@cogitator-ai/rag 0.1.8 → 0.1.10
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/LICENSE +21 -0
- package/dist/chunkers/recursive-chunker.d.ts +2 -2
- package/dist/chunkers/recursive-chunker.d.ts.map +1 -1
- package/dist/chunkers/recursive-chunker.js +49 -88
- package/dist/chunkers/recursive-chunker.js.map +1 -1
- package/dist/chunkers/semantic-chunker.d.ts +2 -0
- package/dist/chunkers/semantic-chunker.d.ts.map +1 -1
- package/dist/chunkers/semantic-chunker.js +68 -37
- package/dist/chunkers/semantic-chunker.js.map +1 -1
- package/dist/loaders/csv-loader.d.ts.map +1 -1
- package/dist/loaders/csv-loader.js +3 -0
- package/dist/loaders/csv-loader.js.map +1 -1
- package/dist/loaders/json-loader.d.ts +1 -0
- package/dist/loaders/json-loader.d.ts.map +1 -1
- package/dist/loaders/json-loader.js +15 -3
- package/dist/loaders/json-loader.js.map +1 -1
- package/dist/loaders/markdown-loader.js +1 -1
- package/dist/loaders/markdown-loader.js.map +1 -1
- package/dist/loaders/pdf-loader.d.ts +1 -0
- package/dist/loaders/pdf-loader.d.ts.map +1 -1
- package/dist/loaders/pdf-loader.js +22 -3
- package/dist/loaders/pdf-loader.js.map +1 -1
- package/dist/loaders/web-loader.d.ts +5 -0
- package/dist/loaders/web-loader.d.ts.map +1 -1
- package/dist/loaders/web-loader.js +127 -8
- package/dist/loaders/web-loader.js.map +1 -1
- package/dist/rag-builder.d.ts.map +1 -1
- package/dist/rag-builder.js +3 -1
- package/dist/rag-builder.js.map +1 -1
- package/dist/rag-pipeline.d.ts +1 -1
- package/dist/rag-pipeline.d.ts.map +1 -1
- package/dist/rag-pipeline.js +5 -6
- package/dist/rag-pipeline.js.map +1 -1
- package/dist/rerankers/cohere-reranker.d.ts +2 -2
- package/dist/rerankers/cohere-reranker.d.ts.map +1 -1
- package/dist/rerankers/cohere-reranker.js +5 -0
- package/dist/rerankers/cohere-reranker.js.map +1 -1
- package/dist/rerankers/llm-reranker.d.ts +1 -1
- package/dist/rerankers/llm-reranker.d.ts.map +1 -1
- package/dist/rerankers/llm-reranker.js +26 -13
- package/dist/rerankers/llm-reranker.js.map +1 -1
- package/dist/retrievers/create-retriever.d.ts +1 -0
- package/dist/retrievers/create-retriever.d.ts.map +1 -1
- package/dist/retrievers/create-retriever.js +5 -0
- package/dist/retrievers/create-retriever.js.map +1 -1
- package/dist/retrievers/mmr-retriever.d.ts.map +1 -1
- package/dist/retrievers/mmr-retriever.js +11 -7
- package/dist/retrievers/mmr-retriever.js.map +1 -1
- package/dist/retrievers/multi-query-retriever.d.ts +2 -0
- package/dist/retrievers/multi-query-retriever.d.ts.map +1 -1
- package/dist/retrievers/multi-query-retriever.js +10 -5
- package/dist/retrievers/multi-query-retriever.js.map +1 -1
- package/dist/schema.d.ts.map +1 -1
- package/dist/schema.js +16 -2
- package/dist/schema.js.map +1 -1
- package/dist/tools.d.ts.map +1 -1
- package/dist/tools.js +6 -4
- package/dist/tools.js.map +1 -1
- package/package.json +16 -16
|
@@ -26,7 +26,7 @@ export class PDFLoader {
|
|
|
26
26
|
return this.loadSingleDocument(pdfParse, buffer, filePath);
|
|
27
27
|
}
|
|
28
28
|
async loadSingleDocument(pdfParse, buffer, source) {
|
|
29
|
-
const result = await pdfParse
|
|
29
|
+
const result = await this.parseWithErrorContext(pdfParse, buffer, source);
|
|
30
30
|
const title = result.info?.Title;
|
|
31
31
|
const metadata = { pages: result.numpages };
|
|
32
32
|
if (title)
|
|
@@ -42,14 +42,33 @@ export class PDFLoader {
|
|
|
42
42
|
];
|
|
43
43
|
}
|
|
44
44
|
async loadSplitPages(pdfParse, buffer, source) {
|
|
45
|
-
const
|
|
45
|
+
const pageTexts = [];
|
|
46
|
+
const result = await this.parseWithErrorContext(pdfParse, buffer, source, {
|
|
47
|
+
pagerender: async (pageData) => {
|
|
48
|
+
if (typeof pageData.getTextContent !== 'function')
|
|
49
|
+
return '';
|
|
50
|
+
const textContent = await pageData.getTextContent();
|
|
51
|
+
const text = textContent.items.map((item) => item.str).join('');
|
|
52
|
+
pageTexts.push(text);
|
|
53
|
+
return text;
|
|
54
|
+
},
|
|
55
|
+
});
|
|
46
56
|
const title = result.info?.Title;
|
|
47
|
-
const pages =
|
|
57
|
+
const pages = pageTexts.filter((p) => p.trim().length > 0);
|
|
48
58
|
if (pages.length === 0) {
|
|
49
59
|
return [this.buildPageDoc(result.text, source, 1, result.numpages, title)];
|
|
50
60
|
}
|
|
51
61
|
return pages.map((text, i) => this.buildPageDoc(text.trim(), source, i + 1, result.numpages, title));
|
|
52
62
|
}
|
|
63
|
+
async parseWithErrorContext(pdfParse, buffer, source, options) {
|
|
64
|
+
try {
|
|
65
|
+
return await pdfParse(buffer, options);
|
|
66
|
+
}
|
|
67
|
+
catch (err) {
|
|
68
|
+
const message = err instanceof Error ? err.message : String(err);
|
|
69
|
+
throw new Error(`PDFLoader: failed to parse "${source}": ${message}`);
|
|
70
|
+
}
|
|
71
|
+
}
|
|
53
72
|
buildPageDoc(content, source, pageNumber, totalPages, title) {
|
|
54
73
|
const metadata = { pageNumber, totalPages };
|
|
55
74
|
if (title)
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"pdf-loader.js","sourceRoot":"","sources":["../../src/loaders/pdf-loader.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,QAAQ,EAAE,MAAM,kBAAkB,CAAC;AAC5C,OAAO,EAAE,OAAO,EAAE,MAAM,WAAW,CAAC;AACpC,OAAO,EAAE,MAAM,EAAE,MAAM,QAAQ,CAAC;
|
|
1
|
+
{"version":3,"file":"pdf-loader.js","sourceRoot":"","sources":["../../src/loaders/pdf-loader.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,QAAQ,EAAE,MAAM,kBAAkB,CAAC;AAC5C,OAAO,EAAE,OAAO,EAAE,MAAM,WAAW,CAAC;AACpC,OAAO,EAAE,MAAM,EAAE,MAAM,QAAQ,CAAC;AAoBhC,KAAK,UAAU,YAAY;IACzB,IAAI,CAAC;QACH,MAAM,GAAG,GAAG,MAAM,MAAM,CAAC,WAAW,CAAC,CAAC;QACtC,OAAO,GAAG,CAAC,OAAgC,CAAC;IAC9C,CAAC;IAAC,MAAM,CAAC;QACP,MAAM,IAAI,KAAK,CAAC,qEAAqE,CAAC,CAAC;IACzF,CAAC;AACH,CAAC;AAMD,MAAM,OAAO,SAAS;IACX,cAAc,GAAG,CAAC,KAAK,CAAC,CAAC;IACjB,UAAU,CAAU;IAErC,YAAY,OAA0B;QACpC,IAAI,CAAC,UAAU,GAAG,OAAO,EAAE,UAAU,IAAI,KAAK,CAAC;IACjD,CAAC;IAED,KAAK,CAAC,IAAI,CAAC,MAAc;QACvB,MAAM,QAAQ,GAAG,MAAM,YAAY,EAAE,CAAC;QACtC,MAAM,QAAQ,GAAG,OAAO,CAAC,MAAM,CAAC,CAAC;QACjC,MAAM,MAAM,GAAG,MAAM,QAAQ,CAAC,QAAQ,CAAC,CAAC;QAExC,IAAI,IAAI,CAAC,UAAU,EAAE,CAAC;YACpB,OAAO,IAAI,CAAC,cAAc,CAAC,QAAQ,EAAE,MAAM,EAAE,QAAQ,CAAC,CAAC;QACzD,CAAC;QAED,OAAO,IAAI,CAAC,kBAAkB,CAAC,QAAQ,EAAE,MAAM,EAAE,QAAQ,CAAC,CAAC;IAC7D,CAAC;IAEO,KAAK,CAAC,kBAAkB,CAC9B,QAAoB,EACpB,MAAc,EACd,MAAc;QAEd,MAAM,MAAM,GAAG,MAAM,IAAI,CAAC,qBAAqB,CAAC,QAAQ,EAAE,MAAM,EAAE,MAAM,CAAC,CAAC;QAC1E,MAAM,KAAK,GAAG,MAAM,CAAC,IAAI,EAAE,KAAK,CAAC;QAEjC,MAAM,QAAQ,GAA4B,EAAE,KAAK,EAAE,MAAM,CAAC,QAAQ,EAAE,CAAC;QACrE,IAAI,KAAK;YAAE,QAAQ,CAAC,KAAK,GAAG,KAAK,CAAC;QAElC,OAAO;YACL;gBACE,EAAE,EAAE,MAAM,EAAE;gBACZ,OAAO,EAAE,MAAM,CAAC,IAAI;gBACpB,MAAM;gBACN,UAAU,EAAE,KAAK;gBACjB,QAAQ;aACT;SACF,CAAC;IACJ,CAAC;IAEO,KAAK,CAAC,cAAc,CAC1B,QAAoB,EACpB,MAAc,EACd,MAAc;QAEd,MAAM,SAAS,GAAa,EAAE,CAAC;QAE/B,MAAM,MAAM,GAAG,MAAM,IAAI,CAAC,qBAAqB,CAAC,QAAQ,EAAE,MAAM,EAAE,MAAM,EAAE;YACxE,UAAU,EAAE,KAAK,EAAE,QAAqB,EAAE,EAAE;gBAC1C,IAAI,OAAO,QAAQ,CAAC,cAAc,KAAK,UAAU;oBAAE,OAAO,EAAE,CAAC;gBAC7D,MAAM,WAAW,GAAG,MAAM,QAAQ,CAAC,cAAc,EAAE,CAAC;gBACpD,MAAM,IAAI,GAAG,WAAW,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;gBAChE,SAAS,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;gBACrB,OAAO,IAAI,CAAC;YACd,CAAC;SACF,CAAC,CAAC;QAEH,MAAM,KAAK,GAAG,MAAM,CAAC,IAAI,EAAE,KAAK,CAAC;QACjC,MAAM,KAAK,GAAG,SAAS,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC;QAE3D,IAAI,KAAK,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;YACvB,OAAO,CAAC,IAAI,CAAC,YAAY,CAAC,MAAM,CAAC,IAAI,EAAE,MAAM,EAAE,CAAC,EAAE,MAAM,CAAC,QAAQ,EAAE,KAAK,CAAC,CAAC,CAAC;QAC7E,CAAC;QAED,OAAO,KAAK,CAAC,GAAG,CAAC,CAAC,IAAI,EAAE,CAAC,EAAE,EAAE,CAC3B,IAAI,CAAC,YAAY,CAAC,IAAI,CAAC,IAAI,EAAE,EAAE,MAAM,EAAE,CAAC,GAAG,CAAC,EAAE,MAAM,CAAC,QAAQ,EAAE,KAAK,CAAC,CACtE,CAAC;IACJ,CAAC;IAEO,KAAK,CAAC,qBAAqB,CACjC,QAAoB,EACpB,MAAc,EACd,MAAc,EACd,OAAmC;QAEnC,IAAI,CAAC;YACH,OAAO,MAAM,QAAQ,CAAC,MAAM,EAAE,OAAO,CAAC,CAAC;QACzC,CAAC;QAAC,OAAO,GAAG,EAAE,CAAC;YACb,MAAM,OAAO,GAAG,GAAG,YAAY,KAAK,CAAC,CAAC,CAAC,GAAG,CAAC,OAAO,CAAC,CAAC,CAAC,MAAM,CAAC,GAAG,CAAC,CAAC;YACjE,MAAM,IAAI,KAAK,CAAC,+BAA+B,MAAM,MAAM,OAAO,EAAE,CAAC,CAAC;QACxE,CAAC;IACH,CAAC;IAEO,YAAY,CAClB,OAAe,EACf,MAAc,EACd,UAAkB,EAClB,UAAkB,EAClB,KAAc;QAEd,MAAM,QAAQ,GAA4B,EAAE,UAAU,EAAE,UAAU,EAAE,CAAC;QACrE,IAAI,KAAK;YAAE,QAAQ,CAAC,KAAK,GAAG,KAAK,CAAC;QAElC,OAAO;YACL,EAAE,EAAE,MAAM,EAAE;YACZ,OAAO;YACP,MAAM;YACN,UAAU,EAAE,KAAK;YACjB,QAAQ;SACT,CAAC;IACJ,CAAC;CACF"}
|
|
@@ -2,12 +2,17 @@ import type { DocumentLoader, RAGDocument } from '@cogitator-ai/types';
|
|
|
2
2
|
export interface WebLoaderOptions {
|
|
3
3
|
selector?: string;
|
|
4
4
|
headers?: Record<string, string>;
|
|
5
|
+
timeoutMs?: number;
|
|
6
|
+
maxResponseBytes?: number;
|
|
5
7
|
}
|
|
6
8
|
export declare class WebLoader implements DocumentLoader {
|
|
7
9
|
readonly supportedTypes: string[];
|
|
8
10
|
private readonly htmlLoader;
|
|
9
11
|
private readonly headers?;
|
|
12
|
+
private readonly timeoutMs;
|
|
13
|
+
private readonly maxResponseBytes;
|
|
10
14
|
constructor(options?: WebLoaderOptions);
|
|
11
15
|
load(source: string): Promise<RAGDocument[]>;
|
|
16
|
+
private readBodyWithLimit;
|
|
12
17
|
}
|
|
13
18
|
//# sourceMappingURL=web-loader.d.ts.map
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"web-loader.d.ts","sourceRoot":"","sources":["../../src/loaders/web-loader.ts"],"names":[],"mappings":"
|
|
1
|
+
{"version":3,"file":"web-loader.d.ts","sourceRoot":"","sources":["../../src/loaders/web-loader.ts"],"names":[],"mappings":"AAEA,OAAO,KAAK,EAAE,cAAc,EAAE,WAAW,EAAE,MAAM,qBAAqB,CAAC;AAoEvE,MAAM,WAAW,gBAAgB;IAC/B,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,OAAO,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC;IACjC,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,gBAAgB,CAAC,EAAE,MAAM,CAAC;CAC3B;AAKD,qBAAa,SAAU,YAAW,cAAc;IAC9C,QAAQ,CAAC,cAAc,WAAqB;IAC5C,OAAO,CAAC,QAAQ,CAAC,UAAU,CAAa;IACxC,OAAO,CAAC,QAAQ,CAAC,OAAO,CAAC,CAAyB;IAClD,OAAO,CAAC,QAAQ,CAAC,SAAS,CAAS;IACnC,OAAO,CAAC,QAAQ,CAAC,gBAAgB,CAAS;gBAE9B,OAAO,CAAC,EAAE,gBAAgB;IAOhC,IAAI,CAAC,MAAM,EAAE,MAAM,GAAG,OAAO,CAAC,WAAW,EAAE,CAAC;YAqCpC,iBAAiB;CA0ChC"}
|
|
@@ -1,22 +1,141 @@
|
|
|
1
|
+
import { lookup } from 'node:dns/promises';
|
|
2
|
+
import { isIP } from 'node:net';
|
|
1
3
|
import { HTMLLoader } from './html-loader.js';
|
|
4
|
+
const BLOCKED_HOSTNAMES = new Set([
|
|
5
|
+
'localhost',
|
|
6
|
+
'127.0.0.1',
|
|
7
|
+
'::1',
|
|
8
|
+
'[::1]',
|
|
9
|
+
'0.0.0.0',
|
|
10
|
+
'169.254.169.254',
|
|
11
|
+
'metadata.google.internal',
|
|
12
|
+
]);
|
|
13
|
+
const MAX_REDIRECTS = 5;
|
|
14
|
+
function isPrivateIP(ip) {
|
|
15
|
+
if (ip.startsWith('10.'))
|
|
16
|
+
return true;
|
|
17
|
+
if (ip.startsWith('192.168.'))
|
|
18
|
+
return true;
|
|
19
|
+
if (ip === '0.0.0.0' || ip === '127.0.0.1' || ip === '::1')
|
|
20
|
+
return true;
|
|
21
|
+
if (ip.startsWith('169.254.'))
|
|
22
|
+
return true;
|
|
23
|
+
if (ip.startsWith('fc') || ip.startsWith('fd') || ip.startsWith('fe80'))
|
|
24
|
+
return true;
|
|
25
|
+
const match172 = /^172\.(\d+)\./.exec(ip);
|
|
26
|
+
if (match172) {
|
|
27
|
+
const second = parseInt(match172[1], 10);
|
|
28
|
+
if (second >= 16 && second <= 31)
|
|
29
|
+
return true;
|
|
30
|
+
}
|
|
31
|
+
return false;
|
|
32
|
+
}
|
|
33
|
+
async function validateUrl(urlString) {
|
|
34
|
+
let url;
|
|
35
|
+
try {
|
|
36
|
+
url = new URL(urlString);
|
|
37
|
+
}
|
|
38
|
+
catch {
|
|
39
|
+
throw new Error(`WebLoader: invalid URL: ${urlString}`);
|
|
40
|
+
}
|
|
41
|
+
if (url.protocol !== 'http:' && url.protocol !== 'https:') {
|
|
42
|
+
throw new Error(`WebLoader: unsupported protocol "${url.protocol}" — only http and https are allowed`);
|
|
43
|
+
}
|
|
44
|
+
const hostname = url.hostname.toLowerCase();
|
|
45
|
+
if (BLOCKED_HOSTNAMES.has(hostname)) {
|
|
46
|
+
throw new Error(`WebLoader: access to "${hostname}" is blocked (loopback / metadata endpoint)`);
|
|
47
|
+
}
|
|
48
|
+
if (isIP(hostname) && isPrivateIP(hostname)) {
|
|
49
|
+
throw new Error(`WebLoader: access to private IP "${hostname}" is blocked`);
|
|
50
|
+
}
|
|
51
|
+
if (!isIP(hostname)) {
|
|
52
|
+
try {
|
|
53
|
+
const { address } = await lookup(hostname);
|
|
54
|
+
if (isPrivateIP(address)) {
|
|
55
|
+
throw new Error(`WebLoader: "${hostname}" resolves to private IP ${address} — blocked`);
|
|
56
|
+
}
|
|
57
|
+
}
|
|
58
|
+
catch (err) {
|
|
59
|
+
if (err instanceof Error && err.message.startsWith('WebLoader:'))
|
|
60
|
+
throw err;
|
|
61
|
+
}
|
|
62
|
+
}
|
|
63
|
+
return url;
|
|
64
|
+
}
|
|
65
|
+
const DEFAULT_TIMEOUT_MS = 30_000;
|
|
66
|
+
const DEFAULT_MAX_RESPONSE_BYTES = 50 * 1024 * 1024;
|
|
2
67
|
export class WebLoader {
|
|
3
68
|
supportedTypes = ['http', 'https'];
|
|
4
69
|
htmlLoader;
|
|
5
70
|
headers;
|
|
71
|
+
timeoutMs;
|
|
72
|
+
maxResponseBytes;
|
|
6
73
|
constructor(options) {
|
|
7
74
|
this.htmlLoader = new HTMLLoader({ selector: options?.selector });
|
|
8
75
|
this.headers = options?.headers;
|
|
76
|
+
this.timeoutMs = options?.timeoutMs ?? DEFAULT_TIMEOUT_MS;
|
|
77
|
+
this.maxResponseBytes = options?.maxResponseBytes ?? DEFAULT_MAX_RESPONSE_BYTES;
|
|
9
78
|
}
|
|
10
79
|
async load(source) {
|
|
11
|
-
|
|
12
|
-
|
|
13
|
-
|
|
14
|
-
|
|
15
|
-
|
|
80
|
+
let currentUrl = source;
|
|
81
|
+
for (let redirectCount = 0; redirectCount <= MAX_REDIRECTS; redirectCount++) {
|
|
82
|
+
await validateUrl(currentUrl);
|
|
83
|
+
const response = await fetch(currentUrl, {
|
|
84
|
+
headers: this.headers ?? {},
|
|
85
|
+
redirect: 'manual',
|
|
86
|
+
signal: AbortSignal.timeout(this.timeoutMs),
|
|
87
|
+
});
|
|
88
|
+
if (response.status >= 300 && response.status < 400) {
|
|
89
|
+
const location = response.headers.get('location');
|
|
90
|
+
if (!location) {
|
|
91
|
+
