IIUC this transforms a node representation of a demangled name to the equivalent string representation. It creates the string itself, and doesn't get the string as a parameter to write into. To me "transform in" was confusing, I think the natural language to use is "transform to".
initialization in-place on demand. Initialize parent metadata
references correctly on struct and enum metadata.
Also includes several minor improvements related to relative
pointers that I was using before deciding to simply switch the
parent reference to an absolute reference to get better access
patterns.
- Added missing ifdef guard in PointerIntEnum header
- Consistent naming convention for ifdef guards
- Consistent 'end namespace swift'
- Consistent single EOL at end of header files
Rename BlotMapVector's template typenames: MapTy -> MapT and VectorTy -> VectorT. This is consistent both with BlotMapVector's other template typename KeyT and ValueT, and with SmallBlotMapVector's MapT and VectorT template typenames.
If this option is enabled, when generating potential bindings for a type
variable, don't propagate IUO type. Instead try the optional type and
the underlying type. This way, untyped bindings will not be given IUO
type when they are initialized with exprs of IUO type.
Previously due to the way that ARC works, it was impossible to trigger any
memory safety issues. That being said the fact that the memory safety here is
non-obvious suggests that the right thing to do is just bite the bullet and
clear the ImmutablePointerSetFactory.
We do this by doing a traversal of our sorted lists in a similar manner as one
would when one is merging two such sets, i.e. one has two iterators and always
advances the iterator that has a value that is less than the other. If we ever
hit a situation where the two iterators equal, we must have a non-empty
intersection.
A unittest that exercises very basic functionality is provided as well.
Specifically this change:
1. Changed the input collection to be a const T & instead of T.
2. The return type of count_if is now properly the difference_type of the input range's iterators.
Introduces import-as-member (IAM) inferene system, to automatically
infer details that would otherwise have to be manually specified with
the swift_name attribute.
Basic functionality present, though there are some issues with
properties at the moment. No hooks, options, or tests yet, stay tuned.
This reverts commit 2262bd579a.
This information isn't necessary for field descriptor lookup,
after all. It's only the fields that need to have generic information,
which is already in the field descriptor.
Previously, the mangling didn't include generics, but these are
needed to key off of the new field descriptor metadata, as well
as to construct type references for the nominal type.
