Run the encode/decode tests as benchmarks. These are trivial microbenchmarks
and not representative of real-world usage, but serve as a useful overview
of the relative cost of various operations and are handy for profiling
specific narrow scenarios.
When run with the -v1 flag, benchmark the v1 implementation for comparison.
Change-Id: I469dd6759bef50e2bd039327095f82d29d70b8fc
Reviewed-on: https://go-review.googlesource.com/c/protobuf/+/171120
Reviewed-by: Joe Tsai <thebrokentoaster@gmail.com>
Using an option instead of a separate method has several useful properties:
It makes it explicit whether the fast-path AppendMarshal is expected to use
cached sizes or not.
It properly plumbs the decision to use cached sizes through the call stack.
Consider the case where message A includes B includes C: If A and C support
cached sizes but B does not, we would like to use the size cache in all
messages which support it. Placing this decision in the options allows this
to work properly with no additional effort.
Placing this option in MarshalOptions permits users to request use of
existing cached sizes. This is a two-edged sword: There are places where
this ability can permit substantial efficiencies, but this is also an
exceedingly sharp-edged API. I believe that on balance the benefits
outweigh the risks, especially since the prerequisites for using
cached sizes are intuitively obvious. (You must have called Size, and
you must not have changed the message.)
This CL adds a Size method to MarshalOptions, rather than adding a SizeOptions
type. Future additions to MarshalOptions may affect the size of the encoded
output (e.g., an option to skip encoding unknown fields) and using the same
options for both Marshal and Size makes it easier to use a consistent
configuration for each.
Change-Id: I6adbb55b717dd03d39f067e1d0b7381945000976
Reviewed-on: https://go-review.googlesource.com/c/protobuf/+/171119
Reviewed-by: Joe Tsai <thebrokentoaster@gmail.com>
Was overwriting the output buffer rather than appending to it.
Change-Id: I6ffb72a440f464f4259cfebc42c1dc75b73fb5ae
Reviewed-on: https://go-review.googlesource.com/c/protobuf/+/171117
Reviewed-by: Joe Tsai <thebrokentoaster@gmail.com>
Remove the Reflection field from MarshalOptions and UnmarshalOptions.
Disable the fast path and use the reflection-based implementation when
the 'protoreflect' build tag is set.
Change-Id: Ic674e3af67501de27fb03ec2712fbed40eae7fef
Reviewed-on: https://go-review.googlesource.com/c/protobuf/+/170896
Reviewed-by: Joe Tsai <thebrokentoaster@gmail.com>
Move all checks for required fields into a proto.IsInitialized function.
Initial testing makes me confident that we can provide a fast-path
implementation of IsInitialized which will perform more than
acceptably. (In the degenerate-but-common case where a message
transitively contains no required fields, this check can be nearly
zero cost.)
Unifying checks into a single function provides consistent behavior
between the wire, text, and json codecs.
Performing the check after decoding eliminates the wire decoder bug
where a split message is incorrectly seen as missing required fields.
Performing the check after decoding also provides consistent and
arguably more correct behavior when the target message was partially
prepopulated.
Change-Id: I9478b7bebb263af00c0d9f66a1f26e31ff553522
Reviewed-on: https://go-review.googlesource.com/c/protobuf/+/170787
Reviewed-by: Herbie Ong <herbie@google.com>
Allow message implementations to provide optimized versions of standard
operations. Generated messages now include a ProtoReflectMethods method,
returning a protoiface.Methods struct containing pointers to assorted
optional functions.
The Methods struct also includes a Flags field indicating support for
optional features such as deterministic marshaling.
Implementation of the fast paths (and tests) will come in later CLs.
Change-Id: Idd1beed0ecf43ec5e5e7b8da2ee1e08d3ce32213
Reviewed-on: https://go-review.googlesource.com/c/protobuf/+/170340
Reviewed-by: Joe Tsai <thebrokentoaster@gmail.com>
Create a single binary for handling generation of protos.
This replaces previous logic spread throughout the repo in:
* regenerate.bash
* cmd/protoc-gen-go/golden_test.go
* cmd/protoc-gen-go-grpc/golden_test.go
* (indirectly) internal/protogen/goldentest
One of the problems with the former approaches is that they relied on
a version of protoc that was specific to a developer's workstation.
This meant that the result of generation was not hermetic.
To address this, we rely on the hard-coded version of protobuf specified
in the test.bash script.
A summary of changes in this CL are:
* The internal_gengo.GenerateFile and internal_gengogrpc.GenerateFile
functions are unified to have consistent signatures. It seems that the
former accepted a *protogen.GeneratedFile to support v1 where gRPC code
was generated into the same file as the base .pb.go file. However, the
same functionality can be achieved by having the function return
the generated file object.
* The test.bash script patches the protobuf toolchain to have properly
specified go_package options in each proto source file.
* The test.bash script accepts a "-regenerate" argument.
* Add generation for the well-known types. Contrary to how these were
laid out in the v1 repo, all the well-known types are placed in the
same Go package.
* Add generation for the conformance proto.
* Remove regenerate.bash
* Remove internal/protogen
* Remove cmd/protoc-gen-go/golden_test.go
* Remove cmd/protoc-gen-go-grpc/golden_test.go
* Add cmd/protoc-gen-go/annotation_test.go
Change-Id: I4a1a97ae6f66e2fabcf4e4d292c95ab2a2db0248
Reviewed-on: https://go-review.googlesource.com/c/164477
Reviewed-by: Damien Neil <dneil@google.com>
Most usages of New actually prefer to interact with the reflective view
rather than the native Go type. Thus, change New to return that instead.
This parallels reflect.New, which returns the reflective view
(i.e., reflect.Value) instead of native type (i.e., interface{}).
We make the equivalent change to KnownFields.NewMessage, List.NewMessage,
and Map.NewMessage for consistency.
Since this is a subtle change where the type system will not always
catch the changed type, this change was made by both changing the type
and renaming the function to NewXXX and manually looking at every usage
of the the function to ensure that the usage correctly operates
on either the native Go type or the reflective view of the type.
After the entire codebase was cleaned up, a rename was performed to convert
NewXXX back to New.
Change-Id: I153fef627b4bf0a427e4039ce0aaec52e20c7950
Reviewed-on: https://go-review.googlesource.com/c/157077
Reviewed-by: Damien Neil <dneil@google.com>
If a field has an ExtensionType defined, use it in decoding.
Change-Id: I85f3da0f52a11578500cf28e4611fa4eb31f0623
Reviewed-on: https://go-review.googlesource.com/c/154581
Reviewed-by: Joe Tsai <thebrokentoaster@gmail.com>
The v2 decoder isn't 100% complete, but it's good enough.
Delete the vendored copy of the v1 Unmarshal implementation.
Change-Id: Ibeabbb2e9109a1ec3df57e71f98b7aa4a583fc5b
Reviewed-on: https://go-review.googlesource.com/c/154577
Reviewed-by: Joe Tsai <thebrokentoaster@gmail.com>
Add proto.Unmarshal.
Test cases all produce identical results to the v1 unmarshaller.
Change-Id: I42259266018a14e88a650c5d83a043cb17a3a15d
Reviewed-on: https://go-review.googlesource.com/c/153918
Reviewed-by: Joe Tsai <thebrokentoaster@gmail.com>
This initial implementation covers marshaling Message without use
of extensions, Any expansion, weak yet.
Change-Id: Ic787939c1d2a4e70e40c3a1654c6e7073052b7d3
Reviewed-on: https://go-review.googlesource.com/c/151677
Reviewed-by: Joe Tsai <thebrokentoaster@gmail.com>