throw new Error(`WebLoader: redirect ${response.status} without Location header from ${currentUrl}`);
|
|
92
|
+
}
|
|
93
|
+
currentUrl = new URL(location, currentUrl).href;
|
|
94
|
+
continue;
|
|
95
|
+
}
|
|
96
|
+
if (!response.ok) {
|
|
97
|
+
throw new Error(`WebLoader: failed to fetch ${currentUrl}: ${response.status} ${response.statusText}`);
|
|
98
|
+
}
|
|
99
|
+
const html = await this.readBodyWithLimit(response, currentUrl);
|
|
100
|
+
const doc = await this.htmlLoader.parseHTML(html, source, 'web');
|
|
101
|
+
return [doc];
|
|
102
|
+
}
|
|
103
|
+
throw new Error(`WebLoader: too many redirects (max ${MAX_REDIRECTS}) for ${source}`);
|
|
104
|
+
}
|
|
105
|
+
async readBodyWithLimit(response, url) {
|
|
106
|
+
const body = response.body;
|
|
107
|
+
if (!body) {
|
|
108
|
+
const text = await response.text();
|
|
109
|
+
if (text.length > this.maxResponseBytes) {
|
|
110
|
+
throw new Error(`WebLoader: response from ${url} exceeds ${this.maxResponseBytes} bytes limit`);
|
|
111
|
+
}
|
|
112
|
+
return text;
|
|
113
|
+
}
|
|
114
|
+
const reader = body.getReader();
|
|
115
|
+
const chunks = [];
|
|
116
|
+
let totalBytes = 0;
|
|
117
|
+
try {
|
|
118
|
+
for (;;) {
|
|
119
|
+
const { done, value } = await reader.read();
|
|
120
|
+
if (done)
|
|
121
|
+
break;
|
|
122
|
+
totalBytes += value.byteLength;
|
|
123
|
+
if (totalBytes > this.maxResponseBytes) {
|
|
124
|
+
throw new Error(`WebLoader: response from ${url} exceeds ${this.maxResponseBytes} bytes limit`);
|
|
125
|
+
}
|
|
126
|
+
chunks.push(value);
|
|
127
|
+
}
|
|
128
|
+
}
|
|
129
|
+
finally {
|
|
130
|
+
reader.releaseLock();
|
|
131
|
+
}
|
|
132
|
+
const merged = new Uint8Array(totalBytes);
|
|
133
|
+
let offset = 0;
|
|
134
|
+
for (const chunk of chunks) {
|
|
135
|
+
merged.set(chunk, offset);
|
|
136
|
+
offset += chunk.byteLength;
|
|
16
137
|
}
|
|
17
|
-
|
|
18
|
-
const doc = await this.htmlLoader.parseHTML(html, source, 'web');
|
|
19
|
-
return [doc];
|
|
138
|
+
return new TextDecoder('utf-8').decode(merged);
|
|
20
139
|
}
|
|
21
140
|
}
|
|
22
141
|
//# sourceMappingURL=web-loader.js.map
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"web-loader.js","sourceRoot":"","sources":["../../src/loaders/web-loader.ts"],"names":[],"mappings":"
|
|
1
|
+
{"version":3,"file":"web-loader.js","sourceRoot":"","sources":["../../src/loaders/web-loader.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,MAAM,EAAE,MAAM,mBAAmB,CAAC;AAC3C,OAAO,EAAE,IAAI,EAAE,MAAM,UAAU,CAAC;AAEhC,OAAO,EAAE,UAAU,EAAE,MAAM,eAAe,CAAC;AAE3C,MAAM,iBAAiB,GAAG,IAAI,GAAG,CAAC;IAChC,WAAW;IACX,WAAW;IACX,KAAK;IACL,OAAO;IACP,SAAS;IACT,iBAAiB;IACjB,0BAA0B;CAC3B,CAAC,CAAC;AAEH,MAAM,aAAa,GAAG,CAAC,CAAC;AAExB,SAAS,WAAW,CAAC,EAAU;IAC7B,IAAI,EAAE,CAAC,UAAU,CAAC,KAAK,CAAC;QAAE,OAAO,IAAI,CAAC;IACtC,IAAI,EAAE,CAAC,UAAU,CAAC,UAAU,CAAC;QAAE,OAAO,IAAI,CAAC;IAC3C,IAAI,EAAE,KAAK,SAAS,IAAI,EAAE,KAAK,WAAW,IAAI,EAAE,KAAK,KAAK;QAAE,OAAO,IAAI,CAAC;IACxE,IAAI,EAAE,CAAC,UAAU,CAAC,UAAU,CAAC;QAAE,OAAO,IAAI,CAAC;IAC3C,IAAI,EAAE,CAAC,UAAU,CAAC,IAAI,CAAC,IAAI,EAAE,CAAC,UAAU,CAAC,IAAI,CAAC,IAAI,EAAE,CAAC,UAAU,CAAC,MAAM,CAAC;QAAE,OAAO,IAAI,CAAC;IAErF,MAAM,QAAQ,GAAG,eAAe,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IAC1C,IAAI,QAAQ,EAAE,CAAC;QACb,MAAM,MAAM,GAAG,QAAQ,CAAC,QAAQ,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC;QACzC,IAAI,MAAM,IAAI,EAAE,IAAI,MAAM,IAAI,EAAE;YAAE,OAAO,IAAI,CAAC;IAChD,CAAC;IAED,OAAO,KAAK,CAAC;AACf,CAAC;AAED,KAAK,UAAU,WAAW,CAAC,SAAiB;IAC1C,IAAI,GAAQ,CAAC;IACb,IAAI,CAAC;QACH,GAAG,GAAG,IAAI,GAAG,CAAC,SAAS,CAAC,CAAC;IAC3B,CAAC;IAAC,MAAM,CAAC;QACP,MAAM,IAAI,KAAK,CAAC,2BAA2B,SAAS,EAAE,CAAC,CAAC;IAC1D,CAAC;IAED,IAAI,GAAG,CAAC,QAAQ,KAAK,OAAO,IAAI,GAAG,CAAC,QAAQ,KAAK,QAAQ,EAAE,CAAC;QAC1D,MAAM,IAAI,KAAK,CACb,oCAAoC,GAAG,CAAC,QAAQ,qCAAqC,CACtF,CAAC;IACJ,CAAC;IAED,MAAM,QAAQ,GAAG,GAAG,CAAC,QAAQ,CAAC,WAAW,EAAE,CAAC;IAC5C,IAAI,iBAAiB,CAAC,GAAG,CAAC,QAAQ,CAAC,EAAE,CAAC;QACpC,MAAM,IAAI,KAAK,CAAC,yBAAyB,QAAQ,6CAA6C,CAAC,CAAC;IAClG,CAAC;IAED,IAAI,IAAI,CAAC,QAAQ,CAAC,IAAI,WAAW,CAAC,QAAQ,CAAC,EAAE,CAAC;QAC5C,MAAM,IAAI,KAAK,CAAC,oCAAoC,QAAQ,cAAc,CAAC,CAAC;IAC9E,CAAC;IAED,IAAI,CAAC,IAAI,CAAC,QAAQ,CAAC,EAAE,CAAC;QACpB,IAAI,CAAC;YACH,MAAM,EAAE,OAAO,EAAE,GAAG,MAAM,MAAM,CAAC,QAAQ,CAAC,CAAC;YAC3C,IAAI,WAAW,CAAC,OAAO,CAAC,EAAE,CAAC;gBACzB,MAAM,IAAI,KAAK,CAAC,eAAe,QAAQ,4BAA4B,OAAO,YAAY,CAAC,CAAC;YAC1F,CAAC;QACH,CAAC;QAAC,OAAO,GAAG,EAAE,CAAC;YACb,IAAI,GAAG,YAAY,KAAK,IAAI,GAAG,CAAC,OAAO,CAAC,UAAU,CAAC,YAAY,CAAC;gBAAE,MAAM,GAAG,CAAC;QAC9E,CAAC;IACH,CAAC;IAED,OAAO,GAAG,CAAC;AACb,CAAC;AASD,MAAM,kBAAkB,GAAG,MAAM,CAAC;AAClC,MAAM,0BAA0B,GAAG,EAAE,GAAG,IAAI,GAAG,IAAI,CAAC;AAEpD,MAAM,OAAO,SAAS;IACX,cAAc,GAAG,CAAC,MAAM,EAAE,OAAO,CAAC,CAAC;IAC3B,UAAU,CAAa;IACvB,OAAO,CAA0B;IACjC,SAAS,CAAS;IAClB,gBAAgB,CAAS;IAE1C,YAAY,OAA0B;QACpC,IAAI,CAAC,UAAU,GAAG,IAAI,UAAU,CAAC,EAAE,QAAQ,EAAE,OAAO,EAAE,QAAQ,EAAE,CAAC,CAAC;QAClE,IAAI,CAAC,OAAO,GAAG,OAAO,EAAE,OAAO,CAAC;QAChC,IAAI,CAAC,SAAS,GAAG,OAAO,EAAE,SAAS,IAAI,kBAAkB,CAAC;QAC1D,IAAI,CAAC,gBAAgB,GAAG,OAAO,EAAE,gBAAgB,IAAI,0BAA0B,CAAC;IAClF,CAAC;IAED,KAAK,CAAC,IAAI,CAAC,MAAc;QACvB,IAAI,UAAU,GAAG,MAAM,CAAC;QAExB,KAAK,IAAI,aAAa,GAAG,CAAC,EAAE,aAAa,IAAI,aAAa,EAAE,aAAa,EAAE,EAAE,CAAC;YAC5E,MAAM,WAAW,CAAC,UAAU,CAAC,CAAC;YAE9B,MAAM,QAAQ,GAAG,MAAM,KAAK,CAAC,UAAU,EAAE;gBACvC,OAAO,EAAE,IAAI,CAAC,OAAO,IAAI,EAAE;gBAC3B,QAAQ,EAAE,QAAQ;gBAClB,MAAM,EAAE,WAAW,CAAC,OAAO,CAAC,IAAI,CAAC,SAAS,CAAC;aAC5C,CAAC,CAAC;YAEH,IAAI,QAAQ,CAAC,MAAM,IAAI,GAAG,IAAI,QAAQ,CAAC,MAAM,GAAG,GAAG,EAAE,CAAC;gBACpD,MAAM,QAAQ,GAAG,QAAQ,CAAC,OAAO,CAAC,GAAG,CAAC,UAAU,CAAC,CAAC;gBAClD,IAAI,CAAC,QAAQ,EAAE,CAAC;oBACd,MAAM,IAAI,KAAK,CACb,uBAAuB,QAAQ,CAAC,MAAM,iCAAiC,UAAU,EAAE,CACpF,CAAC;gBACJ,CAAC;gBACD,UAAU,GAAG,IAAI,GAAG,CAAC,QAAQ,EAAE,UAAU,CAAC,CAAC,IAAI,CAAC;gBAChD,SAAS;YACX,CAAC;YAED,IAAI,CAAC,QAAQ,CAAC,EAAE,EAAE,CAAC;gBACjB,MAAM,IAAI,KAAK,CACb,8BAA8B,UAAU,KAAK,QAAQ,CAAC,MAAM,IAAI,QAAQ,CAAC,UAAU,EAAE,CACtF,CAAC;YACJ,CAAC;YAED,MAAM,IAAI,GAAG,MAAM,IAAI,CAAC,iBAAiB,CAAC,QAAQ,EAAE,UAAU,CAAC,CAAC;YAChE,MAAM,GAAG,GAAG,MAAM,IAAI,CAAC,UAAU,CAAC,SAAS,CAAC,IAAI,EAAE,MAAM,EAAE,KAAK,CAAC,CAAC;YACjE,OAAO,CAAC,GAAG,CAAC,CAAC;QACf,CAAC;QAED,MAAM,IAAI,KAAK,CAAC,sCAAsC,aAAa,SAAS,MAAM,EAAE,CAAC,CAAC;IACxF,CAAC;IAEO,KAAK,CAAC,iBAAiB,CAAC,QAAkB,EAAE,GAAW;QAC7D,MAAM,IAAI,GAAG,QAAQ,CAAC,IAAI,CAAC;QAC3B,IAAI,CAAC,IAAI,EAAE,CAAC;YACV,MAAM,IAAI,GAAG,MAAM,QAAQ,CAAC,IAAI,EAAE,CAAC;YACnC,IAAI,IAAI,CAAC,MAAM,GAAG,IAAI,CAAC,gBAAgB,EAAE,CAAC;gBACxC,MAAM,IAAI,KAAK,CACb,4BAA4B,GAAG,YAAY,IAAI,CAAC,gBAAgB,cAAc,CAC/E,CAAC;YACJ,CAAC;YACD,OAAO,IAAI,CAAC;QACd,CAAC;QAED,MAAM,MAAM,GAAG,IAAI,CAAC,SAAS,EAAE,CAAC;QAChC,MAAM,MAAM,GAAiB,EAAE,CAAC;QAChC,IAAI,UAAU,GAAG,CAAC,CAAC;QAEnB,IAAI,CAAC;YACH,SAAS,CAAC;gBACR,MAAM,EAAE,IAAI,EAAE,KAAK,EAAE,GAAG,MAAM,MAAM,CAAC,IAAI,EAAE,CAAC;gBAC5C,IAAI,IAAI;oBAAE,MAAM;gBAEhB,UAAU,IAAI,KAAK,CAAC,UAAU,CAAC;gBAC/B,IAAI,UAAU,GAAG,IAAI,CAAC,gBAAgB,EAAE,CAAC;oBACvC,MAAM,IAAI,KAAK,CACb,4BAA4B,GAAG,YAAY,IAAI,CAAC,gBAAgB,cAAc,CAC/E,CAAC;gBACJ,CAAC;gBACD,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;YACrB,CAAC;QACH,CAAC;gBAAS,CAAC;YACT,MAAM,CAAC,WAAW,EAAE,CAAC;QACvB,CAAC;QAED,MAAM,MAAM,GAAG,IAAI,UAAU,CAAC,UAAU,CAAC,CAAC;QAC1C,IAAI,MAAM,GAAG,CAAC,CAAC;QACf,KAAK,MAAM,KAAK,IAAI,MAAM,EAAE,CAAC;YAC3B,MAAM,CAAC,GAAG,CAAC,KAAK,EAAE,MAAM,CAAC,CAAC;YAC1B,MAAM,IAAI,KAAK,CAAC,UAAU,CAAC;QAC7B,CAAC;QAED,OAAO,IAAI,WAAW,CAAC,OAAO,CAAC,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC;IACjD,CAAC;CACF"}
