@phoenixaihub/graphrag 0.1.0
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/.github/ISSUE_TEMPLATE/bug_report.md +17 -0
- package/.github/ISSUE_TEMPLATE/feature_request.md +12 -0
- package/.github/pull_request_template.md +5 -0
- package/.github/workflows/ci.yml +22 -0
- package/LICENSE +21 -0
- package/README.md +103 -0
- package/dist/cli.d.ts +3 -0
- package/dist/cli.d.ts.map +1 -0
- package/dist/cli.js +47 -0
- package/dist/cli.js.map +1 -0
- package/dist/commands/index-repo.d.ts +8 -0
- package/dist/commands/index-repo.d.ts.map +1 -0
- package/dist/commands/index-repo.js +79 -0
- package/dist/commands/index-repo.js.map +1 -0
- package/dist/commands/query.d.ts +7 -0
- package/dist/commands/query.d.ts.map +1 -0
- package/dist/commands/query.js +32 -0
- package/dist/commands/query.js.map +1 -0
- package/dist/commands/serve.d.ts +7 -0
- package/dist/commands/serve.d.ts.map +1 -0
- package/dist/commands/serve.js +65 -0
- package/dist/commands/serve.js.map +1 -0
- package/dist/db.d.ts +46 -0
- package/dist/db.d.ts.map +1 -0
- package/dist/db.js +117 -0
- package/dist/db.js.map +1 -0
- package/dist/embeddings.d.ts +4 -0
- package/dist/embeddings.d.ts.map +1 -0
- package/dist/embeddings.js +35 -0
- package/dist/embeddings.js.map +1 -0
- package/dist/index.d.ts +6 -0
- package/dist/index.d.ts.map +1 -0
- package/dist/index.js +6 -0
- package/dist/index.js.map +1 -0
- package/dist/parser.d.ts +12 -0
- package/dist/parser.d.ts.map +1 -0
- package/dist/parser.js +152 -0
- package/dist/parser.js.map +1 -0
- package/package.json +54 -0
- package/src/cli.ts +52 -0
- package/src/commands/index-repo.ts +96 -0
- package/src/commands/query.ts +43 -0
- package/src/commands/serve.ts +90 -0
- package/src/db.ts +168 -0
- package/src/embeddings.ts +39 -0
- package/src/index.ts +5 -0
- package/src/parser.ts +162 -0
- package/tsconfig.json +19 -0
package/dist/db.js.map
ADDED
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"db.js","sourceRoot":"","sources":["../src/db.ts"],"names":[],"mappings":"AAAA,OAAO,QAAQ,MAAM,gBAAgB,CAAC;AACtC,OAAO,IAAI,MAAM,MAAM,CAAC;AACxB,OAAO,EAAE,MAAM,IAAI,CAAC;AA2BpB,MAAM,MAAM,GAAG;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;CAgCd,CAAC;AAEF,MAAM,OAAO,OAAO;IACV,EAAE,CAAoB;IAE9B,YAAY,MAAc;QACxB,MAAM,GAAG,GAAG,IAAI,CAAC,OAAO,CAAC,MAAM,CAAC,CAAC;QACjC,IAAI,CAAC,EAAE,CAAC,UAAU,CAAC,GAAG,CAAC;YAAE,EAAE,CAAC,SAAS,CAAC,GAAG,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC,CAAC;QAChE,IAAI,CAAC,EAAE,GAAG,IAAI,QAAQ,CAAC,MAAM,CAAC,CAAC;QAC/B,IAAI,CAAC,EAAE,CAAC,MAAM,CAAC,oBAAoB,CAAC,CAAC;QACrC,IAAI,CAAC,EAAE,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;IACvB,CAAC;IAED,YAAY,CAAC,CAAS;QACpB,MAAM,IAAI,GAAG,IAAI,CAAC,EAAE,CAAC,OAAO,CAC1B,iHAAiH,CAClH,CAAC;QACF,MAAM,MAAM,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,QAAQ,EAAE,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC,SAAS,EAAE,CAAC,CAAC,OAAO,EAAE,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC,QAAQ,CAAC,CAAC;QAChG,OAAO,MAAM,CAAC,MAAM,CAAC,eAAe,CAAC,CAAC;IACxC,CAAC;IAED,UAAU,CAAC,CAAO;QAChB,MAAM,IAAI,GAAG,IAAI,CAAC,EAAE,CAAC,OAAO,CAC1B,qEAAqE,CACtE,CAAC;QACF,MAAM,MAAM,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,QAAQ,EAAE,CAAC,CAAC,QAAQ,EAAE,CAAC,CAAC,QAAQ,CAAC,CAAC;QAC5D,OAAO,MAAM,CAAC,MAAM,CAAC,eAAe,CAAC,CAAC;IACxC,CAAC;IAED,WAAW,CAAC,CAAQ;QAClB,MAAM,IAAI,GAAG,IAAI,CAAC,EAAE,CAAC,OAAO,CAC1B,qEAAqE,CACtE,CAAC;QACF,MAAM,aAAa,GAAG,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,MAAM,CAAC,IAAI,CAAC,CAAC,CAAC,SAAS,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC;QAC3E,MAAM,MAAM,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,QAAQ,EAAE,CAAC,CAAC,OAAO,EAAE,aAAa,CAAC,CAAC;QAC9D,OAAO,MAAM,CAAC,MAAM,CAAC,eAAe,CAAC,CAAC;IACxC,CAAC;IAED,SAAS,CAAC,EAAU;QAClB,OAAO,IAAI,CAAC,EAAE,CAAC,OAAO,CAAC,qCAAqC,CAAC,CAAC,GAAG,CAAC,EAAE,CAAuB,CAAC;IAC9F,CAAC;IAED,kBAAkB,CAAC,IAAY;QAC7B,OAAO,IAAI,CAAC,EAAE,CAAC,OAAO,CACpB,0CAA0C,CAC3C,CAAC,GAAG,CAAC,IAAI,IAAI,GAAG,CAAa,CAAC;IACjC,CAAC;IAED,cAAc,CAAC,KAAa,EAAE,KAAK,GAAG,CAAC;QACrC,yCAAyC;QACzC,MAAM,KAAK,GAAG,KAAK,CAAC,WAAW,EAAE,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC;QAC/D,IAAI,KAAK,CAAC,MAAM,KAAK,CAAC;YAAE,OAAO,EAAE,CAAC;QAClC,MAAM,UAAU,GAAG,KAAK,CAAC,GAAG,CAAC,GAAG,EAAE,CAAC,4CAA4C,CAAC,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC;QAC/F,MAAM,MAAM,GAAG,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,IAAI,CAAC,GAAG,EAAE,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC;QACxD,OAAO,IAAI,CAAC,EAAE,CAAC,OAAO,CACpB,gCAAgC,UAAU,8BAA8B,CACzE,CAAC,GAAG,CAAC,GAAG,MAAM,EAAE,KAAK,CAAa,CAAC;IACtC,CAAC;IAED,UAAU,CAAC,QAAgB,EAAE,YAA8C,MAAM;QAC/E,MAAM,OAAO,GAAgD,EAAE,CAAC;QAChE,IAAI,SAAS,KAAK,UAAU,IAAI,SAAS,KAAK,MAAM,EAAE,CAAC;YACrD,MAAM,IAAI,GAAG,IAAI,CAAC,EAAE,CAAC,OAAO,CAC1B,qGAAqG,CACtG,CAAC,GAAG,CAAC,QAAQ,CAAyC,CAAC;YACxD,OAAO,CAAC,IAAI,CAAC,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,EAAE,MAAM,EAAE,CAAC,EAAE,QAAQ,EAAE,CAAC,CAAC,QAAQ,EAAE,CAAC,CAAC,CAAC,CAAC;QACxE,CAAC;QACD,IAAI,SAAS,KAAK,UAAU,IAAI,SAAS,KAAK,MAAM,EAAE,CAAC;YACrD,MAAM,IAAI,GAAG,IAAI,CAAC,EAAE,CAAC,OAAO,CAC1B,qGAAqG,CACtG,CAAC,GAAG,CAAC,QAAQ,CAAyC,CAAC;YACxD,OAAO,CAAC,IAAI,CAAC,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,EAAE,MAAM,EAAE,CAAC,EAAE,QAAQ,EAAE,CAAC,CAAC,QAAQ,EAAE,CAAC,CAAC,CAAC,CAAC;QACxE,CAAC;QACD,OAAO,OAAO,CAAC;IACjB,CAAC;IAED,SAAS,CAAC,IAAY;QACpB,OAAO,IAAI,CAAC,EAAE,CAAC,OAAO,CACpB;;;iEAG2D,CAC5D,CAAC,GAAG,CAAC,IAAI,CAAa,CAAC;IAC1B,CAAC;IAED,SAAS,CAAC,QAAgB;QACxB,MAAM,QAAQ,GAAG,IAAI,CAAC,EAAE,CAAC,OAAO,CAAC,6CAA6C,CAAC,CAAC,GAAG,CAAC,QAAQ,CAA0B,CAAC;QACvH,MAAM,GAAG,GAAG,QAAQ,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC;QACpC,IAAI,GAAG,CAAC,MAAM,KAAK,CAAC;YAAE,OAAO;QAC7B,MAAM,YAAY,GAAG,GAAG,CAAC,GAAG,CAAC,GAAG,EAAE,CAAC,GAAG,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;QAClD,IAAI,CAAC,EAAE,CAAC,OAAO,CAAC,0CAA0C,YAAY,GAAG,CAAC,CAAC,GAAG,CAAC,GAAG,GAAG,CAAC,CAAC;QACvF,IAAI,CAAC,EAAE,CAAC,OAAO,CAAC,yCAAyC,YAAY,sBAAsB,YAAY,GAAG,CAAC,CAAC,GAAG,CAAC,GAAG,GAAG,EAAE,GAAG,GAAG,CAAC,CAAC;QAChI,IAAI,CAAC,EAAE,CAAC,OAAO,CAAC,0CAA0C,CAAC,CAAC,GAAG,CAAC,QAAQ,CAAC,CAAC;IAC5E,CAAC;IAED,QAAQ;QACN,MAAM,QAAQ,GAAI,IAAI,CAAC,EAAE,CAAC,OAAO,CAAC,oCAAoC,CAAC,CAAC,GAAG,EAAoB,CAAC,CAAC,CAAC;QAClG,MAAM,KAAK,GAAI,IAAI,CAAC,EAAE,CAAC,OAAO,CAAC,iCAAiC,CAAC,CAAC,GAAG,EAAoB,CAAC,CAAC,CAAC;QAC5F,MAAM,MAAM,GAAI,IAAI,CAAC,EAAE,CAAC,OAAO,CAAC,kCAAkC,CAAC,CAAC,GAAG,EAAoB,CAAC,CAAC,CAAC;QAC9F,MAAM,KAAK,GAAI,IAAI,CAAC,EAAE,CAAC,OAAO,CAAC,qDAAqD,CAAC,CAAC,GAAG,EAAoB,CAAC,CAAC,CAAC;QAChH,OAAO,EAAE,QAAQ,EAAE,KAAK,EAAE,MAAM,EAAE,KAAK,EAAE,CAAC;IAC5C,CAAC;IAED,KAAK;QACH,IAAI,CAAC,EAAE,CAAC,KAAK,EAAE,CAAC;IAClB,CAAC;CACF"}
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"embeddings.d.ts","sourceRoot":"","sources":["../src/embeddings.ts"],"names":[],"mappings":"AAWA,wBAAsB,YAAY,CAAC,IAAI,EAAE,MAAM,GAAG,OAAO,CAAC,MAAM,EAAE,CAAC,CAOlE;AAED,wBAAsB,aAAa,CAAC,KAAK,EAAE,MAAM,EAAE,GAAG,OAAO,CAAC,MAAM,EAAE,EAAE,CAAC,CAQxE;AAED,wBAAgB,gBAAgB,CAAC,CAAC,EAAE,MAAM,EAAE,EAAE,CAAC,EAAE,MAAM,EAAE,GAAG,MAAM,CAQjE"}
|
|
@@ -0,0 +1,35 @@
|
|
|
1
|
+
import OpenAI from 'openai';
|
|
2
|
+
let client = null;
|
|
3
|
+
function getClient() {
|
|
4
|
+
if (!client) {
|
|
5
|
+
client = new OpenAI();
|
|
6
|
+
}
|
|
7
|
+
return client;
|
|
8
|
+
}
|
|
9
|
+
export async function getEmbedding(text) {
|
|
10
|
+
const openai = getClient();
|
|
11
|
+
const resp = await openai.embeddings.create({
|
|
12
|
+
model: 'text-embedding-3-small',
|
|
13
|
+
input: text.substring(0, 8000),
|
|
14
|
+
});
|
|
15
|
+
return resp.data[0].embedding;
|
|
16
|
+
}
|
|
17
|
+
export async function getEmbeddings(texts) {
|
|
18
|
+
const openai = getClient();
|
|
19
|
+
const truncated = texts.map(t => t.substring(0, 8000));
|
|
20
|
+
const resp = await openai.embeddings.create({
|
|
21
|
+
model: 'text-embedding-3-small',
|
|
22
|
+
input: truncated,
|
|
23
|
+
});
|
|
24
|
+
return resp.data.map(d => d.embedding);
