pub.layers.alignment

Alignment records for parallel structure correspondence. Handles interlinear glossing (Leipzig glossing rules), parallel text alignment (translation), cross-tokenization mapping (word-to-morpheme), audio-text forced alignment, and any many-to-many correspondence between annotation elements or sequences.

Types

alignment

Type: Record

An alignment between two parallel sequences. The sequences can be tokenizations, annotation layers, expressions (for parallel text), or tiers. Links establish many-to-many correspondence between elements indexed by position.

Field	Type	Description
expression	at-uri	Primary expression context (for within-document alignments).
kindUri	at-uri	AT-URI of the alignment kind definition node. Community-expandable via knowledge graph.
kind	string	Alignment kind slug (fallback). Known values: tokenization-to-tokenization, interlinear, parallel-text, audio-to-text, layer-to-layer, error-to-correction, custom
subkindUri	at-uri	AT-URI of the alignment subkind definition node. Community-expandable via knowledge graph.
subkind	string	Alignment subkind slug (fallback). Known values: word-to-morpheme, word-to-word, sentence-to-sentence, phrase-to-phrase, morpheme-to-gloss, forced-alignment, manual-alignment, custom
source	ref	Reference to the source sequence. Ref: pub.layers.defs#objectRef
target	ref	Reference to the target sequence. Ref: pub.layers.defs#objectRef
sourceLang	string	BCP-47 language tag for the source (for parallel text alignment).
targetLang	string	BCP-47 language tag for the target.
links	array	The alignment links. Array of ref: pub.layers.defs#alignmentLink
metadata	ref	Ref: pub.layers.defs#annotationMetadata
knowledgeRefs	array	Knowledge graph references. Array of ref: pub.layers.defs#knowledgeRef
features	ref	Ref: pub.layers.defs#featureMap
createdAt	datetime	Record creation timestamp.