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"rag-builder.d.ts","sourceRoot":"","sources":["../src/rag-builder.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EACV,cAAc,EACd,OAAO,EACP,YAAY,EACZ,gBAAgB,EAChB,gBAAgB,EAChB,SAAS,EACT,QAAQ,EAET,MAAM,qBAAqB,CAAC;AAC7B,OAAO,EAA2B,KAAK,sBAAsB,EAAE,MAAM,UAAU,CAAC;AAChF,OAAO,EAAE,WAAW,EAAE,MAAM,gBAAgB,CAAC;AAI7C,qBAAa,kBAAkB;IAC7B,OAAO,CAAC,MAAM,CAAC,CAAiB;IAChC,OAAO,CAAC,OAAO,CAAC,CAAyB;IACzC,OAAO,CAAC,gBAAgB,CAAC,CAAmB;IAC5C,OAAO,CAAC,gBAAgB,CAAC,CAAmB;IAC5C,OAAO,CAAC,SAAS,CAAC,CAAY;IAC9B,OAAO,CAAC,QAAQ,CAAC,CAAW;IAC5B,OAAO,CAAC,WAAW,CAAC,CAAyB;IAE7C,UAAU,CAAC,MAAM,EAAE,cAAc,GAAG,IAAI;IAKxC,WAAW,CAAC,OAAO,EAAE,OAAO,GAAG,YAAY,GAAG,IAAI;IAKlD,oBAAoB,CAAC,OAAO,EAAE,gBAAgB,GAAG,IAAI;IAKrD,oBAAoB,CAAC,OAAO,EAAE,gBAAgB,GAAG,IAAI;IAKrD,aAAa,CAAC,SAAS,EAAE,SAAS,GAAG,IAAI;IAKzC,YAAY,CAAC,QAAQ,EAAE,QAAQ,GAAG,IAAI;IAKtC,UAAU,CAAC,MAAM,EAAE,sBAAsB,GAAG,IAAI;IAKhD,KAAK,IAAI,WAAW;
|
|
1
|
+
{"version":3,"file":"rag-builder.d.ts","sourceRoot":"","sources":["../src/rag-builder.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EACV,cAAc,EACd,OAAO,EACP,YAAY,EACZ,gBAAgB,EAChB,gBAAgB,EAChB,SAAS,EACT,QAAQ,EAET,MAAM,qBAAqB,CAAC;AAC7B,OAAO,EAA2B,KAAK,sBAAsB,EAAE,MAAM,UAAU,CAAC;AAChF,OAAO,EAAE,WAAW,EAAE,MAAM,gBAAgB,CAAC;AAI7C,qBAAa,kBAAkB;IAC7B,OAAO,CAAC,MAAM,CAAC,CAAiB;IAChC,OAAO,CAAC,OAAO,CAAC,CAAyB;IACzC,OAAO,CAAC,gBAAgB,CAAC,CAAmB;IAC5C,OAAO,CAAC,gBAAgB,CAAC,CAAmB;IAC5C,OAAO,CAAC,SAAS,CAAC,CAAY;IAC9B,OAAO,CAAC,QAAQ,CAAC,CAAW;IAC5B,OAAO,CAAC,WAAW,CAAC,CAAyB;IAE7C,UAAU,CAAC,MAAM,EAAE,cAAc,GAAG,IAAI;IAKxC,WAAW,CAAC,OAAO,EAAE,OAAO,GAAG,YAAY,GAAG,IAAI;IAKlD,oBAAoB,CAAC,OAAO,EAAE,gBAAgB,GAAG,IAAI;IAKrD,oBAAoB,CAAC,OAAO,EAAE,gBAAgB,GAAG,IAAI;IAKrD,aAAa,CAAC,SAAS,EAAE,SAAS,GAAG,IAAI;IAKzC,YAAY,CAAC,QAAQ,EAAE,QAAQ,GAAG,IAAI;IAKtC,UAAU,CAAC,MAAM,EAAE,sBAAsB,GAAG,IAAI;IAKhD,KAAK,IAAI,WAAW;IAoCpB,OAAO,CAAC,aAAa;CAMtB"}
|
package/dist/rag-builder.js
CHANGED
|
@@ -54,6 +54,8 @@ export class RAGPipelineBuilder {
|
|
|
54
54
|
new SimilarityRetriever({
|
|
55
55
|
embeddingAdapter: this.embeddingAdapter,
|
|
56
56
|
embeddingService: this.embeddingService,
|
|
57
|
+
defaultTopK: config.retrieval.topK,
|
|
58
|
+
defaultThreshold: config.retrieval.threshold,
|
|
57
59
|
});
|
|
58
60
|
return new RAGPipeline(config, {
|
|
59
61
|
loader: this.loader,
|
|
@@ -62,7 +64,7 @@ export class RAGPipelineBuilder {
|
|
|
62
64
|
embeddingAdapter: this.embeddingAdapter,
|
|
63
65
|
retriever,
|
|
64
66
|
reranker: this.reranker,
|
|
65
|
-
});
|
|
67
|
+
}, true);
|
|
66
68
|
}
|
|
67
69
|
resolveConfig() {
|
|
68
70
|
if (!this.configInput) {
|
package/dist/rag-builder.js.map
CHANGED
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"rag-builder.js","sourceRoot":"","sources":["../src/rag-builder.ts"],"names":[],"mappings":"AAUA,OAAO,EAAE,uBAAuB,EAA+B,MAAM,UAAU,CAAC;AAChF,OAAO,EAAE,WAAW,EAAE,MAAM,gBAAgB,CAAC;AAC7C,OAAO,EAAE,aAAa,EAAE,MAAM,2BAA2B,CAAC;AAC1D,OAAO,EAAE,mBAAmB,EAAE,MAAM,mCAAmC,CAAC;AAExE,MAAM,OAAO,kBAAkB;IACrB,MAAM,CAAkB;IACxB,OAAO,CAA0B;IACjC,gBAAgB,CAAoB;IACpC,gBAAgB,CAAoB;IACpC,SAAS,CAAa;IACtB,QAAQ,CAAY;IACpB,WAAW,CAA0B;IAE7C,UAAU,CAAC,MAAsB;QAC/B,IAAI,CAAC,MAAM,GAAG,MAAM,CAAC;QACrB,OAAO,IAAI,CAAC;IACd,CAAC;IAED,WAAW,CAAC,OAA+B;QACzC,IAAI,CAAC,OAAO,GAAG,OAAO,CAAC;QACvB,OAAO,IAAI,CAAC;IACd,CAAC;IAED,oBAAoB,CAAC,OAAyB;QAC5C,IAAI,CAAC,gBAAgB,GAAG,OAAO,CAAC;QAChC,OAAO,IAAI,CAAC;IACd,CAAC;IAED,oBAAoB,CAAC,OAAyB;QAC5C,IAAI,CAAC,gBAAgB,GAAG,OAAO,CAAC;QAChC,OAAO,IAAI,CAAC;IACd,CAAC;IAED,aAAa,CAAC,SAAoB;QAChC,IAAI,CAAC,SAAS,GAAG,SAAS,CAAC;QAC3B,OAAO,IAAI,CAAC;IACd,CAAC;IAED,YAAY,CAAC,QAAkB;QAC7B,IAAI,CAAC,QAAQ,GAAG,QAAQ,CAAC;QACzB,OAAO,IAAI,CAAC;IACd,CAAC;IAED,UAAU,CAAC,MAA8B;QACvC,IAAI,CAAC,WAAW,GAAG,MAAM,CAAC;QAC1B,OAAO,IAAI,CAAC;IACd,CAAC;IAED,KAAK;QACH,IAAI,CAAC,IAAI,CAAC,MAAM,EAAE,CAAC;YACjB,MAAM,IAAI,KAAK,CAAC,uDAAuD,CAAC,CAAC;QAC3E,CAAC;QACD,IAAI,CAAC,IAAI,CAAC,gBAAgB,EAAE,CAAC;YAC3B,MAAM,IAAI,KAAK,CAAC,2EAA2E,CAAC,CAAC;QAC/F,CAAC;QACD,IAAI,CAAC,IAAI,CAAC,gBAAgB,EAAE,CAAC;YAC3B,MAAM,IAAI,KAAK,CAAC,2EAA2E,CAAC,CAAC;QAC/F,CAAC;QAED,MAAM,MAAM,GAAG,IAAI,CAAC,aAAa,EAAE,CAAC;QACpC,MAAM,OAAO,GAAG,IAAI,CAAC,OAAO,IAAI,aAAa,CAAC,MAAM,CAAC,QAAQ,EAAE,IAAI,CAAC,gBAAgB,CAAC,CAAC;QACtF,MAAM,SAAS,GACb,IAAI,CAAC,SAAS;YACd,IAAI,mBAAmB,CAAC;gBACtB,gBAAgB,EAAE,IAAI,CAAC,gBAAgB;gBACvC,gBAAgB,EAAE,IAAI,CAAC,gBAAgB;
|
|
1
|
+
{"version":3,"file":"rag-builder.js","sourceRoot":"","sources":["../src/rag-builder.ts"],"names":[],"mappings":"AAUA,OAAO,EAAE,uBAAuB,EAA+B,MAAM,UAAU,CAAC;AAChF,OAAO,EAAE,WAAW,EAAE,MAAM,gBAAgB,CAAC;AAC7C,OAAO,EAAE,aAAa,EAAE,MAAM,2BAA2B,CAAC;AAC1D,OAAO,EAAE,mBAAmB,EAAE,MAAM,mCAAmC,CAAC;AAExE,MAAM,OAAO,kBAAkB;IACrB,MAAM,CAAkB;IACxB,OAAO,CAA0B;IACjC,gBAAgB,CAAoB;IACpC,gBAAgB,CAAoB;IACpC,SAAS,CAAa;IACtB,QAAQ,CAAY;IACpB,WAAW,CAA0B;IAE7C,UAAU,CAAC,MAAsB;QAC/B,IAAI,CAAC,MAAM,GAAG,MAAM,CAAC;QACrB,OAAO,IAAI,CAAC;IACd,CAAC;IAED,WAAW,CAAC,OAA+B;QACzC,IAAI,CAAC,OAAO,GAAG,OAAO,CAAC;QACvB,OAAO,IAAI,CAAC;IACd,CAAC;IAED,oBAAoB,CAAC,OAAyB;QAC5C,IAAI,CAAC,gBAAgB,GAAG,OAAO,CAAC;QAChC,OAAO,IAAI,CAAC;IACd,CAAC;IAED,oBAAoB,CAAC,OAAyB;QAC5C,IAAI,CAAC,gBAAgB,GAAG,OAAO,CAAC;QAChC,OAAO,IAAI,CAAC;IACd,CAAC;IAED,aAAa,CAAC,SAAoB;QAChC,IAAI,CAAC,SAAS,GAAG,SAAS,CAAC;QAC3B,OAAO,IAAI,CAAC;IACd,CAAC;IAED,YAAY,CAAC,QAAkB;QAC7B,IAAI,CAAC,QAAQ,GAAG,QAAQ,CAAC;QACzB,OAAO,IAAI,CAAC;IACd,CAAC;IAED,UAAU,CAAC,MAA8B;QACvC,IAAI,CAAC,WAAW,GAAG,MAAM,CAAC;QAC1B,OAAO,IAAI,CAAC;IACd,CAAC;IAED,KAAK;QACH,IAAI,CAAC,IAAI,CAAC,MAAM,EAAE,CAAC;YACjB,MAAM,IAAI,KAAK,CAAC,uDAAuD,CAAC,CAAC;QAC3E,CAAC;QACD,IAAI,CAAC,IAAI,CAAC,gBAAgB,EAAE,CAAC;YAC3B,MAAM,IAAI,KAAK,CAAC,2EAA2E,CAAC,CAAC;QAC/F,CAAC;QACD,IAAI,CAAC,IAAI,CAAC,gBAAgB,EAAE,CAAC;YAC3B,MAAM,IAAI,KAAK,CAAC,2EAA2E,CAAC,CAAC;QAC/F,CAAC;QAED,MAAM,MAAM,GAAG,IAAI,CAAC,aAAa,EAAE,CAAC;QACpC,MAAM,OAAO,GAAG,IAAI,CAAC,OAAO,IAAI,aAAa,CAAC,MAAM,CAAC,QAAQ,EAAE,IAAI,CAAC,gBAAgB,CAAC,CAAC;QACtF,MAAM,SAAS,GACb,IAAI,CAAC,SAAS;YACd,IAAI,mBAAmB,CAAC;gBACtB,gBAAgB,EAAE,IAAI,CAAC,gBAAgB;gBACvC,gBAAgB,EAAE,IAAI,CAAC,gBAAgB;gBACvC,WAAW,EAAE,MAAM,CAAC,SAAS,CAAC,IAAI;gBAClC,gBAAgB,EAAE,MAAM,CAAC,SAAS,CAAC,SAAS;aAC7C,CAAC,CAAC;QAEL,OAAO,IAAI,WAAW,CACpB,MAAM,EACN;YACE,MAAM,EAAE,IAAI,CAAC,MAAM;YACnB,OAAO;YACP,gBAAgB,EAAE,IAAI,CAAC,gBAAgB;YACvC,gBAAgB,EAAE,IAAI,CAAC,gBAAgB;YACvC,SAAS;YACT,QAAQ,EAAE,IAAI,CAAC,QAAQ;SACxB,EACD,IAAI,CACL,CAAC;IACJ,CAAC;IAEO,aAAa;QACnB,IAAI,CAAC,IAAI,CAAC,WAAW,EAAE,CAAC;YACtB,MAAM,IAAI,KAAK,CAAC,uDAAuD,CAAC,CAAC;QAC3E,CAAC;QACD,OAAO,uBAAuB,CAAC,KAAK,CAAC,IAAI,CAAC,WAAW,CAAC,CAAC;IACzD,CAAC;CACF"}