|
|
25
|
+
}
|
|
26
|
+
export function cosineSimilarity(a, b) {
|
|
27
|
+
let dot = 0, normA = 0, normB = 0;
|
|
28
|
+
for (let i = 0; i < a.length; i++) {
|
|
29
|
+
dot += a[i] * b[i];
|
|
30
|
+
normA += a[i] * a[i];
|
|
31
|
+
normB += b[i] * b[i];
|
|
32
|
+
}
|
|
33
|
+
return dot / (Math.sqrt(normA) * Math.sqrt(normB));
|
|
34
|
+
}
|
|
35
|
+
//# sourceMappingURL=embeddings.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"embeddings.js","sourceRoot":"","sources":["../src/embeddings.ts"],"names":[],"mappings":"AAAA,OAAO,MAAM,MAAM,QAAQ,CAAC;AAE5B,IAAI,MAAM,GAAkB,IAAI,CAAC;AAEjC,SAAS,SAAS;IAChB,IAAI,CAAC,MAAM,EAAE,CAAC;QACZ,MAAM,GAAG,IAAI,MAAM,EAAE,CAAC;IACxB,CAAC;IACD,OAAO,MAAM,CAAC;AAChB,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,YAAY,CAAC,IAAY;IAC7C,MAAM,MAAM,GAAG,SAAS,EAAE,CAAC;IAC3B,MAAM,IAAI,GAAG,MAAM,MAAM,CAAC,UAAU,CAAC,MAAM,CAAC;QAC1C,KAAK,EAAE,wBAAwB;QAC/B,KAAK,EAAE,IAAI,CAAC,SAAS,CAAC,CAAC,EAAE,IAAI,CAAC;KAC/B,CAAC,CAAC;IACH,OAAO,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC;AAChC,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,aAAa,CAAC,KAAe;IACjD,MAAM,MAAM,GAAG,SAAS,EAAE,CAAC;IAC3B,MAAM,SAAS,GAAG,KAAK,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,CAAC;IACvD,MAAM,IAAI,GAAG,MAAM,MAAM,CAAC,UAAU,CAAC,MAAM,CAAC;QAC1C,KAAK,EAAE,wBAAwB;QAC/B,KAAK,EAAE,SAAS;KACjB,CAAC,CAAC;IACH,OAAO,IAAI,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC;AACzC,CAAC;AAED,MAAM,UAAU,gBAAgB,CAAC,CAAW,EAAE,CAAW;IACvD,IAAI,GAAG,GAAG,CAAC,EAAE,KAAK,GAAG,CAAC,EAAE,KAAK,GAAG,CAAC,CAAC;IAClC,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;QAClC,GAAG,IAAI,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC;QACnB,KAAK,IAAI,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC;QACrB,KAAK,IAAI,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC;IACvB,CAAC;IACD,OAAO,GAAG,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,GAAG,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC;AACrD,CAAC"}
|
package/dist/index.d.ts
ADDED
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,SAAS,EAAE,MAAM,0BAA0B,CAAC;AACrD,OAAO,EAAE,SAAS,EAAE,MAAM,qBAAqB,CAAC;AAChD,OAAO,EAAE,KAAK,EAAE,MAAM,qBAAqB,CAAC;AAC5C,OAAO,EAAE,OAAO,EAAE,MAAM,SAAS,CAAC;AAClC,OAAO,EAAE,SAAS,EAAE,MAAM,aAAa,CAAC"}
|
package/dist/index.js
ADDED
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"index.js","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,SAAS,EAAE,MAAM,0BAA0B,CAAC;AACrD,OAAO,EAAE,SAAS,EAAE,MAAM,qBAAqB,CAAC;AAChD,OAAO,EAAE,KAAK,EAAE,MAAM,qBAAqB,CAAC;AAC5C,OAAO,EAAE,OAAO,EAAE,MAAM,SAAS,CAAC;AAClC,OAAO,EAAE,SAAS,EAAE,MAAM,aAAa,CAAC"}
|
package/dist/parser.d.ts
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
1
|
+
import type { Entity, Edge } from './db.js';
|
|
2
|
+
interface ParseResult {
|
|
3
|
+
entities: Omit<Entity, 'id'>[];
|
|
4
|
+
edges: Array<{
|
|
5
|
+
sourceName: string;
|
|
6
|
+
targetName: string;
|
|
7
|
+
relation: Edge['relation'];
|
|
8
|
+
}>;
|
|
9
|
+
}
|
|
10
|
+
export declare function parseFile(filePath: string, source?: string): ParseResult;
|
|
11
|
+
export {};
|
|
12
|
+
//# sourceMappingURL=parser.d.ts.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"parser.d.ts","sourceRoot":"","sources":["../src/parser.ts"],"names":[],"mappings":"AAEA,OAAO,KAAK,EAAE,MAAM,EAAE,IAAI,EAAE,MAAM,SAAS,CAAC;AAI5C,UAAU,WAAW;IACnB,QAAQ,EAAE,IAAI,CAAC,MAAM,EAAE,IAAI,CAAC,EAAE,CAAC;IAC/B,KAAK,EAAE,KAAK,CAAC;QAAE,UAAU,EAAE,MAAM,CAAC;QAAC,UAAU,EAAE,MAAM,CAAC;QAAC,QAAQ,EAAE,IAAI,CAAC,UAAU,CAAC,CAAA;KAAE,CAAC,CAAC;CACtF;AAyED,wBAAgB,SAAS,CAAC,QAAQ,EAAE,MAAM,EAAE,MAAM,CAAC,EAAE,MAAM,GAAG,WAAW,CA+ExE"}
|
package/dist/parser.js
ADDED
|
@@ -0,0 +1,152 @@
|
|
|
1
|
+
import fs from 'fs';
|
|
2
|
+
import path from 'path';
|
|
3
|
+
const FUNCTION_PATTERNS = {
|
|
4
|
+
typescript: [
|
|
5
|
+
/^(?:export\s+)?(?:async\s+)?function\s+(\w+)/gm,
|
|
6
|
+
/^(?:export\s+)?const\s+(\w+)\s*=\s*(?:async\s+)?\(/gm,
|
|
7
|
+
