FlexBuffers is a schema-less, dynamic format optimized for reading single value in a document. It's a child format of FlatBuffers.
For more detailed specification, please refer https://google.github.io/flatbuffers/flexbuffers.html
NOTE: I'm planning extend flexbuffers by adding ability to attach metadata to each element. And I'm going to rename project name as it will lose few backward compatibility. (type bit space is too tight to attach metadata information, so FBT_VECTOR_BOOL might be no longer available in forked version)
Flexbuffers is optimized for lookup single value in a document.
About 3x faster than BSON's LookupErr method. (BenchmarkFlexbuffersTraverseByTraverser-4 vs BenchmarkBSONTraverseTree-4)
cd bench
go test -tags unsafe -bench . -benchmemBenchmarkFlexbuffersTraverseByReference-4 1636011 726 ns/op 0 B/op 0 allocs/op
BenchmarkFlexbuffersTraverseByTraverser-4 4406778 272 ns/op 0 B/op 0 allocs/op
BenchmarkBSONTraverseTree-4 1271134 932 ns/op 0 B/op 0 allocs/op
BenchmarkJSONTraverseByFastJson-4 277 5343157 ns/op 10257193 B/op 11229 allocs/op
BenchmarkJSONTraverseByFastJsonWithoutParseTime-4 6981626 151 ns/op 0 B/op 0 allocs/op
BenchmarkMsgpackTraverse-4 1900 848233 ns/op 288 B/op 5 allocs/op
![dependabot[bot] avatar](https://avatars.githubusercontent.com/in/29110?v=4 "dependabot[bot]")
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent