Class: SimpleNodeParser
SimpleNodeParser is the default NodeParser. It splits documents into TextNodes using a splitter, by default SentenceSplitter
Implements
Constructors
constructor
• new SimpleNodeParser(init?)
Parameters
| Name | Type |
|---|---|
init? | Object |
init.chunkOverlap? | number |
init.chunkSize? | number |
init.includeMetadata? | boolean |
init.includePrevNextRel? | boolean |
init.textSplitter? | SentenceSplitter |
Defined in
packages/core/src/NodeParser.ts:93
Properties
includeMetadata
• includeMetadata: boolean
Whether to include metadata in the nodes.
Defined in
packages/core/src/NodeParser.ts:87
includePrevNextRel
• includePrevNextRel: boolean
Whether to include previous and next relationships in the nodes.
Defined in
packages/core/src/NodeParser.ts:91
textSplitter
• textSplitter: SentenceSplitter
The text splitter to use.
Defined in
packages/core/src/NodeParser.ts:83
Methods
getNodesFromDocuments
▸ getNodesFromDocuments(documents): TextNode<Metadata>[]
Generate Node objects from documents
Parameters
| Name | Type |
|---|---|
documents | Document<Metadata>[] |
Returns
Implementation of
NodeParser.getNodesFromDocuments
Defined in
packages/core/src/NodeParser.ts:124
fromDefaults
▸ Static fromDefaults(init?): SimpleNodeParser
Parameters
| Name | Type |
|---|---|
init? | Object |
init.chunkOverlap? | number |
init.chunkSize? | number |
init.includeMetadata? | boolean |
init.includePrevNextRel? | boolean |