignment code into the dynamic struct is skipped because it's a literal.
I think the solution is to make the code that writes into the dynamic struct more tolerant of literals.