When running >bazel test :protobuf_test a few compiler warnings came up. They were centered around the usage of "int" as the loop counter where the value should have been a size_t.
ARM specific CONFIG_KUSER_HELPERS feature can be disabled in Linux kernel,
and in this case, we shouldn't crash. Use gcc built-in functions instead
of arm specific code if they are available.
There seem to already be .proto files out there that have conflicting
enum names, which will not be able to build successfully for some
languages (like C#). To prevent this problem from spreading, let's make
it an error for proto3 but just issue a warning for proto2. This fixes
issue #2179.
A bug was causing generated JSPB code with CommonJS-style imports to
refer incorrectly to nested messages from other .proto files. The
generated code would have things like "test_pb.InnerMessage" instead of
"test_pb.OuterMessage.InnerMessage". This commit fixes the problem by
correctly taking into account any message nesting.
Currently deserialization of a non-packed binary repeated field is quadratic in
the number of elements, because each time we parse a new element we copy over
all elements we have parsed so far. This CL fixes the performance problem by
having the generated deserialization code just call addX() instead of using
getX() and setX().
There seem to already be .proto files out there that have conflicting
enum names, which will not be able to build successfully for some
languages (like C#). To prevent this problem from spreading, let's make
it an error for proto3 but just issue a warning for proto2. This fixes
issue #2179.
A bug was causing generated JSPB code with CommonJS-style imports to
refer incorrectly to nested messages from other .proto files. The
generated code would have things like "test_pb.InnerMessage" instead of
"test_pb.OuterMessage.InnerMessage". This commit fixes the problem by
correctly taking into account any message nesting.
Currently deserialization of a non-packed binary repeated field is quadratic in
the number of elements, because each time we parse a new element we copy over
all elements we have parsed so far. This CL fixes the performance problem by
having the generated deserialization code just call addX() instead of using
getX() and setX().
- Capture the version used to generated.
- Check at compile time and runtime that generated code isn't from a newer
version, also check that the min version required is also supported.
- Keep the old constants/macros/functions to special case the last version
that was working so those generated sources still work until we decide
otherwise.
This pull request includes two implementation: C extension and PHP
package. Both implementations support encode/decode of singular,
repeated and map fields.
- Capture the ObjC prefix used when generating the the file.
- Track the containing type on descriptors.
- Mark descriptors where the message class name got a suffix added to it.
- Expose a fullName property on Descriptors.
- Add helpers for packing/unpacking Any messages.
- Bump the ObjC runtime version number. Since we added methods and invoke them
in the generated code, ensure the code is running against a matching version.
Otherwise, someone could compile against headers, but run with a framework
that is older and get unknown selector failures. This should trip clearer
messaging.
Fixes https://github.com/google/protobuf/issues/1674
At generation time, walk the file's dependencies to see what really contains
extensions so we can generate more minimal code that only links together the
roots that provided extensions. Gets a bunch of otherwise noop code out of
the call flow when the roots are +initialized.
Michael Avrukin
Julien Brianceau
ramrunner
Jisi Liu
Adam Cozzette
Feng Xiao
Bo Yang
Adam Cozzette
Andy Hochhaus
Derek Murray
Sergio Campama
Nickolay Mazurkin
Thomas Van Lenten
Paul Yang
Guo Xiao
Khing
Chih-Hung Hsieh
mike07026