Codeflash / micropie#3
⚡️ Speed up method Server._parse_cookies by 6%
REVIEW
A valid input family runs 1.60× the instructions of the old code.
Performance · App()#_parse_cookies
Worst family with mostly valid inputs: 1.60× the instructions of base (many_empty_segments).
- F5
- F4
- F6
- F4.1
- F2
| Family | Inputs | Ratio range | Slope base → head | Verdict |
|---|---|---|---|---|
F5 | many_empty_segmentssuspected slowdown · plan | 1.03× – 1.60× | 0.55 → 0.63 | slower |
F4 | many_semicolon_segments_without_equalssuspected slowdown · plan | 1.02× – 1.37× | 0.53 → 0.57 | slower |
F6 | malformed cookie header – many semicolon‑separated parts that do NOT contain "="suspected slowdown · plan | 1.02× – 1.37× | 0.53 → 0.57 | slower |
F4.1 | many_semicolon_segments_with_plus_in_middlesuspected slowdown · expand:F4 | 1.02× – 1.35× | 0.52 → 0.56 | slower |
F2 | many_well_formed_cookiesclaimed speedup · plan | 1.01× – 1.01× | 0.71 → 0.71 | neutral |
Receipt
| Target | Backend | Image | Run | Audit | Adjudication |
|---|---|---|---|---|---|
MicroPie:App()#_parse_cookies | Token Factory Sandboxes | f05e47b2-dfb5-4e4e-8bff-cb2a2dcddd6b | bt-micropie#3-ae94a9 | $0.053 | $0.000 |
Nemotron cost on Nebius Token Factory. Backend: where the untrusted code ran. Token Factory Sandboxes (a microVM branched from a kept image, cited by its UUID). Every image is content-addressed (repo at each commit + harness), so every card re-runs from the image and seed 0. Raw data: audit.