TextChunk class

One chunk of source text. Carries enough metadata to cite back to the original document (source path + ordinal within that document) plus arbitrary meta for higher-level info (page number, heading path, etc.).

Chunks are produced by a Chunker; an Embedder turns the text of each chunk into a vector; a CorpusStore persists the pair.

Annotations

Constructors

TextChunk({required String text, required String sourcePath, required int ord, Map<String, Object?> meta = const {}})
const

Properties

hashCode → int
The hash code for this object.
no setterinherited
meta → Map<String, Object?>
Free-form per-chunk metadata. Stored as JSON in the SQLite row. Keep values JSON-encodable (no DateTime, no custom types).
final
ord → int
Index of this chunk within its source document, starting at 0. Adjacent ordinals share overlap characters when produced by a sliding-window chunker.
final
runtimeType → Type
A representation of the runtime type of the object.
no setterinherited
sourcePath → String
Origin file (or logical identifier) for citations. Free-form; the store uses it verbatim.
final
text → String
The chunk's text payload — what gets embedded and shown back to the model at retrieval time.
final

Methods

noSuchMethod(Invocation invocation) → dynamic
Invoked when a nonexistent method or property is accessed.
inherited
toString() → String
A string representation of this object.
override

Operators

operator ==(Object other) → bool
The equality operator.
inherited