Age | Commit message (Collapse) | Author | |
---|---|---|---|
2018-08-27 | Bugfix on bigendianess by casting size_t to unint32_t | Sakala Venkata Krishna Rohit | |
The reason for typecasting size_t to unint32_t is that size_t is 8 bytes and uint32_t is only 4 bytes. If not typecasted Memcpy fails to copy the *correct* four bytes in big endian platforms. | |||
2018-08-24 | Fix json parsing for timestamp (#115) | Paul Yang | |
* Fix json parsing for timestamp * Fix json encoding for timestamp * Implement timestamp_machine to parse timestamp * Remove comparision for size_t and 0 * Increase defined printer size * Increase size * Increase parser size * Mark json_en_timestamp_machine unused | |||
2018-08-22 | Moved upb_msg parts that depend on def to a separate msgfactory.{c,h}. | Josh Haberman | |
Also got rid of the premature "v1" business that was attempting to create a binary compatibility story. Also added an in-progress CMakeLists.txt file. | |||
2018-08-21 | Re-add upb/bindings/stdc++/string.h, which is still used in tests. | Josh Haberman | |
2018-08-21 | Removed a bunch of obsolete code. | Josh Haberman | |
A lot of this code was experimental or temporarily useful, but is no longer needed. | |||
2018-08-12 | Re-add msg handlers to the .c file also. | Joshua Haberman | |
2018-08-12 | Re-add message handlers to upb/handlers.*. | Joshua Haberman | |
These are still being used by the proto2 bindings. | |||
2018-08-12 | Enforced that upb_msg lives in an Arena only, and other simplifying. | Joshua Haberman | |
upb_msg was trying to be general enough that it could either live in an arena or be allocated with malloc()/free(). This was too much complexity for too little benefit. We should commit to just saying that upb_msg is arena-only. I also ripped out the code to glue upb_msg to the existing handlers-based encoder/decoder. upb_msg has its own, small, simple encoder/decoder. I'm trying to whittle down upb_msg to a small and simple core. I updated the Lua extension for these changes. Lua needs some more work to properly create arenas per message. For now I just created a single global arena. | |||
2018-08-09 | Use special eof char to mark end of parsing | Bo Yang | |
2018-08-09 | Move logic away from ragel machine | Bo Yang | |
2018-08-05 | Remove unused code | Bo Yang | |
2018-08-05 | Make value top level | Bo Yang | |
2018-08-01 | Mark unused variable | Bo Yang | |
2018-07-31 | Update ragel generated file | Bo Yang | |
2018-07-30 | Make top level wrapper value parsing work | Bo Yang | |
2018-07-27 | Fix json decoding for wrapper messages | Bo Yang | |
2018-06-27 | Enable ignore unknown field in json parsing | Bo Yang | |
2018-06-14 | Revert "Prepare upb_value for encoding/decoding map." | Bo Yang | |
This reverts commit f30dd0ff0cead9bb1f8f3b1b295392d75705486b. | |||
2018-03-11 | Prepare upb_value for encoding/decoding map. | Bo Yang | |
2018-03-08 | Implement decoding repeated message field. | Bo Yang | |
2018-03-08 | Bytes type should return size of stringview | Bo Yang | |
2018-03-08 | e->limit is invalidated after growing. | Bo Yang | |
2018-03-07 | Merge pull request #101 from TeBoring/php-change | Paul Yang | |
Fix parsing sub-message field. | |||
2018-03-07 | Merge pull request #102 from haberman/descriptor-in-core | Joshua Haberman | |
Added google/protobuf/descriptor.upb.* to core. | |||
2018-03-07 | Added google/protobuf/descriptor.upb.* to core. | Josh Haberman | |
This is in preparation for making upb_def capable of parsing binary descriptors directly. We leave upb/descriptor/descriptor.upbdefs.* in place for now, because upb/descriptor/reader.* still depends on it. Also removed a bit of cruft from the codegen. | |||
2018-03-07 | Fix parsing sub-message field. | Bo Yang | |
Buffer start place needs to be adjusted before parsing sub-message. | |||
2018-03-07 | swap condition branches to remove negation | Bo Yang | |
2018-03-07 | Fix oneof encoding/decoding | Bo Yang | |
2018-03-07 | In case of circular dependency, layout has to be inserted first. | Bo Yang | |
2018-03-07 | Field missing submsg and hasbit information. | Bo Yang | |
2018-03-07 | Conversion from uint64 to uint32 didn't consider sign bit. | Bo Yang | |
2018-03-07 | For encoding upb needs descriptor type instead of type. | Bo Yang | |
2017-09-25 | Modify TODO | Bo Yang | |
2017-09-25 | Change parameter type from enum opcode to int. | Bo Yang | |
If the compiler elects to represent enum E as a char rather than an int (per 6.7.2.2p4), the call to va_start() will have undefined behavior. | |||
2017-09-25 | Initialize tag to make compiler happy. | Bo Yang | |
2017-09-25 | Fixed JIT for unknown fields. (#1) | Joshua Haberman | |
2017-09-20 | Do not expose encode_unknown | Bo Yang | |
2017-09-19 | Remove upb_pb_encoder_encode_unknown | Bo Yang | |
2017-09-19 | Remove unused declaration. | Bo Yang | |
2017-09-19 | Remove upb_addunknown_handlerfunc | Bo Yang | |
2017-09-19 | Use upb_sink_putunknown for reserve unknown | Bo Yang | |
2017-09-19 | Remove upb_addunknown_handlerfunc and upb_handlers_setaddunknown | Bo Yang | |
2017-09-19 | Reserve unknown fields in upb | Bo Yang | |
1. For decoding, an unknownfields will be lazily created on message, which contains bytes of unknown fields. 2. For encoding, if the unknownfields is present on message, all bytes contained in it will be serialized. | |||
2017-07-19 | Fixed amalgamated build. | Joshua Haberman | |
2017-07-19 | Fix amalgamated build. | Joshua Haberman | |
2017-07-19 | Fixed 32-bit build and added generated files. | Joshua Haberman | |
2017-07-19 | Address review comments and fix compile warnings. | Joshua Haberman | |
2017-07-18 | Responded to PR comments. | Joshua Haberman | |
2017-07-18 | Fixes for oneof conformance tests. | Joshua Haberman | |
2017-07-18 | Fixed varint length when buffer is reallocated. | Joshua Haberman | |