API Reference
ReaderChunk
A normalized extraction chunk produced by DocumentReader.
Inheritance
- Object
- ReaderChunk
Constructors
public ReaderChunk() #Inherited Methods
public override Boolean Equals(Object obj) #BooleanParameters
- obj Object
Properties
public ReaderInputKind Kind { get; set; } #The kind of input that produced this chunk.
public ReaderLocation Location { get; set; } #Source location information for citations and debugging.
public String SourceId { get; set; } #Stable identifier for the source document. For file-based reads this is deterministic for a given normalized path.
public String SourceHash { get; set; } #Optional content hash for the source document (for incremental upserts).
public String ChunkHash { get; set; } #Optional content hash for this chunk (for incremental upserts).
public Nullable<DateTime> SourceLastWriteUtc { get; set; } #Optional source last-write timestamp (UTC) when available.
public Nullable<Int64> SourceLengthBytes { get; set; } #Optional source length in bytes when available.
public Nullable<Int32> TokenEstimate { get; set; } #Estimated token count (best-effort heuristic) for prompt budgeting.
public String Text { get; set; } #Plain text representation of the chunk.
public String Markdown { get; set; } #Optional Markdown representation of the chunk.
public IReadOnlyList<ReaderTable> Tables { get; set; } #Optional structured tables extracted from this chunk.
public IReadOnlyList<ReaderVisual> Visuals { get; set; } #Optional structured visual fence metadata extracted from this chunk.
public IReadOnlyList<String> Warnings { get; set; } #Optional warnings about truncation or unsupported content.