Repository navigation
Per-phrase tiers 4/4: verify the mastery slider works end to end on an editor-saved pack #36
Description
Part of #3. Last: depends on Per-phrase tiers 1/4 and at least one of 2/4 or 3/4.
The point of #3 is that the master-difficulty slider (feedBack#48) should work for editor-authored sloppaks, not just GP-imported ones.
Scope
- Author or derive distinct tiers for a phrase, save a sloppak, load it in the host, and confirm the mastery slider changes which notes are shown (the consumer is core's
static/highway.js_rebuildMasteryFilter(), which readsphrases[].levels[]). - Capture this as a repeatable test or a documented manual script with fixtures.
- Update the CHANGELOG note that says the slider is a no-op for editor-saved packs.
Acceptance criteria
- The slider visibly changes the note set for an editor-saved pack with authored tiers.
- A pack without authored tiers behaves as today.
- The documentation no longer says the slider is a no-op.
Out of scope
Difficulty Ladder's own generation (that lives in the difficulty-ladder plugin).
Activity
pullfrog commentedon Oct 2, 2026
pullfrog commentedon Oct 10, 2026
PR opened: #67
Closed #36 (per-phrase tiers 4/4) with a repeatable end-to-end verification suite,
tests/test_mastery_filter_e2e.py(5 tests), that mirrors core's highway mastery filter and drives it over the real save-path output (_repopulate_phrase_levels), plus the docs/CHANGELOG corrections for the "slider is a no-op" framing (new[Unreleased]entry; 1.4.3 note amended with a pointer). All three acceptance criteria pinned: authored tiers visibly change the rendered set per slider position, untiered packs render identically at every position (including the freshly-added-phrase empty-ladder case), and no documentation claims the slider is a no-op anymore.Verification: full pytest suite 502 passed / 5 skipped,
npm test400/400,npm run lint0 errors (3 baseline warnings). A mutation check confirmed the tiered-pack tests fail if the save path's authored-tiers branch regresses. No product code changed.Reviewer (reviewfrog) findings were addressed: mirror-guard wording now scope the manual re-sync to core,
Math.floorused for exactness, the handshape fall-through gate modeled, themax_difficulty > top-tierdivergence pinned and documented, the fresh-phrase empty-ladder acceptance-2 case added, the cascade claim qualified, and the CHANGELOG no-op/chain wording tightened.Manual host-side verification remains the scripted step:
Alt+Pa phrase → save → load in host → drag the master-difficulty slider.Task list (1/7 completed)
big-pickle(free) ｜ 𝕏test(editor): add trailing newline to mastery filter e2e suite (#36)
test(editor): verify mastery slider end to end on editor-saved packs (#…
test(editor): add trailing newline to mastery filter e2e suite (#36)