/^(?:export\s+)?const\s+(\w+)\s*=\s*(?:async\s+)?(?:\([^)]*\)|[^=])\s*=>/gm,
|
|
8
|
+
],
|
|
9
|
+
python: [
|
|
10
|
+
/^def\s+(\w+)\s*\(/gm,
|
|
11
|
+
/^async\s+def\s+(\w+)\s*\(/gm,
|
|
12
|
+
],
|
|
13
|
+
};
|
|
14
|
+
const CLASS_PATTERNS = {
|
|
15
|
+
typescript: [/^(?:export\s+)?(?:abstract\s+)?class\s+(\w+)/gm],
|
|
16
|
+
python: [/^class\s+(\w+)/gm],
|
|
17
|
+
};
|
|
18
|
+
const IMPORT_PATTERNS = {
|
|
19
|
+
typescript: [
|
|
20
|
+
/import\s+(?:{([^}]+)}|(\w+))\s+from\s+['"]([^'"]+)['"]/gm,
|
|
21
|
+
/import\s+\*\s+as\s+(\w+)\s+from\s+['"]([^'"]+)['"]/gm,
|
|
22
|
+
],
|
|
23
|
+
python: [
|
|
24
|
+
/^from\s+(\S+)\s+import\s+(.+)$/gm,
|
|
25
|
+
/^import\s+(\S+)/gm,
|
|
26
|
+
],
|
|
27
|
+
};
|
|
28
|
+
function detectLanguage(filePath) {
|
|
29
|
+
const ext = path.extname(filePath).toLowerCase();
|
|
30
|
+
if (['.ts', '.tsx', '.js', '.jsx', '.mjs', '.cjs'].includes(ext))
|
|
31
|
+
return 'typescript';
|
|
32
|
+
if (ext === '.py')
|
|
33
|
+
return 'python';
|
|
34
|
+
return 'unknown';
|
|
35
|
+
}
|
|
36
|
+
function getLineNumber(source, index) {
|
|
37
|
+
return source.substring(0, index).split('\n').length;
|
|
38
|
+
}
|
|
39
|
+
function extractBlock(lines, startLine, language) {
|
|
40
|
+
// Simple brace/indent block extraction
|
|
41
|
+
const start = startLine - 1;
|
|
42
|
+
if (language === 'python') {
|
|
43
|
+
let end = start + 1;
|
|
44
|
+
const baseIndent = lines[start]?.search(/\S/) ?? 0;
|
|
45
|
+
while (end < lines.length) {
|
|
46
|
+
const line = lines[end];
|
|
47
|
+
if (line.trim() === '') {
|
|
48
|
+
end++;
|
|
49
|
+
continue;
|
|
50
|
+
}
|
|
51
|
+
const indent = line.search(/\S/);
|
|
52
|
+
if (indent <= baseIndent)
|
|
53
|
+
break;
|
|
54
|
+
end++;
|
|
55
|
+
}
|
|
56
|
+
return { endLine: end, code: lines.slice(start, end).join('\n') };
|
|
57
|
+
}
|
|
58
|
+
// Brace-based
|
|
59
|
+
let braces = 0;
|
|
60
|
+
let found = false;
|
|
61
|
+
let end = start;
|
|
62
|
+
for (let i = start; i < lines.length; i++) {
|
|
63
|
+
for (const ch of lines[i]) {
|
|
64
|
+
if (ch === '{') {
|
|
65
|
+
braces++;
|
|
66
|
+
found = true;
|
|
67
|
+
}
|
|
68
|
+
if (ch === '}')
|
|
69
|
+
braces--;
|
|
70
|
+
}
|
|
71
|
+
end = i + 1;
|
|
72
|
+
if (found && braces <= 0)
|
|
73
|
+
break;
|
|
74
|
+
}
|
|
75
|
+
if (!found)
|
|
76
|
+
end = Math.min(start + 1, lines.length);
|
|
77
|
+
return { endLine: end, code: lines.slice(start, end).join('\n') };
|
|
78
|
+
}
|
|
79
|
+
export function parseFile(filePath, source) {
|
|
80
|
+
const content = source ?? fs.readFileSync(filePath, 'utf-8');
|
|
81
|
+
const language = detectLanguage(filePath);
|
|
82
|
+
if (language === 'unknown')
|
|
83
|
+
return { entities: [], edges: [] };
|
|
84
|
+
const lines = content.split('\n');
|
|
85
|
+
const entities = [];
|
|
86
|
+
const edgesList = [];
|
|
87
|
+
// File entity
|
|
88
|
+
entities.push({
|
|
89
|
+
filePath, name: path.basename(filePath), kind: 'file',
|
|
90
|
+
startLine: 1, endLine: lines.length, code: content.substring(0, 500),
|
|
91
|
+
language,
|
|
92
|
+
});
|
|
93
|
+
const langKey = language;
|
|
94
|
+
// Functions
|
|
95
|
+
for (const pattern of (FUNCTION_PATTERNS[langKey] ?? [])) {
|
|
96
|
+
const regex = new RegExp(pattern.source, pattern.flags);
|
|
97
|
+
let match;
|
|
98
|
+
while ((match = regex.exec(content)) !== null) {
|
|
99
|
+
const name = match[1];
|
|
100
|
+
const startLine = getLineNumber(content, match.index);
|
|
101
|
+
const block = extractBlock(lines, startLine, language);
|
|
102
|
+
entities.push({
|
|
103
|
+
filePath, name, kind: 'function',
|
|
104
|
+
startLine, endLine: block.endLine, code: block.code,
|
|
105
|
+
language,
|
|
106
|
+
});
|
|
107
|
+
edgesList.push({ sourceName: path.basename(filePath), targetName: name, relation: 'contains' });
|
|
108
|
+
}
|
|
109
|
+
}
|
|
110
|
+
// Classes
|
|
111
|
+
for (const pattern of (CLASS_PATTERNS[langKey] ?? [])) {
|
|
