| Commit message (Collapse) | Author | Age |
|---|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
* Add a travis stage that tests building Firestore with Xcode 8.3.
* Simulate on a device available in Xcode 8
* Fix compile errors under Xcode 8.3.3
* Remove Firestore_SwiftTests_iOS from the Firestore_Tests_iOS Scheme
I'll create a new target for this but in another PR.
* Add an entry to CHANGELOG.md
|
| |
|
|
|
|
|
|
|
|
|
| |
This is a pretty close port of `FSTExponentialBackoff`. The changes are pretty minor:
* delay is calculated using <chrono> duration types, not plain numbers, which should be a little more type-safe;
* split a piece of code into a ClampDelay function, because it's reasonably close to std::clamp;
* rephrased the class-level comment to make it clearer that the first attempt always has delay = 0;
* added simple tests (other platforms don't have tests for this).
Also make sure that canceling a DelayedOperation is always valid.
|
| |
|
| |
Were showing up in xcode build (but not cmake build)
|
| |\
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Also "fixed" BadFieldValueTagWithOtherValidTagsPresent test by changing
'false' to 'true'. Details: Depending on the version of nanopb, nanopb
would explicitly encode 'false', which shouldn't be done in proto3. When
it's explicitly encoded, the test worked properly. But when it was
(properly) dropped, the invalid tag is the only field that's actually
encoded, thus violating the assumptions of the test, leading to a test
failure. s/false/true fixes it, as now the boolean_value field is
(properly) encoded regardless of version.
|
| | |
| |
| |
| | |
(#1377)
|
| | | |
|
| | |
| |
| |
| |
| |
| |
| | |
Normally, this would be unexpected, as only a single entry in the Value
proto *should* be present. However, the proto docs state that parsers
should be able to handle repeated fields. (In the case of repeated
fields, the last one "wins".)
|
| | |
| |
| |
| | |
... where neither 'found' nor 'missing' fields set.
|
| |/ |
|
| |
|
|
|
|
| |
Note that it isn't possible to *serialize* NoDocuments.
Still TODO:
- Error handling
|
| |
|
| |
Roughly s/google::firebase::v1beta1/v1beta1/g
|
| |
|
|
|
|
|
| |
* [De]serialize non-empty Document instances
Still TODO:
- NoDocument
- ErrorHandling
|
| |
|
|
|
|
|
|
|
| |
* [De]serialize empty Document instances
Still TODO:
- non-empty
- NoDocument
- ErrorHandling
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
| |
This is more interesting than the serializing case, as we should expect
to see occasional corruption of our input byte vector.
|
| |
|
|
|
|
| |
Previously, the tests would compare serialization results against a
precomputed (via protoc) array of bytes. Now they serialize via our
nanopb based class and deserialize via libprotobuf (and vice versa) and
then ensure the result is the same as the input
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* Add a .clang-tidy configuration for Firestore C++
* Fix clang-tidy warnings
* typedef -> using
* const ref + rvalue ref -> pass by value
* NULL -> nullptr
* remove useless default initializations
* remove useless const value-type parameter declarations (definitions
can still use them)
* use auto instead of repeating types in a cast
* Fix typos
* Address use of static method through instance warnings
* Address use after move warnings
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* lint.sh now lints Objective-C++ too
* cpplint checks system-style includes that should be user-style
This prevents it from recognizing project sources as if they were C
system headers and then complaining that they're in the wrong place.
* cpplint checks #imports and #includes
* cpplint checks that C++ system headers aren't #imported
* cpplint checks for C system headers that could be C++ system headers
* cpplint checks that Objective-C sources include their headers
|
| |
|
|
|
|
|
|
| |
Deserializing not handled yet.
Note that the serializing case is fairly uninteresting, as assuming
valid input is passed in, there's no real reason why it should fail (and
if it does fail, it indicates a gross violation of our understanding of
the system.)
|
| | |
|
| |
|
|
|
|
|
| |
* Rename Writer::Encode* methods to Writer::Write*
* Rename: s/stream/writer/ (approximately)
but only where it applies to Writer (rather than pb_ostream_t).
|
| |
|
| |
These can (recursively) contain other FieldValues.
|
| | |
|
| |
|
|
|
|
|
|
| |
This will likely only apply for proto messages that use 'oneof's. (Because we need to serialize these manually.)
The problem was that as we were adding additional types, TypeValue was evolving into a parallel implementation of the FieldValue union.
When serializing/deserializing oneofs we need to supply a custom value to serialize and a custom function to do the work. There's no value in translating from FieldValue to TypeValue just to store as this custom object. We might as well just store the FieldValue model directly and write the custom serializer/deserializer to use the model directly.
|
| | |
|
| |
|
|
| |
Reduces noise while running valgrind so that I can see the leaks that
I'm introducing. :/
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* Build (grpc's) nanopb with -DPB_FIELD_16BIT
We require (at least) 16 bit fields. (By default, nanopb uses 8 bit
fields, ie allowing up to 256 field tags.)
Also note that this patch adds this to grpc's nanopb, rather than to our
nanopb. We'll need to eventually either:
a) we instruct grpc to use our nanopb
b) we rely on grpc's nanopb instead of using our own.
(^ marked as a TODO for now.)
* Add some dependant protos
Imported from protobuf. Nanopb requires these to be present (though
anything using libprotobuf does not, as these are already built into
that.)
* Add generated nanopb protos based off of newly added proto definitions
* Build the nanopb protos
* Serialize and deserialize null
|
| | |
|
| |
|
|
|
| |
Use remote/serializer placeholder class as a hook for the test to ensure
nanopb headers can be found, and test can be linked.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* Rewrite cc_test to take named arguments
Cut down on build file verbosity by having cc_test take SOURCES and
DEPENDS. The separate invocation of target_link_libraries is no longer
necessary.
* Add a cc_library rule to parallel cc_test
This cuts down on build file verbosity.
* Automatically add OBJC_FLAGS to cc_libraries if applicable
* Exclude platform-specific libraries from 'all'
This is makes it possible to declare this kind of library
unconditionally. Usage within a test or as a dependency will actually
trigger building.
* Restore secure_random_test.cc; clean-up comments
|
|
|
* Clean up quoting and other minor issues
* Reorganize CMake build output
Make it clearer which parts of the output pertain to external projects.
* Use a consistent ordering of ExternalProject arguments
* Prevent the top-level build from running in parallel
This prevents spurious failures when running make -j.
* Actually parse arguments in the xcodebuild function
* Use ExternalProject features when available
* submodule limits from CMake 3.0
* shallow clones from CMake 3.6
* git progress output from CMake 3.8
* Only build the parts of leveldb we need
Skip building the tools and other libraries
* Avoid installing ExternalProjects
Consume build output directly so that we can build just the targets we
need. Installing causes all targets to be built.
This doesn't matter as much for these targets but the gRPC build
includes a ton of stuff we don't need so it's worth adopting this as a
general strategy.
* Define an external build for grpc
* Test that grpc can link successfully.
* Add a FindGRPC CMake module
* Actually comment ExternalProjext_GitSource
|
Gil
Konstantin Varlamov
rsgowman