|
package/dist/rag-pipeline.d.ts
CHANGED
|
@@ -11,7 +11,7 @@ export declare class RAGPipeline {
|
|
|
11
11
|
private readonly config;
|
|
12
12
|
private readonly deps;
|
|
13
13
|
private stats;
|
|
14
|
-
constructor(config: RAGPipelineConfig, deps: RAGPipelineDeps);
|
|
14
|
+
constructor(config: RAGPipelineConfig, deps: RAGPipelineDeps, alreadyParsed?: boolean);
|
|
15
15
|
ingest(source: string): Promise<{
|
|
16
16
|
documents: number;
|
|
17
17
|
chunks: number;
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"rag-pipeline.d.ts","sourceRoot":"","sources":["../src/rag-pipeline.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EACV,cAAc,EACd,OAAO,EACP,YAAY,EACZ,gBAAgB,EAChB,gBAAgB,EAChB,SAAS,EACT,QAAQ,EACR,iBAAiB,EACjB,eAAe,EACf,eAAe,EAGhB,MAAM,qBAAqB,CAAC;AAG7B,MAAM,WAAW,eAAe;IAC9B,MAAM,EAAE,cAAc,CAAC;IACvB,OAAO,EAAE,OAAO,GAAG,YAAY,CAAC;IAChC,gBAAgB,EAAE,gBAAgB,CAAC;IACnC,gBAAgB,EAAE,gBAAgB,CAAC;IACnC,SAAS,EAAE,SAAS,CAAC;IACrB,QAAQ,CAAC,EAAE,QAAQ,CAAC;CACrB;AAED,qBAAa,WAAW;IACtB,OAAO,CAAC,QAAQ,CAAC,MAAM,CAAoB;IAC3C,OAAO,CAAC,QAAQ,CAAC,IAAI,CAAkB;IACvC,OAAO,CAAC,KAAK,CAAkE;gBAEnE,MAAM,EAAE,iBAAiB,EAAE,IAAI,EAAE,eAAe;
|
|
1
|
+
{"version":3,"file":"rag-pipeline.d.ts","sourceRoot":"","sources":["../src/rag-pipeline.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EACV,cAAc,EACd,OAAO,EACP,YAAY,EACZ,gBAAgB,EAChB,gBAAgB,EAChB,SAAS,EACT,QAAQ,EACR,iBAAiB,EACjB,eAAe,EACf,eAAe,EAGhB,MAAM,qBAAqB,CAAC;AAG7B,MAAM,WAAW,eAAe;IAC9B,MAAM,EAAE,cAAc,CAAC;IACvB,OAAO,EAAE,OAAO,GAAG,YAAY,CAAC;IAChC,gBAAgB,EAAE,gBAAgB,CAAC;IACnC,gBAAgB,EAAE,gBAAgB,CAAC;IACnC,SAAS,EAAE,SAAS,CAAC;IACrB,QAAQ,CAAC,EAAE,QAAQ,CAAC;CACrB;AAED,qBAAa,WAAW;IACtB,OAAO,CAAC,QAAQ,CAAC,MAAM,CAAoB;IAC3C,OAAO,CAAC,QAAQ,CAAC,IAAI,CAAkB;IACvC,OAAO,CAAC,KAAK,CAAkE;gBAEnE,MAAM,EAAE,iBAAiB,EAAE,IAAI,EAAE,eAAe,EAAE,aAAa,UAAQ;IAK7E,MAAM,CAAC,MAAM,EAAE,MAAM,GAAG,OAAO,CAAC;QAAE,SAAS,EAAE,MAAM,CAAC;QAAC,MAAM,EAAE,MAAM,CAAA;KAAE,CAAC;IA4BtE,KAAK,CAAC,IAAI,EAAE,MAAM,EAAE,OAAO,CAAC,EAAE,OAAO,CAAC,eAAe,CAAC,GAAG,OAAO,CAAC,eAAe,EAAE,CAAC;IAYzF,QAAQ;;;;;YAIM,aAAa;YAIb,UAAU;CAkBzB"}
|
package/dist/rag-pipeline.js
CHANGED
|
@@ -3,8 +3,8 @@ export class RAGPipeline {
|
|
|
3
3
|
config;
|
|
4
4
|
deps;
|
|
5
5
|
stats = { documentsIngested: 0, chunksStored: 0, queriesProcessed: 0 };
|
|
6
|
-
constructor(config, deps) {
|
|
7
|
-
this.config = RAGPipelineConfigSchema.parse(config);
|
|
6
|
+
constructor(config, deps, alreadyParsed = false) {
|
|
7
|
+
this.config = alreadyParsed ? config : RAGPipelineConfigSchema.parse(config);
|
|
8
8
|
this.deps = deps;
|
|
9
9
|
}
|
|
10
10
|
async ingest(source) {
|
|
@@ -19,9 +19,7 @@ export class RAGPipeline {
|
|
|
19
19
|
if (vectors.length !== chunks.length) {
|
|
20
20
|
throw new Error(`Embedding count mismatch: got ${vectors.length} vectors for ${chunks.length} chunks`);
|
|
21
21
|
}
|
|
22
|
-
|
|
23
|
-
await this.storeChunk(chunks[i], vectors[i], doc);
|
|
24
|
-
}
|
|
22
|
+
await Promise.all(chunks.map((chunk, i) => this.storeChunk(chunk, vectors[i], doc)));
|
|
25
23
|
totalChunks += chunks.length;
|
|
26
24
|
}
|
|
27
25
|
this.stats.documentsIngested += documents.length;
|
|
@@ -29,7 +27,8 @@ export class RAGPipeline {
|
|
|
29
27
|
return { documents: documents.length, chunks: totalChunks };
|
|
30
28
|
}
|
|
31
29
|
async query(text, options) {
|
|
32
|
-
const
|
|
30
|
+
const merged = { ...this.config.retrieval, ...options };
|
|
31
|
+
const results = await this.deps.retriever.retrieve(text, merged);
|
|
33
32
|
this.stats.queriesProcessed++;
|
|
34
33
|
if (this.deps.reranker && this.config.reranking?.enabled) {
|
|
35
34
|
return this.deps.reranker.rerank(text, results, this.config.reranking.topN);
|
package/dist/rag-pipeline.js.map
CHANGED
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"rag-pipeline.js","sourceRoot":"","sources":["../src/rag-pipeline.ts"],"names":[],"mappings":"AAcA,OAAO,EAAE,uBAAuB,EAAE,MAAM,UAAU,CAAC;AAWnD,MAAM,OAAO,WAAW;IACL,MAAM,CAAoB;IAC1B,IAAI,CAAkB;IAC/B,KAAK,GAAG,EAAE,iBAAiB,EAAE,CAAC,EAAE,YAAY,EAAE,CAAC,EAAE,gBAAgB,EAAE,CAAC,EAAE,CAAC;IAE/E,YAAY,MAAyB,EAAE,IAAqB;
|
|
1
|
+
{"version":3,"file":"rag-pipeline.js","sourceRoot":"","sources":["../src/rag-pipeline.ts"],"names":[],"mappings":"AAcA,OAAO,EAAE,uBAAuB,EAAE,MAAM,UAAU,CAAC;AAWnD,MAAM,OAAO,WAAW;IACL,MAAM,CAAoB;IAC1B,IAAI,CAAkB;IAC/B,KAAK,GAAG,EAAE,iBAAiB,EAAE,CAAC,EAAE,YAAY,EAAE,CAAC,EAAE,gBAAgB,EAAE,CAAC,EAAE,CAAC;IAE/E,YAAY,MAAyB,EAAE,IAAqB,EAAE,aAAa,GAAG,KAAK;QACjF,IAAI,CAAC,MAAM,GAAG,aAAa,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,uBAAuB,CAAC,KAAK,CAAC,MAAM,CAAC,CAAC;QAC7E,IAAI,CAAC,IAAI,GAAG,IAAI,CAAC;IACnB,CAAC;IAED,KAAK,CAAC,MAAM,CAAC,MAAc;QACzB,MAAM,SAAS,GAAG,MAAM,IAAI,CAAC,IAAI,CAAC,MAAM,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;QACtD,IAAI,WAAW,GAAG,CAAC,CAAC;QAEpB,KAAK,MAAM,GAAG,IAAI,SAAS,EAAE,CAAC;YAC5B,MAAM,MAAM,GAAG,MAAM,IAAI,CAAC,aAAa,CAAC,GAAG,CAAC,CAAC;YAC7C,IAAI,MAAM,CAAC,MAAM,KAAK,CAAC;gBAAE,SAAS;YAElC,MAAM,KAAK,GAAG,MAAM,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC;YAC3C,MAAM,OAAO,GAAG,MAAM,IAAI,CAAC,IAAI,CAAC,gBAAgB,CAAC,UAAU,CAAC,KAAK,CAAC,CAAC;YAEnE,IAAI,OAAO,CAAC,MAAM,KAAK,MAAM,CAAC,MAAM,EAAE,CAAC;gBACrC,MAAM,IAAI,KAAK,CACb,iCAAiC,OAAO,CAAC,MAAM,gBAAgB,MAAM,CAAC,MAAM,SAAS,CACtF,CAAC;YACJ,CAAC;YAED,MAAM,OAAO,CAAC,GAAG,CAAC,MAAM,CAAC,GAAG,CAAC,CAAC,KAAK,EAAE,CAAC,EAAE,EAAE,CAAC,IAAI,CAAC,UAAU,CAAC,KAAK,EAAE,OAAO,CAAC,CAAC,CAAC,EAAE,GAAG,CAAC,CAAC,CAAC,CAAC;YAErF,WAAW,IAAI,MAAM,CAAC,MAAM,CAAC;QAC/B,CAAC;QAED,IAAI,CAAC,KAAK,CAAC,iBAAiB,IAAI,SAAS,CAAC,MAAM,CAAC;QACjD,IAAI,CAAC,KAAK,CAAC,YAAY,IAAI,WAAW,CAAC;QAEvC,OAAO,EAAE,SAAS,EAAE,SAAS,CAAC,MAAM,EAAE,MAAM,EAAE,WAAW,EAAE,CAAC;IAC9D,CAAC;IAED,KAAK,CAAC,KAAK,CAAC,IAAY,EAAE,OAAkC;QAC1D,MAAM,MAAM,GAA6B,EAAE,GAAG,IAAI,CAAC,MAAM,CAAC,SAAS,EAAE,GAAG,OAAO,EAAE,CAAC;QAClF,MAAM,OAAO,GAAG,MAAM,IAAI,CAAC,IAAI,CAAC,SAAS,CAAC,QAAQ,CAAC,IAAI,EAAE,MAAM,CAAC,CAAC;QACjE,IAAI,CAAC,KAAK,CAAC,gBAAgB,EAAE,CAAC;QAE9B,IAAI,IAAI,CAAC,IAAI,CAAC,QAAQ,IAAI,IAAI,CAAC,MAAM,CAAC,SAAS,EAAE,OAAO,EAAE,CAAC;YACzD,OAAO,IAAI,CAAC,IAAI,CAAC,QAAQ,CAAC,MAAM,CAAC,IAAI,EAAE,OAAO,EAAE,IAAI,CAAC,MAAM,CAAC,SAAS,CAAC,IAAI,CAAC,CAAC;QAC9E,CAAC;QAED,OAAO,OAAO,CAAC;IACjB,CAAC;IAED,QAAQ;QACN,OAAO,EAAE,GAAG,IAAI,CAAC,KAAK,EAAE,CAAC;IAC3B,CAAC;IAEO,KAAK,CAAC,aAAa,CAAC,GAAgB;QAC1C,OAAO,IAAI,CAAC,IAAI,CAAC,OAAO,CAAC,KAAK,CAAC,GAAG,CAAC,OAAO,EAAE,GAAG,CAAC,EAAE,CAAC,CAAC;IACtD,CAAC;IAEO,KAAK,CAAC,UAAU,CAAC,KAAoB,EAAE,MAAgB,EAAE,GAAgB;QAC/E,MAAM,MAAM,GAAG,MAAM,IAAI,CAAC,IAAI,CAAC,gBAAgB,CAAC,YAAY,CAAC;YAC3D,QAAQ,EAAE,KAAK,CAAC,EAAE;YAClB,UAAU,EAAE,UAAU;YACtB,MAAM;YACN,OAAO,EAAE,KAAK,CAAC,OAAO;YACtB,QAAQ,EAAE;gBACR,GAAG,KAAK,CAAC,QAAQ;gBACjB,UAAU,EAAE,GAAG,CAAC,EAAE;gBAClB,MAAM,EAAE,GAAG,CAAC,MAAM;gBAClB,KAAK,EAAE,KAAK,CAAC,KAAK;aACnB;SACF,CAAC,CAAC;QAEH,IAAI,CAAC,MAAM,CAAC,OAAO,EAAE,CAAC;YACpB,MAAM,IAAI,KAAK,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC;QAChC,CAAC;IACH,CAAC;CACF"}
