Age | Commit message (Collapse) | Author | |
---|---|---|---|
2014-12-05 | Update JSON parser to emit UTF8 to string fields appropriately. | Chris Fallin | |
2014-12-04 | Synced to post-code-review JSON printer. | Chris Fallin | |
2014-12-04 | Merge pull request #5 from cfallin/master | Joshua Haberman | |
JSON printer: remove dependency on YAJL, using a hand-rolled printer instead. | |||
2014-12-04 | JSON printer: remove dependency on YAJL, using a hand-rolled printer instead. | Chris Fallin | |
2014-12-03 | Sync from Google-internal development. | Josh Haberman | |
Most notably, a JSON parser is now provided. | |||
2014-11-18 | Fixed building of googlepbtests. | Josh Haberman | |
2014-11-18 | Sync to Google-internal development. | Josh Haberman | |
2014-09-01 | I think I have the right 32-bit C++ libraries now. | Josh Haberman | |
2014-09-01 | Maybe I finally got the right C++ library for 32-bit? | Josh Haberman | |
2014-09-01 | 32-bit builds also needs 32-bit C++ library. | Josh Haberman | |
2014-09-01 | 32-bit build needs 32-bit libraries installed. | Josh Haberman | |
2014-09-01 | Added lua and core32 Travis builds, and rewrote README.md | Josh Haberman | |
2014-09-01 | Added compiled version of test.proto. | Josh Haberman | |
This avoids requiring protoc to be installed just to build/run the basic tests. When upb has its own .proto file parser we should be able to remove this precompiled version from the repository. | |||
2014-09-01 | Enable more Travis configurations. | Josh Haberman | |
2014-08-31 | Rewrite of build system. | Josh Haberman | |
Notable changes: - We now only build things by default that require no dependencies. So you can build upb even if you don't have Lua or Google protobuf installed. - Checked in a pre-built version of the JIT, so you don't need Lua installed at build time to run DynASM. It will still notice if you change the .dasc file and attempt to re-run DynASM in that case. - The build system now builds all modules of upb into separate libraries, reflecting the modularity that is already inherent in upb's design. This should make it easier to trim the fat. - removed the GDB JIT interface. I wasn't using it much; using a .so is easier and more robust. | |||
2014-08-30 | Fix in test_cpp for bad macro expansion. | Josh Haberman | |
2014-08-30 | Fixed build and added Travis CI support. | Josh Haberman | |
2014-08-09 | Added optimization note. | Josh Haberman | |
2014-08-09 | Make the absence of perf-cppflags give a good default build. | Josh Haberman | |
Defaults are now: - thread-safe with GCC/Clang - Debugging not enabled (enable with -UNDEBUG) | |||
2014-08-09 | Fixed some bugs and added accessors. | Josh Haberman | |
2014-08-01 | Beginnings of a prototype Ruby extension. | Josh Haberman | |
2014-07-29 | Added UPB_UNTRACKED_REF and some more handler overload options. | Josh Haberman | |
2014-07-24 | Sync from Google-internal development. | Josh Haberman | |
2014-07-02 | Sync from internal Google development. | Josh Haberman | |
2014-06-26 | Sync from internal Google development. | Josh Haberman | |
2014-02-04 | Sync from Google development. | Josh Haberman | |
2014-01-21 | Sync to internal Google development. | Josh Haberman | |
2013-12-20 | Sync from Google-internal development. | Josh Haberman | |
2013-10-24 | Build fix for Mac OS X. | Josh Haberman | |
2013-10-24 | Fixes to make the open-source build compile on Linux. | Josh Haberman | |
2013-10-24 | Merge from Google-internal development: | Josh Haberman | |
- rewritten decoder; interpreted decoder is bytecode-based, JIT decoder no longer falls back to the interpreter. - C++ improvements: C++11-compatible iterators, upb::reffed_ptr for RAII refcounting, better upcast/downcast support. - removed the gross upb_value abstraction from public upb.h. | |||
2013-05-28 | Fix to compile on Linux. | Josh Haberman | |
2013-05-28 | Fixes to compile on GCC 4.2, as shipped with XCode. | Josh Haberman | |
2013-05-28 | Merge from Google-internal development. | Josh Haberman | |
- Better error reporting for upb::Def setters. - error reporting for upb::Handlers setters. - made the start/endmsg handlers a little less special-cased. | |||
2013-05-25 | Updated benchmarks to new API. | Josh Haberman | |
2013-05-25 | Merge branch 'master' of github.com:haberman/upb | Josh Haberman | |
Conflicts: upb/google/proto2.cc upb/handlers.h | |||
2013-05-25 | Synced with Google-internal development. | Josh Haberman | |
C++ handlers are now type-safe; SinkFrame is gone. Various other changes. | |||
2013-05-11 | Updated benchmarks to new APIs. | Joshua Haberman | |
2013-05-11 | Open source fixes: builds on OS X again. | Joshua Haberman | |
2013-05-11 | Synced with 3 months of Google-internal development. | Josh Haberman | |
Major changes: - Got rid of all bytestream interfaces in favor of using regular handlers. - new Pipeline object represents a upb pipeline, does bump allocation internally to manage memory. - proto2 support now can handle extensions. | |||
2013-02-15 | Sync with 8 months of Google-internal development. | Josh Haberman | |
Many things have changed and been simplified. The memory-management story for upb_def and upb_handlers is much more robust; upb_def and upb_handlers should be fairly stable interfaces now. There is still much work to do for the runtime component (upb_sink). | |||
2012-03-31 | Properly export setjmp/longjmp. | Joshua Haberman | |
2012-03-31 | Fix broken setjmp with working one. | Joshua Haberman | |
2012-03-31 | Got decoder & textprinter compiling in kernel mode. | Joshua Haberman | |
2012-03-31 | Sync from internal Google development. | Joshua Haberman | |
2012-03-24 | Got our performance back by tweaking AssignToString. | Josh Haberman | |
plain.parsetoproto2_googlemessage1.upb_jit: 254 -> 324 (27.56) omitfp.parsetoproto2_googlemessage1.upb_jit: 220 -> 332 (50.91) | |||
2012-03-24 | Fix duplicate symbol on OS X. | Josh Haberman | |
2012-03-24 | Updated DynASM from upstream (Lua 5.2 support). | Josh Haberman | |
2012-03-24 | Sync from internal Google development. | Joshua Haberman | |
Many improvements, too many to mention. One significant perf regression warrants investigation: omitfp.parsetoproto2_googlemessage1.upb_jit: 343 -> 252 (-26.53) plain.parsetoproto2_googlemessage1.upb_jit: 334 -> 251 (-24.85) 25% regression for this benchmark is bad, but since I don't think there's any fundamental design issue that caused it I'm going to go ahead with the commit anyway. Can investigate and fix later. Other benchmarks were neutral or showed slight improvement. | |||
2011-12-23 | Fixes to un-break "make descriptorgen" | Joshua Haberman | |