process
fun processOnce(text: String, sourceId: String, context: SourceAnalysisContext, historyStore: ChunkHistoryStore? = null, contentHasher: ContentHasher = Sha256ContentHasher, additionalGrounding: List<String> = emptyList()): ChunkPropositionResult?
Process a text once, with hash-based deduplication. Ideal for one-shot ingestion of documents, notes, or other static text.
Return
Result with propositions and entities, or null if already processed
Note: This method does NOT persist anything — it returns UNSAVED results. The caller must persist the returned entities and propositions via PersistablePropositions.persist within its own transaction scope; nothing is written to any repository otherwise.
Parameters
text
The text to process
source
Identifier for the source (used for chunk metadata and bookmark tracking)
context
Analysis context with schema, entity resolver, etc.
history
Tracks what's been processed; null to skip dedup
content
Strategy for computing content hashes