|
|
@@ -4,8 +4,8 @@ export interface CohereRerankerConfig {
|
|
|
4
4
|
model?: string;
|
|
5
5
|
}
|
|
6
6
|
export declare class CohereReranker implements Reranker {
|
|
7
|
-
private apiKey;
|
|
8
|
-
private model;
|
|
7
|
+
private readonly apiKey;
|
|
8
|
+
private readonly model;
|
|
9
9
|
constructor(config: CohereRerankerConfig);
|
|
10
10
|
rerank(query: string, results: RetrievalResult[], topN?: number): Promise<RetrievalResult[]>;
|
|
11
11
|
}
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"cohere-reranker.d.ts","sourceRoot":"","sources":["../../src/rerankers/cohere-reranker.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,QAAQ,EAAE,eAAe,EAAE,MAAM,qBAAqB,CAAC;AAErE,MAAM,WAAW,oBAAoB;IACnC,MAAM,EAAE,MAAM,CAAC;IACf,KAAK,CAAC,EAAE,MAAM,CAAC;CAChB;
|
|
1
|
+
{"version":3,"file":"cohere-reranker.d.ts","sourceRoot":"","sources":["../../src/rerankers/cohere-reranker.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,QAAQ,EAAE,eAAe,EAAE,MAAM,qBAAqB,CAAC;AAErE,MAAM,WAAW,oBAAoB;IACnC,MAAM,EAAE,MAAM,CAAC;IACf,KAAK,CAAC,EAAE,MAAM,CAAC;CAChB;AAaD,qBAAa,cAAe,YAAW,QAAQ;IAC7C,OAAO,CAAC,QAAQ,CAAC,MAAM,CAAS;IAChC,OAAO,CAAC,QAAQ,CAAC,KAAK,CAAS;gBAEnB,MAAM,EAAE,oBAAoB;IAKlC,MAAM,CACV,KAAK,EAAE,MAAM,EACb,OAAO,EAAE,eAAe,EAAE,EAC1B,IAAI,CAAC,EAAE,MAAM,GACZ,OAAO,CAAC,eAAe,EAAE,CAAC;CAwC9B"}
|
|
@@ -1,5 +1,6 @@
|
|
|
1
1
|
const DEFAULT_MODEL = 'rerank-v3.5';
|
|
2
2
|
const COHERE_RERANK_URL = 'https://api.cohere.com/v2/rerank';
|
|
3
|
+
const DEFAULT_TIMEOUT_MS = 30_000;
|
|
3
4
|
export class CohereReranker {
|
|
4
5
|
apiKey;
|
|
5
6
|
model;
|
|
@@ -23,12 +24,16 @@ export class CohereReranker {
|
|
|
23
24
|
documents: results.map((r) => r.content),
|
|
24
25
|
top_n: n,
|
|
25
26
|
}),
|
|
27
|
+
signal: AbortSignal.timeout(DEFAULT_TIMEOUT_MS),
|
|
26
28
|
});
|
|
27
29
|
if (!response.ok) {
|
|
28
30
|
const text = await response.text();
|
|
29
31
|
throw new Error(`Cohere rerank failed (${response.status}): ${text}`);
|
|
30
32
|
}
|
|
31
33
|
const data = (await response.json());
|
|
34
|
+
if (!Array.isArray(data.results)) {
|
|
35
|
+
throw new Error('Cohere rerank returned unexpected response: missing results array');
|
|
36
|
+
}
|
|
32
37
|
const mapped = data.results
|
|
33
38
|
.filter(({ index }) => index >= 0 && index < results.length)
|
|
34
39
|
.map(({ index, relevance_score }) => ({
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"cohere-reranker.js","sourceRoot":"","sources":["../../src/rerankers/cohere-reranker.ts"],"names":[],"mappings":"AAcA,MAAM,aAAa,GAAG,aAAa,CAAC;AACpC,MAAM,iBAAiB,GAAG,kCAAkC,CAAC;
|
|
1
|
+
{"version":3,"file":"cohere-reranker.js","sourceRoot":"","sources":["../../src/rerankers/cohere-reranker.ts"],"names":[],"mappings":"AAcA,MAAM,aAAa,GAAG,aAAa,CAAC;AACpC,MAAM,iBAAiB,GAAG,kCAAkC,CAAC;AAC7D,MAAM,kBAAkB,GAAG,MAAM,CAAC;AAElC,MAAM,OAAO,cAAc;IACR,MAAM,CAAS;IACf,KAAK,CAAS;IAE/B,YAAY,MAA4B;QACtC,IAAI,CAAC,MAAM,GAAG,MAAM,CAAC,MAAM,CAAC;QAC5B,IAAI,CAAC,KAAK,GAAG,MAAM,CAAC,KAAK,IAAI,aAAa,CAAC;IAC7C,CAAC;IAED,KAAK,CAAC,MAAM,CACV,KAAa,EACb,OAA0B,EAC1B,IAAa;QAEb,IAAI,OAAO,CAAC,MAAM,KAAK,CAAC;YAAE,OAAO,EAAE,CAAC;QAEpC,MAAM,CAAC,GAAG,IAAI,IAAI,OAAO,CAAC,MAAM,CAAC;QAEjC,MAAM,QAAQ,GAAG,MAAM,KAAK,CAAC,iBAAiB,EAAE;YAC9C,MAAM,EAAE,MAAM;YACd,OAAO,EAAE;gBACP,cAAc,EAAE,kBAAkB;gBAClC,aAAa,EAAE,UAAU,IAAI,CAAC,MAAM,EAAE;aACvC;YACD,IAAI,EAAE,IAAI,CAAC,SAAS,CAAC;gBACnB,KAAK,EAAE,IAAI,CAAC,KAAK;gBACjB,KAAK;gBACL,SAAS,EAAE,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,OAAO,CAAC;gBACxC,KAAK,EAAE,CAAC;aACT,CAAC;YACF,MAAM,EAAE,WAAW,CAAC,OAAO,CAAC,kBAAkB,CAAC;SAChD,CAAC,CAAC;QAEH,IAAI,CAAC,QAAQ,CAAC,EAAE,EAAE,CAAC;YACjB,MAAM,IAAI,GAAG,MAAM,QAAQ,CAAC,IAAI,EAAE,CAAC;YACnC,MAAM,IAAI,KAAK,CAAC,yBAAyB,QAAQ,CAAC,MAAM,MAAM,IAAI,EAAE,CAAC,CAAC;QACxE,CAAC;QAED,MAAM,IAAI,GAAG,CAAC,MAAM,QAAQ,CAAC,IAAI,EAAE,CAAyB,CAAC;QAE7D,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,IAAI,CAAC,OAAO,CAAC,EAAE,CAAC;YACjC,MAAM,IAAI,KAAK,CAAC,mEAAmE,CAAC,CAAC;QACvF,CAAC;QAED,MAAM,MAAM,GAAG,IAAI,CAAC,OAAO;aACxB,MAAM,CAAC,CAAC,EAAE,KAAK,EAAE,EAAE,EAAE,CAAC,KAAK,IAAI,CAAC,IAAI,KAAK,GAAG,OAAO,CAAC,MAAM,CAAC;aAC3D,GAAG,CAAC,CAAC,EAAE,KAAK,EAAE,eAAe,EAAE,EAAE,EAAE,CAAC,CAAC;YACpC,GAAG,OAAO,CAAC,KAAK,CAAC;YACjB,KAAK,EAAE,eAAe;SACvB,CAAC,CAAC,CAAC;QAEN,OAAO,MAAM,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC;IAC5B,CAAC;CACF"}
|
|
@@ -3,7 +3,7 @@ export interface LLMRerankerConfig {
|
|
|
3
3
|
generateFn: (prompt: string) => Promise<string>;
|
|
4
4
|
}
|
|
5
5
|
export declare class LLMReranker implements Reranker {
|
|
6
|
-
private generateFn;
|
|
6
|
+
private readonly generateFn;
|
|
7
7
|
constructor(config: LLMRerankerConfig);
|
|
8
8
|
rerank(query: string, results: RetrievalResult[], topN?: number): Promise<RetrievalResult[]>;
|
|
9
9
|
private buildPrompt;
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"llm-reranker.d.ts","sourceRoot":"","sources":["../../src/rerankers/llm-reranker.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,QAAQ,EAAE,eAAe,EAAE,MAAM,qBAAqB,CAAC;AAErE,MAAM,WAAW,iBAAiB;IAChC,UAAU,EAAE,CAAC,MAAM,EAAE,MAAM,KAAK,OAAO,CAAC,MAAM,CAAC,CAAC;CACjD;
|
|
1
|
+
{"version":3,"file":"llm-reranker.d.ts","sourceRoot":"","sources":["../../src/rerankers/llm-reranker.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,QAAQ,EAAE,eAAe,EAAE,MAAM,qBAAqB,CAAC;AAErE,MAAM,WAAW,iBAAiB;IAChC,UAAU,EAAE,CAAC,MAAM,EAAE,MAAM,KAAK,OAAO,CAAC,MAAM,CAAC,CAAC;CACjD;AAID,qBAAa,WAAY,YAAW,QAAQ;IAC1C,OAAO,CAAC,QAAQ,CAAC,UAAU,CAAsC;gBAErD,MAAM,EAAE,iBAAiB;IAI/B,MAAM,CACV,KAAK,EAAE,MAAM,EACb,OAAO,EAAE,eAAe,EAAE,EAC1B,IAAI,CAAC,EAAE,MAAM,GACZ,OAAO,CAAC,eAAe,EAAE,CAAC;IA2B7B,OAAO,CAAC,WAAW;IAenB,OAAO,CAAC,WAAW;CA2BpB"}
|
|
@@ -1,3 +1,4 @@
|
|
|
1
|
+
const MAX_DOCUMENTS_IN_PROMPT = 50;
|
|
1
2
|
export class LLMReranker {
|
|
2
3
|
generateFn;
|
|
3
4
|
constructor(config) {
|
|
@@ -21,13 +22,15 @@ export class LLMReranker {
|
|
|
21
22
|
}));
|
|
22
23
|
return topN ? reranked.slice(0, topN) : reranked;
|
|
23
24
|
}
|
|
24
|
-
catch {
|
|
25
|
+
catch (error) {
|
|
26
|
+
console.warn('[LLMReranker] Reranking failed, returning original order:', error);
|
|
25
27
|
const fallback = [...results];
|
|
26
28
|
return topN ? fallback.slice(0, topN) : fallback;
|
|
27
29
|
}
|
|
28
30
|
}
|
|
29
31
|
buildPrompt(query, results) {
|
|
30
|
-
const
|
|
32
|
+
const limited = results.slice(0, MAX_DOCUMENTS_IN_PROMPT);
|
|
33
|
+
const docs = limited.map((r, i) => `[${i}] ${r.content}`).join('\n\n');
|
|
31
34
|
return [
|
|
32
35
|
"Score each document's relevance to the query on a scale of 0-10.",
|
|
33
36
|
'Return ONLY a JSON array: [{ "index": number, "score": number }]',
|
|
@@ -39,18 +42,28 @@ export class LLMReranker {
|
|
|
39
42
|
].join('\n');
|
|
40
43
|
}
|
|
41
44
|
parseScores(response, count) {
|
|
42
|
-
const
|
|
43
|
-
if (!