112
|
+
const regex = new RegExp(pattern.source, pattern.flags);
|
|
113
|
+
let match;
|
|
114
|
+
while ((match = regex.exec(content)) !== null) {
|
|
115
|
+
const name = match[1];
|
|
116
|
+
const startLine = getLineNumber(content, match.index);
|
|
117
|
+
const block = extractBlock(lines, startLine, language);
|
|
118
|
+
entities.push({
|
|
119
|
+
filePath, name, kind: 'class',
|
|
120
|
+
startLine, endLine: block.endLine, code: block.code,
|
|
121
|
+
language,
|
|
122
|
+
});
|
|
123
|
+
edgesList.push({ sourceName: path.basename(filePath), targetName: name, relation: 'contains' });
|
|
124
|
+
}
|
|
125
|
+
}
|
|
126
|
+
// Imports
|
|
127
|
+
for (const pattern of (IMPORT_PATTERNS[langKey] ?? [])) {
|
|
128
|
+
const regex = new RegExp(pattern.source, pattern.flags);
|
|
129
|
+
let match;
|
|
130
|
+
while ((match = regex.exec(content)) !== null) {
|
|
131
|
+
const startLine = getLineNumber(content, match.index);
|
|
132
|
+
const importSource = language === 'typescript' ? (match[3] ?? match[2] ?? match[1]) : match[1];
|
|
133
|
+
const importedNames = language === 'typescript'
|
|
134
|
+
? (match[1] ?? match[2] ?? '').split(',').map(s => s.trim()).filter(Boolean)
|
|
135
|
+
: (match[2] ?? match[1] ?? '').split(',').map(s => s.trim()).filter(Boolean);
|
|
136
|
+
for (const name of importedNames) {
|
|
137
|
+
entities.push({
|
|
138
|
+
filePath, name: name.replace(/\s+as\s+\w+/, ''), kind: 'import',
|
|
139
|
+
startLine, endLine: startLine, code: match[0],
|
|
140
|
+
language,
|
|
141
|
+
});
|
|
142
|
+
edgesList.push({
|
|
143
|
+
sourceName: path.basename(filePath),
|
|
144
|
+
targetName: name.replace(/\s+as\s+\w+/, ''),
|
|
145
|
+
relation: 'imports',
|
|
146
|
+
});
|
|
147
|
+
}
|
|
148
|
+
}
|
|
149
|
+
}
|
|
150
|
+
return { entities, edges: edgesList };
|
|
151
|
+
}
|
|
152
|
+
//# sourceMappingURL=parser.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"parser.js","sourceRoot":"","sources":["../src/parser.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,MAAM,IAAI,CAAC;AACpB,OAAO,IAAI,MAAM,MAAM,CAAC;AAUxB,MAAM,iBAAiB,GAA6B;IAClD,UAAU,EAAE;QACV,gDAAgD;QAChD,sDAAsD;QACtD,2EAA2E;KAC5E;IACD,MAAM,EAAE;QACN,qBAAqB;QACrB,6BAA6B;KAC9B;CACF,CAAC;AAEF,MAAM,cAAc,GAA6B;IAC/C,UAAU,EAAE,CAAC,gDAAgD,CAAC;IAC9D,MAAM,EAAE,CAAC,kBAAkB,CAAC;CAC7B,CAAC;AAEF,MAAM,eAAe,GAA6B;IAChD,UAAU,EAAE;QACV,0DAA0D;QAC1D,sDAAsD;KACvD;IACD,MAAM,EAAE;QACN,kCAAkC;QAClC,mBAAmB;KACpB;CACF,CAAC;AAEF,SAAS,cAAc,CAAC,QAAgB;IACtC,MAAM,GAAG,GAAG,IAAI,CAAC,OAAO,CAAC,QAAQ,CAAC,CAAC,WAAW,EAAE,CAAC;IACjD,IAAI,CAAC,KAAK,EAAE,MAAM,EAAE,KAAK,EAAE,MAAM,EAAE,MAAM,EAAE,MAAM,CAAC,CAAC,QAAQ,CAAC,GAAG,CAAC;QAAE,OAAO,YAAY,CAAC;IACtF,IAAI,GAAG,KAAK,KAAK;QAAE,OAAO,QAAQ,CAAC;IACnC,OAAO,SAAS,CAAC;AACnB,CAAC;AAED,SAAS,aAAa,CAAC,MAAc,EAAE,KAAa;IAClD,OAAO,MAAM,CAAC,SAAS,CAAC,CAAC,EAAE,KAAK,CAAC,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC,MAAM,CAAC;AACvD,CAAC;AAED,SAAS,YAAY,CAAC,KAAe,EAAE,SAAiB,EAAE,QAAgB;IACxE,uCAAuC;IACvC,MAAM,KAAK,GAAG,SAAS,GAAG,CAAC,CAAC;IAC5B,IAAI,QAAQ,KAAK,QAAQ,EAAE,CAAC;QAC1B,IAAI,GAAG,GAAG,KAAK,GAAG,CAAC,CAAC;QACpB,MAAM,UAAU,GAAG,KAAK,CAAC,KAAK,CAAC,EAAE,MAAM,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QACnD,OAAO,GAAG,GAAG,KAAK,CAAC,MAAM,EAAE,CAAC;YAC1B,MAAM,IAAI,GAAG,KAAK,CAAC,GAAG,CAAC,CAAC;YACxB,IAAI,IAAI,CAAC,IAAI,EAAE,KAAK,EAAE,EAAE,CAAC;gBAAC,GAAG,EAAE,CAAC;gBAAC,SAAS;YAAC,CAAC;YAC5C,MAAM,MAAM,GAAG,IAAI,CAAC,MAAM,CAAC,IAAI,CAAC,CAAC;YACjC,IAAI,MAAM,IAAI,UAAU;gBAAE,MAAM;YAChC,GAAG,EAAE,CAAC;QACR,CAAC;QACD,OAAO,EAAE,OAAO,EAAE,GAAG,EAAE,IAAI,EAAE,KAAK,CAAC,KAAK,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC;IACpE,CAAC;IAED,cAAc;IACd,IAAI,MAAM,GAAG,CAAC,CAAC;IACf,IAAI,KAAK,GAAG,KAAK,CAAC;IAClB,IAAI,GAAG,GAAG,KAAK,CAAC;IAChB,KAAK,IAAI,CAAC,GAAG,KAAK,EAAE,CAAC,GAAG,KAAK,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;QAC1C,KAAK,MAAM,EAAE,IAAI,KAAK,CAAC,CAAC,CAAC,EAAE,CAAC;YAC1B,IAAI,EAAE,KAAK,GAAG,EAAE,CAAC;gBAAC,MAAM,EAAE,CAAC;