Ensure that the init funcs for files within a Go package run in the
dependency order of the source .proto files. That is, if a.proto and b.proto
are part of the same Go package, and a.proto imports b.proto, then b.pb.go's
init funcs must run before a.pb.go's.
Change-Id: I0e86ff22e5c4cab9df7a73fe4805390fadd34b0d
Reviewed-on: https://go-review.googlesource.com/c/protobuf/+/166419
Reviewed-by: Joe Tsai <thebrokentoaster@gmail.com>
Reviewed-by: Herbie Ong <herbie@google.com>
Using arrays in the generated reflection information adds unnecessary
eq and hash functions being added to the package. Change to slices
to reduce bloat.
Change-Id: I1a4f6d59021644d93dd6c24679b9233141e89a75
Reviewed-on: https://go-review.googlesource.com/c/164640
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>
Use the full path (including the extension) for the generation of
the per-file variable name. Several reasons for this:
* The current logic is buggy in the case where pathType == pathTypeImport
since the prefix variable will be mangled with the Go import path.
* The extension is technically part of the path.
Thus, "path/to/foo.proto" and "path/to/foo.protodevel" are two
distinctly different imports.
* Style-wise, it subjectively looks better. Rather than being a mixture
of camelCase and snake_case, it is all snake_case for the common case:
before: ProtoFile_google_protobuf_any
after: File_google_protobuf_any_proto
* Since the extension is almost always ".proto", this results in a
suffix of "_proto", which provides an additional layer of protection
against possible name conflicts. The previous approach could possibly
have a conflict between "Foo.proto" and a message named ProtoFile
with a sub-message called Foo.
Also, use the per-file variable name for the raw descriptor variables
instead of the hashed version.
Change-Id: Ic91e326b7593e5985cee6ececc60539c27fe32fe
Reviewed-on: https://go-review.googlesource.com/c/164379
Reviewed-by: Damien Neil <dneil@google.com>
Calling a helper function directly should reduce binary bloat slightly.
Change-Id: I6068dc4cd00c8d90d2e6e6d99633b81388bc8781
Reviewed-on: https://go-review.googlesource.com/c/164679
Reviewed-by: Damien Neil <dneil@google.com>
For a file "foo/bar.proto", put the FileDescriptor in "ProtoFile_foo_bar"
rather than "Bar_fileDescriptor".
Avoid name clashes when a package contains "a/foo.proto" and "b/foo.proto".
Don't camelcase the filename: These vars weren't fully camelcased to begin
with, and leaving the filename relatively unchanged is clearer and more
predictable.
Move "ProtoFile" from the end of the var name to the start, so that vars
will sort better in packages with multiple descriptors.
These changes do add a chance of name collision when the input filename
begins with an uppercase letter: Foo.proto becomes "ProtoFile_Foo", which
could be the result of camelcasing "proto_file.foo". The readability
benefits seem worth it.
Change-Id: If27d3a0d7b5bf3535aa1607a8579eb057c74d2dc
Reviewed-on: https://go-review.googlesource.com/c/163199
Reviewed-by: Joe Tsai <thebrokentoaster@gmail.com>
Reviewed-by: Herbie Ong <herbie@google.com>
This CL takes a significantly different approach to generating support
for protobuf reflection. The previous approach involved generating a
large number of Go literals to represent the reflection information.
While that approach was correct, it resulted in too much binary bloat.
The approach taken here initializes the reflection information from
the raw descriptor proto, which is a relatively dense representation
of the protobuf reflection information. In order to keep initialization
cost low, several measures were taken:
* At program init, the bare minimum is parsed in order to initialize
naming information for enums, messages, extensions, and services declared
in the file. This is done because those top-level declarations are often
relevant for registration.
* Only upon first are most of the other data structures for protobuf
reflection actually initialized.
* Instead of using proto.Unmarshal, a hand-written unmarshaler is used.
This allows us to avoid a dependendency on the descriptor proto and also
because the API for the descriptor proto is fundamentally non-performant
since it requires an allocation for every primitive field.
At a high-level, the new implementation lives in internal/fileinit.