|
|
45
|
+
const candidates = response.match(/\[[\s\S]*?\]/g);
|
|
46
|
+
if (!candidates)
|
|
44
47
|
throw new Error('No JSON array found in response');
|
|
45
|
-
|
|
46
|
-
|
|
47
|
-
|
|
48
|
-
|
|
49
|
-
|
|
50
|
-
|
|
51
|
-
|
|
52
|
-
|
|
53
|
-
|
|
48
|
+
for (let i = candidates.length - 1; i >= 0; i--) {
|
|
49
|
+
try {
|
|
50
|
+
const parsed = JSON.parse(candidates[i]);
|
|
51
|
+
if (!Array.isArray(parsed))
|
|
52
|
+
continue;
|
|
53
|
+
const scores = parsed.filter((item) => typeof item === 'object' &&
|
|
54
|
+
item !== null &&
|
|
55
|
+
typeof item.index === 'number' &&
|
|
56
|
+
typeof item.score === 'number' &&
|
|
57
|
+
item.index >= 0 &&
|
|
58
|
+
item.index < count);
|
|
59
|
+
if (scores.length > 0)
|
|
60
|
+
return scores;
|
|
61
|
+
}
|
|
62
|
+
catch {
|
|
63
|
+
continue;
|
|
64
|
+
}
|
|
65
|
+
}
|
|
66
|
+
throw new Error('No valid JSON array found in response');
|
|
54
67
|
}
|
|
55
68
|
}
|
|
56
69
|
//# sourceMappingURL=llm-reranker.js.map
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"llm-reranker.js","sourceRoot":"","sources":["../../src/rerankers/llm-reranker.ts"],"names":[],"mappings":"AAMA,MAAM,OAAO,WAAW;
|
|
1
|
+
{"version":3,"file":"llm-reranker.js","sourceRoot":"","sources":["../../src/rerankers/llm-reranker.ts"],"names":[],"mappings":"AAMA,MAAM,uBAAuB,GAAG,EAAE,CAAC;AAEnC,MAAM,OAAO,WAAW;IACL,UAAU,CAAsC;IAEjE,YAAY,MAAyB;QACnC,IAAI,CAAC,UAAU,GAAG,MAAM,CAAC,UAAU,CAAC;IACtC,CAAC;IAED,KAAK,CAAC,MAAM,CACV,KAAa,EACb,OAA0B,EAC1B,IAAa;QAEb,IAAI,OAAO,CAAC,MAAM,KAAK,CAAC;YAAE,OAAO,EAAE,CAAC;QAEpC,IAAI,CAAC;YACH,MAAM,MAAM,GAAG,IAAI,CAAC,WAAW,CAAC,KAAK,EAAE,OAAO,CAAC,CAAC;YAChD,MAAM,QAAQ,GAAG,MAAM,IAAI,CAAC,UAAU,CAAC,MAAM,CAAC,CAAC;YAC/C,MAAM,MAAM,GAAG,IAAI,CAAC,WAAW,CAAC,QAAQ,EAAE,OAAO,CAAC,MAAM,CAAC,CAAC;YAE1D,MAAM,MAAM,GAAG,OAAO,CAAC,GAAG,CAAC,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE;gBACvC,MAAM,KAAK,GAAG,MAAM,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,KAAK,KAAK,CAAC,CAAC,CAAC;gBAChD,OAAO,EAAE,MAAM,EAAE,QAAQ,EAAE,KAAK,EAAE,KAAK,IAAI,CAAC,EAAE,CAAC;YACjD,CAAC,CAAC,CAAC;YAEH,MAAM,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,QAAQ,GAAG,CAAC,CAAC,QAAQ,CAAC,CAAC;YAC/C,MAAM,QAAQ,GAAG,MAAM,CAAC,GAAG,CAAC,CAAC,EAAE,MAAM,EAAE,QAAQ,EAAE,EAAE,EAAE,CAAC,CAAC;gBACrD,GAAG,MAAM;gBACT,KAAK,EAAE,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,QAAQ,GAAG,EAAE,CAAC,CAAC;aAC/C,CAAC,CAAC,CAAC;YAEJ,OAAO,IAAI,CAAC,CAAC,CAAC,QAAQ,CAAC,KAAK,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,CAAC,CAAC,QAAQ,CAAC;QACnD,CAAC;QAAC,OAAO,KAAK,EAAE,CAAC;YACf,OAAO,CAAC,IAAI,CAAC,2DAA2D,EAAE,KAAK,CAAC,CAAC;YACjF,MAAM,QAAQ,GAAG,CAAC,GAAG,OAAO,CAAC,CAAC;YAC9B,OAAO,IAAI,CAAC,CAAC,CAAC,QAAQ,CAAC,KAAK,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,CAAC,CAAC,QAAQ,CAAC;QACnD,CAAC;IACH,CAAC;IAEO,WAAW,CAAC,KAAa,EAAE,OAA0B;QAC3D,MAAM,OAAO,GAAG,OAAO,CAAC,KAAK,CAAC,CAAC,EAAE,uBAAuB,CAAC,CAAC;QAC1D,MAAM,IAAI,GAAG,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,OAAO,EAAE,CAAC,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;QAEvE,OAAO;YACL,kEAAkE;YAClE,kEAAkE;YAClE,EAAE;YACF,UAAU,KAAK,EAAE;YACjB,EAAE;YACF,YAAY;YACZ,IAAI;SACL,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;IACf,CAAC;IAEO,WAAW,CAAC,QAAgB,EAAE,KAAa;QACjD,MAAM,UAAU,GAAG,QAAQ,CAAC,KAAK,CAAC,eAAe,CAAC,CAAC;QACnD,IAAI,CAAC,UAAU;YAAE,MAAM,IAAI,KAAK,CAAC,iCAAiC,CAAC,CAAC;QAEpE,KAAK,IAAI,CAAC,GAAG,UAAU,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC,IAAI,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC;YAChD,IAAI,CAAC;gBACH,MAAM,MAAM,GAAY,IAAI,CAAC,KAAK,CAAC,UAAU,CAAC,CAAC,CAAC,CAAC,CAAC;gBAClD,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,MAAM,CAAC;oBAAE,SAAS;gBAErC,MAAM,MAAM,GAAG,MAAM,CAAC,MAAM,CAC1B,CAAC,IAAI,EAA4C,EAAE,CACjD,OAAO,IAAI,KAAK,QAAQ;oBACxB,IAAI,KAAK,IAAI;oBACb,OAAQ,IAAgC,CAAC,KAAK,KAAK,QAAQ;oBAC3D,OAAQ,IAAgC,CAAC,KAAK,KAAK,QAAQ;oBAC1D,IAA0B,CAAC,KAAK,IAAI,CAAC;oBACrC,IAA0B,CAAC,KAAK,GAAG,KAAK,CAC5C,CAAC;gBAEF,IAAI,MAAM,CAAC,MAAM,GAAG,CAAC;oBAAE,OAAO,MAAM,CAAC;YACvC,CAAC;YAAC,MAAM,CAAC;gBACP,SAAS;YACX,CAAC;QACH,CAAC;QAED,MAAM,IAAI,KAAK,CAAC,uCAAuC,CAAC,CAAC;IAC3D,CAAC;CACF"}
|
|
@@ -26,6 +26,7 @@ interface MultiQueryDeps {
|
|
|
26
26
|
strategy: 'multi-query';
|
|
27
27
|
baseRetriever: Retriever;
|
|
28
28
|
expandQuery: (query: string) => Promise<string[]>;
|
|
29
|
+
topK?: number;
|
|
29
30
|
}
|
|
30
31
|
export type CreateRetrieverConfig = SimilarityDeps | MMRDeps | HybridDeps | MultiQueryDeps;
|
|
31
32
|
export declare function createRetriever(config: CreateRetrieverConfig): Retriever;
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"create-retriever.d.ts","sourceRoot":"","sources":["../../src/retrievers/create-retriever.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EACV,SAAS,EACT,gBAAgB,EAChB,gBAAgB,EAChB,mBAAmB,EACpB,MAAM,qBAAqB,CAAC;AAC7B,OAAO,KAAK,EAAE,YAAY,EAAE,MAAM,sBAAsB,CAAC;AAMzD,UAAU,iBAAiB;IACzB,gBAAgB,EAAE,gBAAgB,CAAC;IACnC,gBAAgB,EAAE,gBAAgB,CAAC;CACpC;AAED,UAAU,cAAe,SAAQ,iBAAiB;IAChD,QAAQ,EAAE,YAAY,CAAC;IACvB,IAAI,CAAC,EAAE,MAAM,CAAC;IACd,SAAS,CAAC,EAAE,MAAM,CAAC;CACpB;AAED,UAAU,OAAQ,SAAQ,iBAAiB;IACzC,QAAQ,EAAE,KAAK,CAAC;IAChB,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,IAAI,CAAC,EAAE,MAAM,CAAC;IACd,SAAS,CAAC,EAAE,MAAM,CAAC;CACpB;AAED,UAAU,UAAU;IAClB,QAAQ,EAAE,QAAQ,CAAC;IACnB,YAAY,EAAE,YAAY,CAAC;IAC3B,OAAO,CAAC,EAAE,mBAAmB,CAAC;IAC9B,IAAI,CAAC,EAAE,MAAM,CAAC;IACd,SAAS,CAAC,EAAE,MAAM,CAAC;CACpB;AAED,UAAU,cAAc;IACtB,QAAQ,EAAE,aAAa,CAAC;IACxB,aAAa,EAAE,SAAS,CAAC;IACzB,WAAW,EAAE,CAAC,KAAK,EAAE,MAAM,KAAK,OAAO,CAAC,MAAM,EAAE,CAAC,CAAC;
|
|
1
|
+
{"version":3,"file":"create-retriever.d.ts","sourceRoot":"","sources":["../../src/retrievers/create-retriever.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EACV,SAAS,EACT,gBAAgB,EAChB,gBAAgB,EAChB,mBAAmB,EACpB,MAAM,qBAAqB,CAAC;AAC7B,OAAO,KAAK,EAAE,YAAY,EAAE,MAAM,sBAAsB,CAAC;AAMzD,UAAU,iBAAiB;IACzB,gBAAgB,EAAE,gBAAgB,CAAC;IACnC,gBAAgB,EAAE,gBAAgB,CAAC;CACpC;AAED,UAAU,cAAe,SAAQ,iBAAiB;IAChD,QAAQ,EAAE,YAAY,CAAC;IACvB,IAAI,CAAC,EAAE,MAAM,CAAC;IACd,SAAS,CAAC,EAAE,MAAM,CAAC;CACpB;AAED,UAAU,OAAQ,SAAQ,iBAAiB;IACzC,QAAQ,EAAE,KAAK,CAAC;IAChB,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,IAAI,CAAC,EAAE,MAAM,CAAC;IACd,SAAS,CAAC,EAAE,MAAM,CAAC;CACpB;AAED,UAAU,UAAU;IAClB,QAAQ,EAAE,QAAQ,CAAC;IACnB,YAAY,EAAE,YAAY,CAAC;IAC3B,OAAO,CAAC,EAAE,mBAAmB,CAAC;IAC9B,IAAI,CAAC,EAAE,MAAM,CAAC;IACd,SAAS,CAAC,EAAE,MAAM,CAAC;CACpB;AAED,UAAU,cAAc;IACtB,QAAQ,EAAE,aAAa,CAAC;IACxB,aAAa,EAAE,SAAS,CAAC;IACzB,WAAW,EAAE,CAAC,KAAK,EAAE,MAAM,KAAK,OAAO,CAAC,MAAM,EAAE,CAAC,CAAC;IAClD,IAAI,CAAC,EAAE,MAAM,CAAC;CACf;AAED,MAAM,MAAM,qBAAqB,GAAG,cAAc,GAAG,OAAO,GAAG,UAAU,GAAG,cAAc,CAAC;AAE3F,wBAAgB,eAAe,CAAC,MAAM,EAAE,qBAAqB,GAAG,SAAS,CAuCxE"}
