Chiebot-Mirror/protobuf - protobuf - Gitea: Git with a cup of tea

Author	SHA1	Message	Date
Joshua Haberman	a5e337f88c	Allow dynamic tree shaking for groups, and disallow groups for map entry sub-messages. The initial motivation for this CL was to fix a bug found by fuzzing. But the fuzz bug pointed out a few edge cases that this CL corrects: 1. The core bug is that we were allowing a map entry sub-message to be linked to a group field. This is not allowed in protobuf schemas, but we did not check for this edge case in `upb_MiniTable_SetSubMessage()`, so we were de facto allowing it. This triggered some bad behavior in the parser whereby we pushed a limit without checking its validity first. 2. To defend against this, I added asserts in `upb_MiniTable_SetSubMessage()` to verify the type of the field we are linking, to ensure that a group field is not linked to a map entry sub-message. But this should probably be changed to return an error instead of relying on asserts for this. 3. I changed the fuzz util code that builds the MiniTable so that it will never violate this new invariant. The fuzz util code now can run into situations where a group field has no valid non-map-entry sub-message to select. In those cases it will simply not register any sub-message for that field. 4. Previously group did not support leaving sub-messages unregistered. Previously I added this feature for sub-messages but not for groups. There is no reason why dynamic tree shaking should not work for group fields, so I extended the feature to support groups also. PiperOrigin-RevId: 504913630	2 years ago
Joshua Haberman	7cd8f6c940	Ported more cases of wire format parsing to upb_WireReader. PiperOrigin-RevId: 498502557	2 years ago
Eric Salo	a9f9bfea75	fix some compiler warnings in ecis PiperOrigin-RevId: 498425946	2 years ago
Protobuf Team Bot	466408978f	Add extern C blocks to wire_reader and eps_copy_input_stream. PiperOrigin-RevId: 498262528	2 years ago
Joshua Haberman	112f037da6	Migrated the text format encoder to upb_WireReader. Also fixed a few bugs and added new functionality to upb_WireReader. PiperOrigin-RevId: 498223970	2 years ago
Joshua Haberman	75488a0742	Created a new upb_WireReader interface for parsing wire data directly. The overall motivation for this interface is to consolidate many places in upb that are parsing wire format data directly. This interface is not yet complete, but this is a good start. We have enough to port the wire format parsing in accessors.c to this interface. We can follow up by porting more places that do wire format parsing. PiperOrigin-RevId: 498109788	2 years ago
Joshua Haberman	a48af3f824	Moved aliasing logic for string field parsing into EpsCopyInputStream. Moving the logic down to EpsCopyInputStream makes it easier to test and reuse this functionality. We also implement aliasing for the final bytes of the patch buffer, which has never been supported before. We used to always force a copy for any data parsed out of the patch buffer at the end of the stream. Much of this logic is ported directly from the C++ EpsCopyInputStream class. PiperOrigin-RevId: 498091644	2 years ago
Joshua Haberman	4a581e5cd0	Fixed Lexan builds that were broken by cl/494400285. PiperOrigin-RevId: 494428378	2 years ago
Joshua Haberman	68d1d91475	Separated out buffering code into upb_EpsCopyInputStream. This mirrors the structure of C++ protobuf, which has an EpsCopyInputStream class. This will lay the foundation for making EpsCopyInputStream capable of true streaming, by reading its input from a ZeroCopyInputStream. It also lets us test EpsCopyInputStream separately from the decoder: see the new unit test that fuzzes upb_EpsCopyInputStream. After this CL is submitted, the two decoders (the normal decoder and the fast decoder) should no longer be accessing the members of upb_EpsCopyInputStream. PiperOrigin-RevId: 494400285	2 years ago