@mastra/rag 1.0.6 → 1.0.7-alpha.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/.turbo/turbo-build.log +1 -1
- package/CHANGELOG.md +12 -0
- package/dist/document/document.d.ts +9 -8
- package/dist/document/document.d.ts.map +1 -1
- package/dist/document/transformers/character.d.ts +4 -26
- package/dist/document/transformers/character.d.ts.map +1 -1
- package/dist/document/transformers/html.d.ts +8 -3
- package/dist/document/transformers/html.d.ts.map +1 -1
- package/dist/document/transformers/json.d.ts +4 -4
- package/dist/document/transformers/json.d.ts.map +1 -1
- package/dist/document/transformers/latex.d.ts +2 -8
- package/dist/document/transformers/latex.d.ts.map +1 -1
- package/dist/document/transformers/markdown.d.ts +2 -8
- package/dist/document/transformers/markdown.d.ts.map +1 -1
- package/dist/document/transformers/sentence.d.ts +31 -0
- package/dist/document/transformers/sentence.d.ts.map +1 -0
- package/dist/document/transformers/text.d.ts +3 -3
- package/dist/document/transformers/text.d.ts.map +1 -1
- package/dist/document/transformers/token.d.ts +4 -15
- package/dist/document/transformers/token.d.ts.map +1 -1
- package/dist/document/types.d.ts +85 -14
- package/dist/document/types.d.ts.map +1 -1
- package/dist/document/validation.d.ts +3 -0
- package/dist/document/validation.d.ts.map +1 -0
- package/dist/index.cjs +414 -80
- package/dist/index.cjs.map +1 -1
- package/dist/index.js +414 -80
- package/dist/index.js.map +1 -1
- package/dist/tools/document-chunker.d.ts.map +1 -1
- package/package.json +5 -5
- package/src/document/document.test.ts +294 -39
- package/src/document/document.ts +69 -41
- package/src/document/transformers/character.ts +15 -43
- package/src/document/transformers/html.ts +9 -9
- package/src/document/transformers/json.ts +8 -3
- package/src/document/transformers/latex.ts +3 -11
- package/src/document/transformers/markdown.ts +3 -11
- package/src/document/transformers/sentence.ts +314 -0
- package/src/document/transformers/text.ts +10 -10
- package/src/document/transformers/token.ts +6 -17
- package/src/document/types.ts +66 -15
- package/src/document/validation.ts +147 -0
- package/src/tools/document-chunker.ts +12 -8
|
@@ -3,22 +3,26 @@ import { z } from 'zod';
|
|
|
3
3
|
|
|
4
4
|
import type { MDocument, ChunkParams } from '../document';
|
|
5
5
|
|
|
6
|
+
const DEFAULT_CHUNK_PARAMS = {
|
|
7
|
+
strategy: 'recursive' as const,
|
|
8
|
+
maxSize: 512,
|
|
9
|
+
overlap: 50,
|
|
10
|
+
separators: ['\n'],
|
|
11
|
+
} satisfies ChunkParams;
|
|
12
|
+
|
|
6
13
|
export const createDocumentChunkerTool = ({
|
|
7
14
|
doc,
|
|
8
|
-
params =
|
|
9
|
-
strategy: 'recursive',
|
|
10
|
-
size: 512,
|
|
11
|
-
overlap: 50,
|
|
12
|
-
separator: '\n',
|
|
13
|
-
},
|
|
15
|
+
params = DEFAULT_CHUNK_PARAMS,
|
|
14
16
|
}: {
|
|
15
17
|
doc: MDocument;
|
|
16
18
|
params?: ChunkParams;
|
|
17
19
|
}): ReturnType<typeof createTool> => {
|
|
18
20
|
return createTool({
|
|
19
|
-
id: `Document Chunker ${params.strategy} ${params.
|
|
21
|
+
id: `Document Chunker ${params.strategy} ${params.maxSize}`,
|
|
20
22
|
inputSchema: z.object({}),
|
|
21
|
-
description: `Chunks document using ${params.strategy} strategy with
|
|
23
|
+
description: `Chunks document using ${params.strategy} strategy with maxSize ${params.maxSize} and ${
|
|
24
|
+
params.overlap || 0
|
|
25
|
+
} overlap`,
|
|
22
26
|
execute: async () => {
|
|
23
27
|
const chunks = await doc.chunk(params);
|
|
24
28
|
|