You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: benchmark/README.md
+25Lines changed: 25 additions & 0 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -81,6 +81,19 @@ mise run benchmark-json-canada
81
81
mise run benchmark-json-catalog
82
82
```
83
83
84
+
### SQLite Parsing (`sqlite_parse/`)
85
+
86
+
Parses 81 realistic SQLite statements (13KB of SQL) using a parser generated by Gale from the ANTLR4 `SQLite.g4` grammar. Compares Gale's auto-generated recursive descent parser (Wado, compiled to Wasm) against Rust's `sqlparser-rs` (hand-written recursive descent, native).
Both implementations parse 81 SQL statements per iteration. The Gale parser is auto-generated from ANTLR4's `SQLite.g4` grammar; `sqlparser-rs` is a hand-tuned native Rust parser.
232
+
210
233
## Profiling Wado Programs
211
234
212
235
`wado run --profile <mode>` enables runtime profiling via wasmtime's profiling infrastructure.
@@ -308,6 +331,7 @@ samply record wado run --profile perfmap benchmark/count_prime/count_prime.wado
308
331
- Zig benchmarks use `-OReleaseFast`
309
332
- JSON benchmarks compare Wado's `core:json` (pure Wado, Wasm) against Rust's `serde_json` (native). Rust uses `BTreeMap` for map fields to match Wado's `TreeMap`
310
333
- JSON test data from [nativejson-benchmark](https://github.com/miloyip/nativejson-benchmark)
0 commit comments