Repository navigation
v0.5.166
What's Changed
- fix(xsd): dismiss content-model findings that reject substitution members (#466) by @ronaldtse in #467
Full Changelog: v0.5.165...v0.5.166
Performance scoreboard
CPU-time medians of an interleaved A/B run against raw nokogiri on the same document (ratio > 1.0 means moxml is faster). Pair: moxml + leptris 1.9.383.0 on Ruby 3.3.13; 9 repetitions.
| Row | Ratio | Kind | Status |
|---|---|---|---|
parse |
3.05x | parity | met |
serialize |
1.62x | parity | met |
xpath //field (count) |
3.04x | parity | met |
xpath //field (each) |
0.59x | structural | tracked (below parity) |
attr pairs listing |
7.35x | parity | met |
build add_element |
0.98x | structural | tracked (at parity) |
walk+read (warm) |
0.67x | structural | tracked (below parity) |
walk+read (cold) |
0.37x | structural | tracked (below parity), paused: leptris-ruby#527 (1.9.380 crossing regression) |
children recursion (cold) |
0.46x | structural | tracked (below parity), paused: leptris-ruby#527 (1.9.380 crossing regression) |
plan records (typed) |
1.09x | data | informational |
Parity rows are the commitment (0.95 measurement floor); structural rows are tracked (10% regression tolerance) with their gaps documented in benchmark/README.md.
Contributors
ronaldtse