Several changes were made to other parts of the repository:
* cmd/protoc-gen-go:
* Stop compressing the raw descriptors. While compression does reduce
the size of the descriptors by approximately 2x, it is a pre-mature
optimization since the descriptors themselves are around 1% of the total
binary bloat that is due to generated protobufs.
* Seeding protobuf reflection from the raw descriptor significantly
simplifies the generator implementation since it is no longer responsible
for constructing a tree of Go literals to represent the same information.
* We remove the generation of the shadow types and instead call
protoimpl.MessageType.MessageOf. Unfortunately, this incurs an allocation
for every call to ProtoReflect since we need to allocate a tuple that wraps
a pointer to the message value, and a pointer to message type.
* internal/impl:
* We add a MessageType.GoType field and make it required that it is
set prior to first use. This is done so that we can avoid calling
MessageType.init except for when it is actually needed. The allows code
to call (*FooMessage)(nil).ProtoReflect().Type() without fearing that the
init code will run, possibly triggering a recursive deadlock (where the
init code depends on getting the Type of some dependency which may be
declared within the same file).
* internal/cmd/generate-types:
* The code to generate reflect/prototype/protofile_list_gen.go was copied
and altered to generated internal/fileinit.desc_list_gen.go.
At a high-level this CL adds significant technical complexity.
However, this is offset by several possible future changes:
* The prototype package can be drastically simplified. We can probably
reimplement internal/legacy to use internal/fileinit instead, allowing us
to drop another dependency on the prototype package. As a result, we can
probably delete most of the constructor types in that package.
* With the prototype package significantly pruned, and the fact that generated
code no longer depend on depends on that package, we can consider merging
what's left of prototype into protodesc.
Change-Id: I6090f023f2e1b6afaf62bd3ae883566242e30715
Reviewed-on: https://go-review.googlesource.com/c/158539
Reviewed-by: Herbie Ong <herbie@google.com>
Reviewed-by: Joe Tsai <thebrokentoaster@gmail.com>
A "_ProtoFile" suffix can potentially conflict with a sub-message named
"ProtoFile" nested within a message that matches the camel-cased
form of the basename of the .proto source file.
Avoid unlikely conflicts and rename this to use a "_protoFile" suffix,
which can never conflict except with an enum value that is also named
"protoFile" (which is a violation of the style guide).
Change-Id: Ie9d22f9f741a63021b8f76906b20c6c2f599885b
Reviewed-on: https://go-review.googlesource.com/c/157218
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>
Drop the protoreflect.ProtoEnum type (containing a single method
returning a protoreflect.Enum) and make generated enum types
directly implement protoreflect.Enum instead.
Messages have a two-level type split (ProtoMessage and Message) to
minimize conflicts between reflection methods and field names. Enums
need no such split, since enums do not have fields and therefore have
no source of conflicts.
Change-Id: I2b6222e9404253e6bfef2217859e1b760ffcd29b
Reviewed-on: https://go-review.googlesource.com/c/156902
Reviewed-by: Joe Tsai <thebrokentoaster@gmail.com>
Reviewed-by: Damien Neil <dneil@google.com>
Add fields to the Message and Field builder structs which hold the value
of MessageOptions.map_entry, FieldOptions.packed, and FieldOptions.weak
options. Remove all access to the contents of options messages from the
prototype package.
Change IsPacked to always return false for unpackable field types,
which is consistent with the equivalent C++ API.
This change helps avoid dependency cycles between prototype and the
options messages. (Previously this was resolved by accessing options
with reflection, but just breaking the dependency from prototype to the
options message is cleaner and simpler.)
Change-Id: I756aefe2e04cfa8fea31eaaaa0b5a99d4ac9e851
Reviewed-on: https://go-review.googlesource.com/c/153517
Reviewed-by: Joe Tsai <thebrokentoaster@gmail.com>
Remove the Mutable methods from KnownFields, List, and Map, replacing
them with methods which return a new, empty message value without adding
that value to the collection.
The new API is simpler, since it clearly applies only to message values,
and more orthogonal, since it provides a way to create a value without
mutating the collection. This latter point is particularly useful in
map deserialization, where the key may be unknown at the time the value
is deserialized.