gBAAC,KAAK,GAAG,IAAI,CAAC;YAAC,CAAC;YAC3C,IAAI,EAAE,KAAK,GAAG;gBAAE,MAAM,EAAE,CAAC;QAC3B,CAAC;QACD,GAAG,GAAG,CAAC,GAAG,CAAC,CAAC;QACZ,IAAI,KAAK,IAAI,MAAM,IAAI,CAAC;YAAE,MAAM;IAClC,CAAC;IACD,IAAI,CAAC,KAAK;QAAE,GAAG,GAAG,IAAI,CAAC,GAAG,CAAC,KAAK,GAAG,CAAC,EAAE,KAAK,CAAC,MAAM,CAAC,CAAC;IACpD,OAAO,EAAE,OAAO,EAAE,GAAG,EAAE,IAAI,EAAE,KAAK,CAAC,KAAK,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC;AACpE,CAAC;AAED,MAAM,UAAU,SAAS,CAAC,QAAgB,EAAE,MAAe;IACzD,MAAM,OAAO,GAAG,MAAM,IAAI,EAAE,CAAC,YAAY,CAAC,QAAQ,EAAE,OAAO,CAAC,CAAC;IAC7D,MAAM,QAAQ,GAAG,cAAc,CAAC,QAAQ,CAAC,CAAC;IAC1C,IAAI,QAAQ,KAAK,SAAS;QAAE,OAAO,EAAE,QAAQ,EAAE,EAAE,EAAE,KAAK,EAAE,EAAE,EAAE,CAAC;IAE/D,MAAM,KAAK,GAAG,OAAO,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC;IAClC,MAAM,QAAQ,GAAyB,EAAE,CAAC;IAC1C,MAAM,SAAS,GAAyB,EAAE,CAAC;IAE3C,cAAc;IACd,QAAQ,CAAC,IAAI,CAAC;QACZ,QAAQ,EAAE,IAAI,EAAE,IAAI,CAAC,QAAQ,CAAC,QAAQ,CAAC,EAAE,IAAI,EAAE,MAAM;QACrD,SAAS,EAAE,CAAC,EAAE,OAAO,EAAE,KAAK,CAAC,MAAM,EAAE,IAAI,EAAE,OAAO,CAAC,SAAS,CAAC,CAAC,EAAE,GAAG,CAAC;QACpE,QAAQ;KACT,CAAC,CAAC;IAEH,MAAM,OAAO,GAAG,QAAQ,CAAC;IAEzB,YAAY;IACZ,KAAK,MAAM,OAAO,IAAI,CAAC,iBAAiB,CAAC,OAAO,CAAC,IAAI,EAAE,CAAC,EAAE,CAAC;QACzD,MAAM,KAAK,GAAG,IAAI,MAAM,CAAC,OAAO,CAAC,MAAM,EAAE,OAAO,CAAC,KAAK,CAAC,CAAC;QACxD,IAAI,KAAK,CAAC;QACV,OAAO,CAAC,KAAK,GAAG,KAAK,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;YAC9C,MAAM,IAAI,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC;YACtB,MAAM,SAAS,GAAG,aAAa,CAAC,OAAO,EAAE,KAAK,CAAC,KAAK,CAAC,CAAC;YACtD,MAAM,KAAK,GAAG,YAAY,CAAC,KAAK,EAAE,SAAS,EAAE,QAAQ,CAAC,CAAC;YACvD,QAAQ,CAAC,IAAI,CAAC;gBACZ,QAAQ,EAAE,IAAI,EAAE,IAAI,EAAE,UAAU;gBAChC,SAAS,EAAE,OAAO,EAAE,KAAK,CAAC,OAAO,EAAE,IAAI,EAAE,KAAK,CAAC,IAAI;gBACnD,QAAQ;aACT,CAAC,CAAC;YACH,SAAS,CAAC,IAAI,CAAC,EAAE,UAAU,EAAE,IAAI,CAAC,QAAQ,CAAC,QAAQ,CAAC,EAAE,UAAU,EAAE,IAAI,EAAE,QAAQ,EAAE,UAAU,EAAE,CAAC,CAAC;QAClG,CAAC;IACH,CAAC;IAED,UAAU;IACV,KAAK,MAAM,OAAO,IAAI,CAAC,cAAc,CAAC,OAAO,CAAC,IAAI,EAAE,CAAC,EAAE,CAAC;QACtD,MAAM,KAAK,GAAG,IAAI,MAAM,CAAC,OAAO,CAAC,MAAM,EAAE,OAAO,CAAC,KAAK,CAAC,CAAC;QACxD,IAAI,KAAK,CAAC;QACV,OAAO,CAAC,KAAK,GAAG,KAAK,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;YAC9C,MAAM,IAAI,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC;YACtB,MAAM,SAAS,GAAG,aAAa,CAAC,OAAO,EAAE,KAAK,CAAC,KAAK,CAAC,CAAC;YACtD,MAAM,KAAK,GAAG,YAAY,CAAC,KAAK,EAAE,SAAS,EAAE,QAAQ,CAAC,CAAC;YACvD,QAAQ,CAAC,IAAI,CAAC;gBACZ,QAAQ,EAAE,IAAI,EAAE,IAAI,EAAE,OAAO;gBAC7B,SAAS,EAAE,OAAO,EAAE,KAAK,CAAC,OAAO,EAAE,IAAI,EAAE,KAAK,CAAC,IAAI;gBACnD,QAAQ;aACT,CAAC,CAAC;YACH,SAAS,CAAC,IAAI,CAAC,EAAE,UAAU,EAAE,IAAI,CAAC,QAAQ,CAAC,QAAQ,CAAC,EAAE,UAAU,EAAE,IAAI,EAAE,QAAQ,EAAE,UAAU,EAAE,CAAC,CAAC;QAClG,CAAC;IACH,CAAC;IAED,UAAU;IACV,KAAK,MAAM,OAAO,IAAI,CAAC,eAAe,CAAC,OAAO,CAAC,IAAI,EAAE,CAAC,EAAE,CAAC;QACvD,MAAM,KAAK,GAAG,IAAI,MAAM,CAAC,OAAO,CAAC,MAAM,EAAE,OAAO,CAAC,KAAK,CAAC,CAAC;QACxD,IAAI,KAAK,CAAC;QACV,OAAO,CAAC,KAAK,GAAG,KAAK,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;YAC9C,MAAM,SAAS,GAAG,aAAa,CAAC,OAAO,EAAE,KAAK,CAAC,KAAK,CAAC,CAAC;YACtD,MAAM,YAAY,GAAG,QAAQ,KAAK,YAAY,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,IAAI,KAAK,CAAC,CAAC,CAAC,IAAI,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC;YAC/F,MAAM,aAAa,GAAG,QAAQ,KAAK,YAAY;gBAC7C,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,IAAI,KAAK,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC,MAAM,CAAC,OAAO,CAAC;gBAC5E,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,IAAI,KAAK,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC;YAE/E,KAAK,MAAM,IAAI,IAAI,aAAa,EAAE,CAAC;gBACjC,QAAQ,CAAC,IAAI,CAAC;oBACZ,QAAQ,EAAE,IAAI,EAAE,IAAI,CAAC,OAAO,CAAC,aAAa,EAAE,EAAE,CAAC,EAAE,IAAI,EAAE,QAAQ;oBAC/D,SAAS,EAAE,OAAO,EAAE,SAAS,EAAE,IAAI,EAAE,KAAK,CAAC,CAAC,CAAC;oBAC7C,QAAQ;iBACT,CAAC,CAAC;gBACH,SAAS,CAAC,IAAI,CAAC;oBACb,UAAU,EAAE,IAAI,CAAC,QAAQ,CAAC,QAAQ,CAAC;oBACnC,UAAU,EAAE,IAAI,CAAC,OAAO,CAAC,aAAa,EAAE,EAAE,CAAC;oBAC3C,QAAQ,EAAE,SAAS;iBACpB,CAAC,CAAC;YACL,CAAC;QACH,CAAC;IACH,CAAC;IAED,OAAO,EAAE,QAAQ,EAAE,KAAK,EAAE,SAAS,EAAE,CAAC;AACxC,CAAC"}
