This change moves the logic for computing the package name part of an
import directory from python_rules.bzl to protobuf.bzl, and reinstates
the workspace root part of the directory. The effect is that it is now
possible for a py_library_rule to depend on a proto_library in a
different repository whose source files are generated.
This change replaces the simple `basename` logic with a more general
"remove prefix" logic. This allows py_proto_library to work even when
the input proto_library's sources are contained in strict
subdirectories of their package. (This situation is not uncommon when
using remote, non-Bazel repositories that only get a single BUILD file.)
* Add the daily xDS interop GCP resource cleanup script
* Add keep config feature && speed-up checking phase
* DO-NOT-SUBMIT: borrow Python test to ad-hoc run the new kokoro job
* Pin Python 3.6
* Kokoro is using pyenv to manage Python versions
* Install "dataclassses" package for 3.6.1
* Remove the latest 'test=' subprocess argument
* Revert "DO-NOT-SUBMIT: borrow Python test to ad-hoc run the new kokoro job"
This reverts commit 86d69dcb92.
* Clean the keep config
The original pipe implementation made a few assumptions:
* it should be used outside of arenas, implying allocation was expensive
* pipes would be fairly deep
With our expected implementation of the filter stack, pipes will max out at one to three stages, and so optimizations enabled by Filter seem to be actual pessimizations for our usual case.
Additionally we'll be operating inside of arenas and consequently memory allocation can be trivially cheap - and in doing so we can enable a far simpler implementation.
Co-authored-by: ctiller <ctiller@users.noreply.github.com>
* Revert "Revert "Replace work serializer with a mutex in c-ares resolver (#27858)" (#28324)"
This reverts commit b972b76816.
* Fix synchronization of grpc_ares_resolver_address
* WIP
* introduce XdsResourceType API and change Listener parsing to use it
* converted RouteConfig parsing
* convert cluster and endpoint parsing
* cleanup
* clang-format
* attempt to work around compiler problems
* move XdsResourceType to its own file, and move endpoint code out of XdsApi
* move cluster parsing to its own file
* move route config parsing to its own file
* move listener parsing to its own file
* clang-format
* minor cleanup
* plumbed XdsResourceType throughout XdsClient
* a bit of cleanup
* more cleanup
* construct full resource names before calling XdsApi::CreateAdsRequest()
* remove some unneeded code
* clean up includes and have XdsResourceType initialize the upb symtab
* more cleanup of unnecessary code
* more cleanup
* update comment
* clang-format
* add missing virtual dtor
* fix build
* remove resource-type-specific methods from XdsClient API
* have each resource type register itself upon instantiation
* remove comment
* add missing virtual dtor
* clang-format
* use a templated base class for XdsResourceType implementations
* clang-format
* Revert "Temporarily remove binder fuzzers it fails to build with old llvm (#27599)"
This reverts commit 6b922f871f.
* Migrate to protobuf based structured fuzzing
* Fix crash happening due to recent change
* Revert "Revert "[App Net] Switch Router to Mesh and Add unique string to Scope (#28145)" (#28176)"
This reverts commit cc968b2158.
* Allow scope to be None
* Revert "Revert "XdsClient: remove resource-type-specific methods from XdsClient API (#28231)" (#28301)"
This reverts commit 3e779b68fe.
* remove global resource type registry
* Add python3 to list of acceptable pythons in python wrapper
* Attempt to make fake server python3 compatible
* py3 fixes for test server
* more py3 fixes for test server
* require python3 for test/core/http/test_server.py
Co-authored-by: Alexander Polcyn <apolcyn@google.com>