1. Fix C++ tests.
* Add missing files to Makefile.am and fix distcheck in tests.sh
* Remove BUILT_SOURCES from conformance/Makefile.am.
* Add some missing override keyword.
* Add a type cast to int64 because our StrCat() in stubs can't handle size_t.
2. Fix Java tests.
* Add missing test dependency on guava in pom.xml.
* Include newly referenced test data in test resources.
* Manually fix map_lite_test.proto which is overwritten because it's mapped
from map_test.proto in google3.
* Add back "optimize_for = LITE_RUNTIME" which is still needed to keep the
opensource test passing as it's still running lite tests.
* Add a type cast in newBuilder() because without it the code doesn't compile
with openjdk javac 1.8 (the compiler can't figure if it's the right type
due to complex generic typing).
3. Fix Python tests.
* Remove/replace references to <hash_map>.
* Suppress more warnings in setup.py.
* Replace incorrect header inclusion for google/protobuf/proto_api.h.
* Add strings::EndsWith to google/protobuf/stubs/strutil.h because it's
referenced in the updated python C extension code.
* Replace proto2 with google::protobuf. The proto2 name is leaked to
opensource because we removed the subsitition rule for proto2 namespace
but only fixed C++ source code and forgot to update python C extension code.
Likely to be java language issue. Varargs are considered arrays, thus
using generic with varargs will cause unchecked warning about generic
array creation.
* Down-integrate internal changes to github.
* Update conformance test failure list.
* Explicitly import used class in nano test to avoid random test fail.
* Update _GNUC_VER to use the correct implementation of atomic operation
on Mac.
* maps_test.js: check whether Symbol is defined before using it (#2524)
Symbol is not yet available on older versions of Node.js and so this
test fails with them. This change just directly checks whether Symbol is
available before we try to use it.
* Added well_known_types_embed.cc to CLEANFILES so that it gets cleaned up
* Updated Makefile.am to fix out-of-tree builds
* Added Bazel genrule for generating well_known_types_embed.cc
In pull request #2517 I made this change for the CMake and autotools
builds but forgot to do it for the Bazel build.
* Update _GNUC_VER to use the correct implementation of atomic operation on Mac.
* Add new js file in extra dist.
* Bump version number to 3.2.0
* Fixed issue with autoloading - Invalid paths (#2538)
* PHP fix int64 decoding (#2516)
* fix int64 decoding
* fix int64 decoding + tests
* Fix int64 decoding on 32-bit machines.
* Fix warning in compiler/js/embed.cc
embed.cc: In function ‘std::string CEscape(const string&)’:
embed.cc:51:32: warning: comparison between signed and unsigned integer expressions [-Wsign-compare]
for (int i = 0; i < str.size(); ++i) {
^
* Fix include in auto-generated well_known_types_embed.cc
Restore include style fix (e3da722) that has been trampled by
auto-generation of well_known_types_embed.cc
* Fixed cross compilations with the Autotools build
Pull request #2517 caused cross compilations to start failing, because
the js_embed binary was being built to run on the target platform
instead of on the build machine. This change updates the Autotools build
to use the AX_PROG_CXX_FOR_BUILD macro to find a suitable compiler for
the build machine and always use that when building js_embed.
* Minor fix for autocreated object repeated fields and maps.
- If setting/clearing a repeated field/map that was objects, check the class
before checking the autocreator.
- Just to be paranoid, don’t mutate within copy/mutableCopy for the autocreated
classes to ensure there is less chance of issues if someone does something
really crazy threading wise.
- Some more tests for the internal AutocreatedArray/AutocreatedDictionary
classes to ensure things are working as expected.
- Add Xcode 8.2 to the full_mac_build.sh supported list.
* Fix generation of extending nested messages in JavaScript (#2439)
* Fix generation of extending nested messages in JavaScript
* Added missing test8.proto to build
* Fix generated code when there is no namespace but there is enum definition.
* Decoding unknown field should succeed.
* Add embed.cc in src/Makefile.am to fix dist check.
* Fixed "make distcheck" for the Autotools build
To make the test pass I needed to fix out-of-tree builds and update
EXTRA_DIST and CLEANFILES.
* Remove redundent embed.cc from src/Makefile.am
* Update version number to 3.2.0-rc.1 (#2578)
* Change protoc-artifacts version to 3.2.0-rc.1
* Update version number to 3.2.0rc2
* Update change logs for 3.2.0 release.
* Update php README
* Update upb, fixes some bugs (including a hash table problem). (#2611)
* Update upb, fixes some bugs (including a hash table problem).
* Ruby: added a test for the previous hash table corruption.
Verified that this triggers the bug in the currently released
version.
* Ruby: bugfix for SEGV.
* Ruby: removed old code for dup'ing defs.
* Reverting deployment target to 7.0 (#2618)
The Protobuf library doesn’t require the 7.1 deployment target so
reverting it back to 7.0
* Fix typo that breaks builds on big-endian (#2632)
* Bump version number to 3.2.0
Integer.MAX_SIZE (0x7FFFFFF) #2228
M java/core/src/main/java/com/google/protobuf/CodedInputStream.java
Set DEFAULT_SIZE_LIMIT to Integer.MAX_SIZE (Was 64MB). This is how it was
in pre-2.7.0 pb. Changed size check to an overflow-conscious test (as it
is later in tryRefillBuffer (making sizeLimit a long was to disruptive).
M java/core/src/test/java/com/google/protobuf/CodedInputStreamTest.java
Add two tests that echo tests recently added over in c++ to test parse
of message sizes that are approach and are beyond the size limit.
Currently some public API methods are defined in GenreatedMessage.java
and they have a generric return type:
class GeneratedMessage {
class Builder<BuilderType extends Builder<BuilderType>> {
public BuilderType setField(...);
public BuilderType setExtension(...);
}
}
With these definitions, the compiled byte code of a callsite will have
a direct reference to GeneratedMessage. For example:
fooBuilder.setField(...);
becomes:
##: invokevirtual // Method Builder.setField:(...)LGeneratedMessage.Builder
##: checkcast // class Builder
This will prevent us from updating generated classes to subclass a
different versioned GeneratedMessageV3 class in the future (we can't do
it in a binary compatible way).
This change addresses the problem by overriding these methods directly
in the generated class:
class Foo {
class Builder extends GeneratedMessage.Builder<Builder> {
public Builder setField(...) {
return super.setField(...);
}
}
}
After this, fooBuilder.setField(...) will be compiled to:
##: invokevirtual // Method Builder.setField:(...)LFoo.Builder
The callsites will no longer reference GeneratedMessage directly and we
can change Foo to subclass GeneratedMessageV3 without breaking binary
compatiblity.
The downside of this change is:
1. It increases generated code size (though it saves some instructions
on the callsites).
2. We can never stop generating these overrides because doing that
will break binary compatibility.
Change-Id: I879afbbc1325a66324a51565e017143489b06e97
Adam Cozzette
Adrián Nieto Rodríguez
Josh Haberman
Feng Xiao
Jonathan Haber
Anuraag Agrawal
Jisi Liu
alanbur
Anuraag Agrawal
Paul Yang
goorov
Michael Stack
Time
nmittler
Bo Yang
Patrick Strawderman
Otto Kekäläinen