|
package/package.json
ADDED
|
@@ -0,0 +1,54 @@
|
|
|
1
|
+
{
|
|
2
|
+
"name": "@phoenixaihub/graphrag",
|
|
3
|
+
"version": "0.1.0",
|
|
4
|
+
"description": "Graph RAG for private codebases — index code into a knowledge graph with semantic search and MCP server. Self-hosted, no code leaves your machine.",
|
|
5
|
+
"type": "module",
|
|
6
|
+
"main": "./dist/index.js",
|
|
7
|
+
"types": "./dist/index.d.ts",
|
|
8
|
+
"bin": {
|
|
9
|
+
"graphrag": "./dist/cli.js"
|
|
10
|
+
},
|
|
11
|
+
"scripts": {
|
|
12
|
+
"build": "tsc",
|
|
13
|
+
"dev": "tsc --watch",
|
|
14
|
+
"lint": "tsc --noEmit",
|
|
15
|
+
"test": "node --test dist/**/*.test.js",
|
|
16
|
+
"prepublishOnly": "npm run build"
|
|
17
|
+
},
|
|
18
|
+
"publishConfig": {
|
|
19
|
+
"access": "public"
|
|
20
|
+
},
|
|
21
|
+
"keywords": [
|
|
22
|
+
"rag",
|
|
23
|
+
"graph",
|
|
24
|
+
"codebase",
|
|
25
|
+
"knowledge-graph",
|
|
26
|
+
"mcp",
|
|
27
|
+
"tree-sitter",
|
|
28
|
+
"semantic-search",
|
|
29
|
+
"embeddings",
|
|
30
|
+
"code-intelligence"
|
|
31
|
+
],
|
|
32
|
+
"author": "Phoenix AI Hub",
|
|
33
|
+
"license": "MIT",
|
|
34
|
+
"repository": {
|
|
35
|
+
"type": "git",
|
|
36
|
+
"url": "https://github.com/phoenix-assistant/graph-rag-private-codebases.git"
|
|
37
|
+
},
|
|
38
|
+
"engines": {
|
|
39
|
+
"node": ">=18"
|
|
40
|
+
},
|
|
41
|
+
"dependencies": {
|
|
42
|
+
"better-sqlite3": "^11.7.0",
|
|
43
|
+
"commander": "^13.1.0",
|
|
44
|
+
"glob": "^11.0.1",
|
|
45
|
+
"openai": "^4.78.0",
|
|
46
|
+
"@anthropic-ai/sdk": "^0.39.0",
|
|
47
|
+
"@modelcontextprotocol/sdk": "^1.12.1"
|
|
48
|
+
},
|
|
49
|
+
"devDependencies": {
|
|
50
|
+
"@types/better-sqlite3": "^7.6.12",
|
|
51
|
+
"@types/node": "^22.10.0",
|
|
52
|
+
"typescript": "^5.7.0"
|
|
53
|
+
}
|
|
54
|
+
}
|
package/src/cli.ts
ADDED
|
@@ -0,0 +1,52 @@
|
|
|
1
|
+
#!/usr/bin/env node
|
|
2
|
+
import { Command } from 'commander';
|
|
3
|
+
import { indexRepo } from './commands/index-repo.js';
|
|
4
|
+
import { queryCode } from './commands/query.js';
|
|
5
|
+
import { serve } from './commands/serve.js';
|
|
6
|
+
|
|
7
|
+
const program = new Command();
|
|
8
|
+
|
|
9
|
+
program
|
|
10
|
+
.name('graphrag')
|
|
11
|
+
.description('Graph RAG for private codebases — self-hosted code intelligence')
|
|
12
|
+
.version('0.1.0');
|
|
13
|
+
|
|
14
|
+
program
|
|
15
|
+
.command('index <repo>')
|
|
16
|
+
.description('Index a codebase into a knowledge graph')
|
|
17
|
+
.option('--extensions <exts>', 'File extensions to index (comma-separated)', 'ts,tsx,js,jsx,py')
|
|
18
|
+
.option('--db <path>', 'Database path', '.graphrag/index.db')
|
|
19
|
+
.option('--embed', 'Generate embeddings (requires OPENAI_API_KEY)', false)
|
|
20
|
+
.action(async (repo: string, opts) => {
|
|
21
|
+
await indexRepo(repo, {
|
|
22
|
+
extensions: opts.extensions.split(','),
|
|
23
|
+
dbPath: opts.db,
|
|
24
|
+
embed: opts.embed,
|
|
25
|
+
});
|
|
26
|
+
});
|
|
27
|
+
|
|
28
|
+
program
|
|
29
|
+
.command('query <question>')
|
|
30
|
+
.description('Semantic search over indexed codebase with graph context')
|
|
31
|
+
.option('--db <path>', 'Database path', '.graphrag/index.db')
|
|
32
|
+
.option('--limit <n>', 'Max results', '5')
|
|
33
|
+
.action(async (question: string, opts) => {
|
|
34
|
+
await queryCode(question, {
|
|
35
|
+
dbPath: opts.db,
|
|
36
|
+
limit: parseInt(opts.limit, 10),
|
|
37
|
+
});
|
|
38
|
+
});
|
|
39
|
+
|
|
40
|
+
program
|
|
41
|
+
.command('serve')
|
|
42
|
+
.description('Start MCP server for code intelligence')
|
|
43
|
+
.option('--db <path>', 'Database path', '.graphrag/index.db')
|
|
44
|
+
.option('--port <n>', 'Port number', '3700')
|
|
45
|
+
.action(async (opts) => {
|
|
46
|
+
await serve({
|
|
47
|
+
dbPath: opts.db,
|
|
48
|
+
port: parseInt(opts.port, 10),
|
|
49
|
+
});
|
|
50
|
+
});
|
|
51
|
+
|
|
52
|
+
program.parse();
|
|
