Swift importing ObjC drops methods/properties if the type is only a forward
declaration since the type is incomplete. Historically the generator has always
use forward declarations to reduce how much will have rebuild when a proto file
does change; but that puts it at odds with Swift. If ObjC Protos end up spanning
Swift modules, the Swift import behavior could become a problem; so this option
provides a control for the behavior. The current behavior is to continue forward
declarations, but eventually the default will be changed.
Generate the WKTs using imports instead of forward decls.
Overdue followup to https://github.com/protocolbuffers/protobuf/pull/7173
Since all the files are at the same level as the runtime headers, there is no
need for things to be framework based imports, they should all just work like
the other headers do.
- Directly generate the bundled header imports into the preamble section when
generating for a bundled proto.
- Update the preamble generation to skip the CPP wrapper when generating for a
bundled proto file.
- Regenerate the WKTs.
- Update GPBProtocolBuffer.h/GPBWellKnownTypes.h to also skip the CPP wrapping.
GPB_USE_PROTOBUF_FRAMEWORK_IMPORTS in the podspec and non bundled files still
has to exist because that comes into play for those files to find the runtime
headers.
* Fix comment syntax in any.proto
Remove extra spaces which cause incorrect indentation in godoc.
Remove the "====" style title which is not rendered by godoc.
* Run ./generate_descriptor_proto.sh
Since the keys are `Class`-s, there's no need to hash/copy/etc. them. This
avoids causing `+initialize` on the classes just when building up a registry.
- Update semaphore comment to new scope.
- Use an atomic swap to avoid needing to use the semaphore.
This means the semaphore is create only when extension are auto created (less
memory usage).
- Move the min up to iOS 9. Things should still work with 8, just can't have
the project set to that and be able to run the tests, so moving it up to
support running the tests.
- Remove the bitcode setting from the unittest bundle, was causing an error
with Xcode 12.
* Fix a typo
* Fix lots of spelling errors
* Fix a few more spelling mistakes
* s/parsable/parseable/
* Don't touch the third party files
* Cloneable is the preferred C# term
* Copyable is the preferred C++ term
* Revert "s/parsable/parseable/"
This reverts commit 534ecf7675.
* Revert unparseable->unparsable corrections
Removing language-specific targets from the top-level BUILD file
will allow users to keep their workspaces smaller and easier to
maintain by not depending on language rules they don't need.
Similar work was done for Java in #7190.
Option to add a prefix to generated #imports incase ObjC Protos are
used in a build system where one wants to avoid adding a header
search path and have more complete imports.
This was causing problems in OCMock due to recursion when OCMock was trying
to mock a protobuf.
`_forwardStackInvocation:` was being called as part of the resolution of `descriptor`.
- Bump the version marker in the generated code.
- Set the flag to say clear on zero is known.
- Set clear on zero for proto3 optional fields that aren't tagged as optional.
- Also tweak the call from some C generated code to use different apis with
better validation.
- Mark the ObjC generator as supporting the proto3 optional feature.
- Regenerate the WKTs to get the new flags.
- With the new flags, the syntax isn't needed, so don't pass it.
- To keep library binary compatibility, rename the helpers since one
"Internal" method was called from generated code. Keep a shim with
the old signature to support those generated sources.
- Add a Descriptor flag to capture if the field should clear on being zeroed.
- Update the runtime to use the new clear on zero flag.
- Add a flag on message initialization to indicate the sources were generated
with/without this support so the runtime can backfill the flag for older
generated sources.
- Don't make an OneofGenerator for synthetic oneofs.
- Update the field calculations that determine if hasbits are needed to
know about synthetic oneofs and ignore them.
