Parity Dashboard

neokapi vs Okapi Framework parity — round-trip byte/canonical/semantic tier comparison per (format, engine, fixture). Each non-byte-equal row carries raw/normalized first-diff offsets and the comparison snippet so you can scan for patterns (line endings, whitespace, encoding…) without re-running the test. Canon and sem rows show the raw byte gap that the normalizer (or semantic comparator) is currently bridging. Severity, GitHub issue links, and spec references are loaded from core/formats/<format>/parity-annotations.yaml. Generated 2026-05-21T23:44:45Z.

bridge

731 / 731 faithful (100.0%)

731 byte-equal (100.0%) + 0 faithful-canon ⓘ

byte 731
canon 0
div 0

native

725 / 727 faithful (99.7%)

314 byte-equal (43.2%) + 368 faithful-canon + 43 faithful-div ⓘ

byte 314
canon 370 (368 faithful)
div 43
skip 27

Divergence severity (click a row to filter)

23 (format, engine) rows · 413 fixtures matching

openxmlnativebyte0canon149div36· 185 total▸ 185

ttxnativebyte0canon1div3· 4 total▸ 4

txmlnativebyte0canon0div3· 3 total▸ 3

htmlnativebyte54canon14div1· 69 total▸ 15

idmlnativebyte0canon70div0· 70 total▸ 70

xliffnativebyte0canon35div0skip1· 36 total▸ 35

xmlnativebyte0canon23div0· 23 total▸ 23

xliff2nativebyte0canon18div0· 18 total▸ 18

odfnativebyte0canon15div0· 15 total▸ 15

tmxnativebyte0canon10div0skip2· 12 total▸ 10

icmlnativebyte0canon9div0skip5· 14 total▸ 9

ponativebyte20canon4div0· 24 total▸ 4

propertiesnativebyte3canon4div0· 7 total▸ 4

markdownnativebyte43canon3div0· 46 total▸ 3

dtdnativebyte0canon2div0· 2 total▸ 2

jsonnativebyte67canon2div0skip1· 70 total▸ 2

tsnativebyte6canon2div0skip1· 9 total▸ 2

vttnativebyte0canon2div0· 2 total▸ 2

wikinativebyte2canon2div0· 4 total▸ 2

yamlnativebyte25canon2div0skip3· 30 total▸ 2

doxygennativebyte5canon1div0· 6 total▸ 1

phpcontentnativebyte0canon1div0· 1 total▸ 1

texnativebyte4canon1div0· 5 total▸ 1