Drop the Mutable interface, since it is no longer necessary.
Change-Id: Ic5f3d06a2aa331a5d5cd2b4e670a3dba4a74f77c
Reviewed-on: https://go-review.googlesource.com/c/153278
Reviewed-by: Joe Tsai <thebrokentoaster@gmail.com>
These properties of descriptors are currently missing and makes it impossible
to convert a FileDescriptorProto into one of the structured Go representations
and convert it back to a proto message without loss of information.
Furthermore, ReservedRanges and ReservedNames has semantic importance
to text serialization.
Change-Id: Ic33c30020ad51912b143156b95f47a4fb8da3503
Reviewed-on: https://go-review.googlesource.com/c/153019
Reviewed-by: Damien Neil <dneil@google.com>
Add pseudo-hidden functions to register the concrete Go type used for
the optional types. Also, augment protoc-gen-go to specially generate the
descriptor proto to register these types.
This change does not add validation logic yet to ensure that the correct option
types are passed to the API.
Change-Id: I5decc897e14b4bf570a61cf17b57a066a2a0f9d7
Reviewed-on: https://go-review.googlesource.com/c/153017
Reviewed-by: Damien Neil <dneil@google.com>
The following TODOs were addressed:
* Consistently collect all enums, messages, and extensions in a breadth-first order.
The practical affect of this is that the declaration order in a Go file may change.
This simplifies reflection generation, which relies on consistent ordering.
* Removal of placeholder declarations (e.g., "var _ = proto.Marshal") since
protogen is intelligent about including imports as necessary.
* Always generate a default variable or constant for explicit empty strings.
The practical effect of this is the addition of new declarations in some cases.
However, it simplifies our logic such that it matches the protobuf data model.
* Generate the registration statements in a consistent order.
Change-Id: I627bb72589432bb65d53b50965ea88e5f7983977
Reviewed-on: https://go-review.googlesource.com/c/152778
Reviewed-by: Damien Neil <dneil@google.com>
In CL/152020, we checked in pre-generated versions of descriptor and plugin.
This CL makes it such that they are generated by protoc-gen-go.
We modify protoc-gen-go to avoid reflection support by default
since the binary size increase is still an issue to investigate.
Reflection support is temporarily enabled by setting a special
PROTOC_GEN_GO_ENABLE_REFLECT environment variable.
Reflection support is always enabled for descriptor and plugin.
Furthermore, we change descriptor to depend on the protoapi package
instead of the proto package. The reason we do not switch to protoapi
for all generated protos is because we still depend on v1 proto
for the table-driven InternalMessageInfo type. Dropping it from descriptor
is semantically correct, but does incur slight performance cost.
It does not seem appropriate to drop it for all generated messages.
We could move InternalMessageInfo to protoapi, but the logic behind that
is significant.
Change-Id: I5c3fff7f6eab1a5a2399049d42fa6bf42d4c93f9
Reviewed-on: https://go-review.googlesource.com/c/152547
Reviewed-by: Damien Neil <dneil@google.com>
In order to generate descriptor.proto, the generated code would want to depend
on the prototype package to construct the reflection data structures.
However, this is a problem since descriptor itself is one of the dependencies
for prototype. To break this dependency, we do the following:
* Avoid using concrete *descriptorpb.XOptions messages in the public API, and
instead just use protoreflect.ProtoMessage. We do lose some type safety here
as a result.
* Use protobuf reflection to interpret the Options message.
* Split out NewFileFromDescriptorProto into a separate protodesc package since
constructing protobuf reflection from the descriptor proto obviously depends
on the descriptor protos themselves.
As part of this CL, we check in a pre-generated version of descriptor and plugin
that supports protobuf reflection natively and switchover all usages of those
protos to the new definitions. These files were generated by protoc-gen-go
from CL/150074, but hand-modified to remove dependencies on the v1 proto runtime.
Change-Id: I81e03c42eeab480b03764e2fcbe1aae0e058fc57
Reviewed-on: https://go-review.googlesource.com/c/152020
Reviewed-by: Damien Neil <dneil@google.com>
