@huukhanhnguyen/model / nodeMethods/filterSynthesis

nodeMethods/filterSynthesis#

synthesizeDataFilter()#

synthesizeDataFilter(siblings, targetIndex): SynthesizedFilter<DataFilterTier>

A filter expression naming siblings[targetIndex] among the data documents one node emitted, in the (data, index) => dialect every per-document callback arg takes.

Tiers, in test order: type (the only face / curve / point in the stream) · sizeRank (uniquely largest or smallest bbox) · point (bbox centre) · ordinal (index === n — the array-copy case, where every copy is congruent and nothing qualitative separates them).

Parameters#

siblings#

ObjectData[]

targetIndex#

number

Returns#

SynthesizedFilter<DataFilterTier>

synthesizeFaceFilter()#

synthesizeFaceFilter(data, faceIndex): SynthesizedFilter<FaceFilterTier>

A filter expression naming face faceIndex of one brep wrap, in the (points, normal) => dialect faceExtrude consumes.

Tiers, in test order: normal (the only face pointing that way) · plane (that direction at that offset) · areaRank (uniquely largest or smallest, as a scale-stable threshold) · point (bbox centre of the face) · ordinal (last resort).

ordinal is NOT an index: this dialect hands the predicate no face index, so the last resort is the tightest available fingerprint (plane AND centre) and is flagged brittle — it may still name more than one face when two faces genuinely coincide.

Parameters#

data#

BrepData

faceIndex#

number

Returns#

SynthesizedFilter<FaceFilterTier>