refactor(core): retire the dead crease-line path + gitignore render artifacts #11
Loading…
Add table
Add a link
Reference in a new issue
No description provided.
Delete branch "v0.10-cleanup"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Cleanup #6 — retire dead code and ignore render scratch. No behaviour change;
beloch foldoutput is identical.Since #10 switched
fold_stringto the face-based path (eval_folded→to_json_folded), the old crease-line pipeline is dead. This removes it:lib/planarize.ml,lib/faces.ml— deleted.Eval.eval(thecrease listevaluator) and theEval.creasetype — deleted;corners,folded, andeval_foldedstay.lib/state.ml— trimmed to justprovenance(the vertex/edge/t/add_*machinery only fedPlanarize).beloch.ml— drops thePlanarize/Facesre-exports.planarize/faces/statetest groups — deleted (69 tests, was 76; the removed tests only exercised the dead path)..gitignore— addslib/parser.conflictsand root-level render scratch (/*.fold,/*.svg,/*.png).Assumptions made at forks (no input requested):
Geomhelpers (signed_area,ccw_compare,clip_to_unit_square,segment_intersection) rather than pruning them. They were only used byFaces/Planarize, but they are small, tested, exact-geometry utilities with no.mli(so no unused-value warning), and plausibly reused by future slices (e.g.signed_area/ccw_comparefor layer geometry). Pruning them would be churn for little gain.lib/parser.conflictsis gitignored, not just deleted — it is an empty, stale menhir--explainartifact; a fresh build does not regenerate it (the grammar is conflict-free), but--explainwould write it again if a future grammar change introduced a conflict, so ignoring it is the durable fix.*.fold/*.svg/*.pngare gitignored (notexamples/, which is.belonly) — these arebeloch fold/fold2svgscratch outputs that shouldn't be committed.StatenamedState(now just provenance) rather than renaming toProvenance, to keep the diff minimal and thepr.State.axiomaccess sites stable.fold2svgitself (the legend wording fix) is a separate PR so this one stays core-only.