planetscale / planetscale/vtprotobuf
Avoid proto reflection for builtin types
Nobody has claimed this yet.
- Dominant language
- Go
- Stars
- 1.1k
- Forks
- 112
- PR merge metrics
- No merged PRs in 30d
Description
The optimized generated code for marshal, unmarshal, clone etc. still resorts to generic, protoreflect-based logic for builtin/well-known types such as google.protobuf.Timestamp, google.protobuf.Duration etc. Since these types are well known, it would be nice if optimized unrolled code could be used for operations on these types as well - especially as there seems to be a significant performance penalty for the "context-switch" to reflection with each individual proto.Clone invocation (benchmark).
I'm happy to send a PR, but one thing I'd ask the library maintainers to chime in on is whether it's acceptable to have the generated code reference global helper functions from a package within this module (e.g., github.com/planetscale/vtprotobuf/support/...), or whether it would be preferable to just generate package-private helper functions for all referenced types on demand.
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start with the optimized generated marshal, unmarshal, and clone paths for builtin and well-known types such as google.protobuf.Timestamp and google.protobuf.Duration. Review the linked proto.Clone benchmark and the proposed helper-function options; done means these operations no longer fall back to protoreflect-based logic while retaining correct behavior.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- go
- Domain
- devtools, performance
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 25/100