|
|
@@ -30,7 +30,12 @@ export function createRetriever(config) {
|
|
|
30
30
|
return new MultiQueryRetriever({
|
|
31
31
|
baseRetriever: config.baseRetriever,
|
|
32
32
|
expandQuery: config.expandQuery,
|
|
33
|
+
defaultTopK: config.topK,
|
|
33
34
|
});
|
|
35
|
+
default: {
|
|
36
|
+
const _exhaustive = config;
|
|
37
|
+
throw new Error(`Unknown retriever strategy: ${JSON.stringify(_exhaustive)}`);
|
|
38
|
+
}
|
|
34
39
|
}
|
|
35
40
|
}
|
|
36
41
|
//# sourceMappingURL=create-retriever.js.map
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"create-retriever.js","sourceRoot":"","sources":["../../src/retrievers/create-retriever.ts"],"names":[],"mappings":"AAOA,OAAO,EAAE,mBAAmB,EAAE,MAAM,wBAAwB,CAAC;AAC7D,OAAO,EAAE,YAAY,EAAE,MAAM,iBAAiB,CAAC;AAC/C,OAAO,EAAE,eAAe,EAAE,MAAM,oBAAoB,CAAC;AACrD,OAAO,EAAE,mBAAmB,EAAE,MAAM,yBAAyB,CAAC;
|
|
1
|
+
{"version":3,"file":"create-retriever.js","sourceRoot":"","sources":["../../src/retrievers/create-retriever.ts"],"names":[],"mappings":"AAOA,OAAO,EAAE,mBAAmB,EAAE,MAAM,wBAAwB,CAAC;AAC7D,OAAO,EAAE,YAAY,EAAE,MAAM,iBAAiB,CAAC;AAC/C,OAAO,EAAE,eAAe,EAAE,MAAM,oBAAoB,CAAC;AACrD,OAAO,EAAE,mBAAmB,EAAE,MAAM,yBAAyB,CAAC;AAqC9D,MAAM,UAAU,eAAe,CAAC,MAA6B;IAC3D,QAAQ,MAAM,CAAC,QAAQ,EAAE,CAAC;QACxB,KAAK,YAAY;YACf,OAAO,IAAI,mBAAmB,CAAC;gBAC7B,gBAAgB,EAAE,MAAM,CAAC,gBAAgB;gBACzC,gBAAgB,EAAE,MAAM,CAAC,gBAAgB;gBACzC,WAAW,EAAE,MAAM,CAAC,IAAI;gBACxB,gBAAgB,EAAE,MAAM,CAAC,SAAS;aACnC,CAAC,CAAC;QAEL,KAAK,KAAK;YACR,OAAO,IAAI,YAAY,CAAC;gBACtB,gBAAgB,EAAE,MAAM,CAAC,gBAAgB;gBACzC,gBAAgB,EAAE,MAAM,CAAC,gBAAgB;gBACzC,aAAa,EAAE,MAAM,CAAC,MAAM;gBAC5B,WAAW,EAAE,MAAM,CAAC,IAAI;gBACxB,gBAAgB,EAAE,MAAM,CAAC,SAAS;aACnC,CAAC,CAAC;QAEL,KAAK,QAAQ;YACX,OAAO,IAAI,eAAe,CAAC;gBACzB,YAAY,EAAE,MAAM,CAAC,YAAY;gBACjC,cAAc,EAAE,MAAM,CAAC,OAAO;gBAC9B,WAAW,EAAE,MAAM,CAAC,IAAI;gBACxB,gBAAgB,EAAE,MAAM,CAAC,SAAS;aACnC,CAAC,CAAC;QAEL,KAAK,aAAa;YAChB,OAAO,IAAI,mBAAmB,CAAC;gBAC7B,aAAa,EAAE,MAAM,CAAC,aAAa;gBACnC,WAAW,EAAE,MAAM,CAAC,WAAW;gBAC/B,WAAW,EAAE,MAAM,CAAC,IAAI;aACzB,CAAC,CAAC;QAEL,OAAO,CAAC,CAAC,CAAC;YACR,MAAM,WAAW,GAAU,MAAM,CAAC;YAClC,MAAM,IAAI,KAAK,CAAC,+BAA+B,IAAI,CAAC,SAAS,CAAC,WAAW,CAAC,EAAE,CAAC,CAAC;QAChF,CAAC;IACH,CAAC;AACH,CAAC"}
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"mmr-retriever.d.ts","sourceRoot":"","sources":["../../src/retrievers/mmr-retriever.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EACV,gBAAgB,EAChB,gBAAgB,EAEhB,SAAS,EACT,eAAe,EACf,eAAe,EAChB,MAAM,qBAAqB,CAAC;AAE7B,MAAM,WAAW,kBAAkB;IACjC,gBAAgB,EAAE,gBAAgB,CAAC;IACnC,gBAAgB,EAAE,gBAAgB,CAAC;IACnC,aAAa,CAAC,EAAE,MAAM,CAAC;IACvB,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,gBAAgB,CAAC,EAAE,MAAM,CAAC;CAC3B;
|
|
1
|
+
{"version":3,"file":"mmr-retriever.d.ts","sourceRoot":"","sources":["../../src/retrievers/mmr-retriever.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EACV,gBAAgB,EAChB,gBAAgB,EAEhB,SAAS,EACT,eAAe,EACf,eAAe,EAChB,MAAM,qBAAqB,CAAC;AAE7B,MAAM,WAAW,kBAAkB;IACjC,gBAAgB,EAAE,gBAAgB,CAAC;IACnC,gBAAgB,EAAE,gBAAgB,CAAC;IACnC,aAAa,CAAC,EAAE,MAAM,CAAC;IACvB,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,gBAAgB,CAAC,EAAE,MAAM,CAAC;CAC3B;AAyBD,qBAAa,YAAa,YAAW,SAAS;IAC5C,OAAO,CAAC,QAAQ,CAAC,OAAO,CAAmB;IAC3C,OAAO,CAAC,QAAQ,CAAC,SAAS,CAAmB;IAC7C,OAAO,CAAC,QAAQ,CAAC,aAAa,CAAS;IACvC,OAAO,CAAC,QAAQ,CAAC,WAAW,CAAS;IACrC,OAAO,CAAC,QAAQ,CAAC,gBAAgB,CAAS;gBAE9B,MAAM,EAAE,kBAAkB;IAQhC,QAAQ,CAAC,KAAK,EAAE,MAAM,EAAE,OAAO,CAAC,EAAE,OAAO,CAAC,eAAe,CAAC,GAAG,OAAO,CAAC,eAAe,EAAE,CAAC;IAwB7F,OAAO,CAAC,SAAS;IAuCjB,OAAO,CAAC,iBAAiB;CAa1B"}
|
|
@@ -3,11 +3,13 @@ const DEFAULT_THRESHOLD = 0.0;
|
|
|
3
3
|
const DEFAULT_LAMBDA = 0.7;
|
|
4
4
|
const CANDIDATE_MULTIPLIER = 3;
|
|
5
5
|
function cosineSimilarity(a, b) {
|
|
6
|
-
|
|
6
|
+
if (a.length !== b.length) {
|
|
7
|
+
throw new Error(`Vector dimension mismatch: ${a.length} vs ${b.length}`);
|
|
8
|
+
}
|
|
7
9
|
let dot = 0;
|
|
8
10
|
let normA = 0;
|
|
9
11
|
let normB = 0;
|
|
10
|
-
for (let i = 0; i <
|
|
12
|
+
for (let i = 0; i < a.length; i++) {
|
|
11
13
|
dot += a[i] * b[i];
|
|
12
14
|
normA += a[i] * a[i];
|
|
13
15
|
normB += b[i] * b[i];
|
|
@@ -44,11 +46,11 @@ export class MMRRetriever {
|
|
|
44
46
|
const candidates = result.data;
|
|
45
47
|
if (candidates.length === 0)
|
|
46
48
|
return [];
|
|
47
|
-
const
|
|
49
|
+
const lambda = options?.mmrLambda ?? this.defaultLambda;
|
|
50
|
+
const selected = this.selectMMR(candidates, topK, lambda);
|
|
48
51
|
return selected.map((entry) => this.toRetrievalResult(entry));
|
|
49
52
|
}
|
|
50
|
-
selectMMR(candidates, topK) {
|
|
51
|
-
const lambda = this.defaultLambda;
|
|
53
|
+
selectMMR(candidates, topK, lambda) {
|
|
52
54
|
const remaining = new Set(candidates.map((_, i) => i));
|
|
53
55
|
const selected = [];
|
|
54
56
|
while (selected.length < topK && remaining.size > 0) {
|
|
@@ -56,12 +58,14 @@ export class MMRRetriever {
|
|
|
56
58
|
let bestScore = -Infinity;
|
|
57
59
|
for (const idx of remaining) {
|
|
58
60
|
const relevance = candidates[idx].score;
|
|
59
|
-
let maxSelectedSim =
|
|
61
|
+
let maxSelectedSim = -Infinity;
|
|
60
62
|
for (const sel of selected) {
|
|
61
63
|
const sim = cosineSimilarity(candidates[idx].vector, sel.vector);
|
|
62
64
|
if (sim > maxSelectedSim)
|
|
63
65
|
maxSelectedSim = sim;
|
|
64
66
|
}
|
|
67
|
+
if (maxSelectedSim === -Infinity)
|
|
68
|
+
maxSelectedSim = 0;
|
|
65
69
|
const mmrScore = lambda * relevance - (1 - lambda) * maxSelectedSim;
|
|
66
70
|
if (mmrScore > bestScore) {
|
|
67
71
|
bestScore = mmrScore;
|
|
@@ -84,7 +88,7 @@ export class MMRRetriever {
|
|
|
84
88
|
content: entry.content,
|
|
85
89
|
score: entry.mmrScore,
|
|
86
90
|
source: entry.sourceType,
|
|
87
|
-
metadata: entry.metadata,
|
|
91
|
+
metadata: { ...entry.metadata, originalScore: entry.score },
|
|
88
92
|
};
|
|
89
93
|
}
|
|
90
94
|
}
|