@@ -0,0 +1,96 @@
|
|
|
1
|
+
import path from 'path';
|
|
2
|
+
import { glob } from 'glob';
|
|
3
|
+
import { GraphDB } from '../db.js';
|
|
4
|
+
import { parseFile } from '../parser.js';
|
|
5
|
+
import { getEmbeddings } from '../embeddings.js';
|
|
6
|
+
|
|
7
|
+
interface IndexOptions {
|
|
8
|
+
extensions: string[];
|
|
9
|
+
dbPath: string;
|
|
10
|
+
embed: boolean;
|
|
11
|
+
}
|
|
12
|
+
|
|
13
|
+
export async function indexRepo(repoPath: string, opts: IndexOptions): Promise<void> {
|
|
14
|
+
const absRepo = path.resolve(repoPath);
|
|
15
|
+
const dbPath = path.isAbsolute(opts.dbPath) ? opts.dbPath : path.join(absRepo, opts.dbPath);
|
|
16
|
+
const db = new GraphDB(dbPath);
|
|
17
|
+
|
|
18
|
+
const patterns = opts.extensions.map(ext => `**/*.${ext}`);
|
|
19
|
+
const files = await glob(patterns, {
|
|
20
|
+
cwd: absRepo,
|
|
21
|
+
ignore: ['**/node_modules/**', '**/dist/**', '**/.git/**', '**/venv/**', '**/__pycache__/**'],
|
|
22
|
+
absolute: true,
|
|
23
|
+
});
|
|
24
|
+
|
|
25
|
+
console.log(`Indexing ${files.length} files from ${absRepo}`);
|
|
26
|
+
|
|
27
|
+
const entityNameToId = new Map<string, number>();
|
|
28
|
+
let totalEntities = 0;
|
|
29
|
+
let totalEdges = 0;
|
|
30
|
+
const pendingEdges: Array<{ sourceName: string; targetName: string; relation: string }> = [];
|
|
31
|
+
const chunksToEmbed: Array<{ entityId: number; content: string }> = [];
|
|
32
|
+
|
|
33
|
+
for (const file of files) {
|
|
34
|
+
const relPath = path.relative(absRepo, file);
|
|
35
|
+
db.clearFile(relPath);
|
|
36
|
+
|
|
37
|
+
const result = parseFile(file);
|
|
38
|
+
|
|
39
|
+
for (const entity of result.entities) {
|
|
40
|
+
const e = { ...entity, filePath: relPath };
|
|
41
|
+
const id = db.insertEntity(e);
|
|
42
|
+
entityNameToId.set(e.name, id);
|
|
43
|
+
totalEntities++;
|
|
44
|
+
|
|
45
|
+
// Create chunk for non-import entities
|
|
46
|
+
if (e.kind !== 'import') {
|
|
47
|
+
chunksToEmbed.push({ entityId: id, content: `${e.kind} ${e.name}\n${e.code}` });
|
|
48
|
+
}
|
|
49
|
+
}
|
|
50
|
+
|
|
51
|
+
pendingEdges.push(...result.edges);
|
|
52
|
+
}
|
|
53
|
+
|
|
54
|
+
// Resolve and insert edges
|
|
55
|
+
for (const edge of pendingEdges) {
|
|
56
|
+
const sourceId = entityNameToId.get(edge.sourceName);
|
|
57
|
+
const targetId = entityNameToId.get(edge.targetName);
|
|
58
|
+
if (sourceId && targetId) {
|
|
59
|
+
db.insertEdge({ sourceId, targetId, relation: edge.relation as any });
|
|
60
|
+
totalEdges++;
|
|
61
|
+
}
|
|
62
|
+
}
|
|
63
|
+
|
|
64
|
+
// Generate embeddings in batches
|
|
65
|
+
if (opts.embed && process.env.OPENAI_API_KEY) {
|
|
66
|
+
console.log(`Generating embeddings for ${chunksToEmbed.length} chunks...`);
|
|
67
|
+
const batchSize = 100;
|
|
68
|
+
for (let i = 0; i < chunksToEmbed.length; i += batchSize) {
|
|
69
|
+
const batch = chunksToEmbed.slice(i, i + batchSize);
|
|
70
|
+
const embeddings = await getEmbeddings(batch.map(c => c.content));
|
|
71
|
+
for (let j = 0; j < batch.length; j++) {
|
|
72
|
+
db.insertChunk({
|
|
73
|
+
entityId: batch[j].entityId,
|
|
74
|
+
content: batch[j].content,
|
|
75
|
+
embedding: new Float64Array(embeddings[j]),
|
|
76
|
+
});
|
|
77
|
+
}
|
|
78
|
+
console.log(` Embedded ${Math.min(i + batchSize, chunksToEmbed.length)}/${chunksToEmbed.length}`);
|
|
79
|
+
}
|
|
80
|
+
} else {
|
|
81
|
+
// Store chunks without embeddings
|
|
82
|
+
for (const chunk of chunksToEmbed) {
|
|
83
|
+
db.insertChunk({ entityId: chunk.entityId, content: chunk.content });
|
|
84
|
+
}
|
|
85
|
+
}
|
|
86
|
+
|
|
87
|
+
const stats = db.getStats();
|
|
88
|
+
console.log(`\nDone! Indexed:`);
|
|
89
|
+
console.log(` Files: ${stats.files}`);
|
|
90
|
+
console.log(` Entities: ${stats.entities}`);
|
|
91
|
+
console.log(` Edges: ${stats.edges}`);
|
|
92
|
+
console.log(` Chunks: ${stats.chunks}`);
|
|
93
|
+
console.log(`\nDatabase: ${dbPath}`);
|
|
94
|
+
|
|
95
|
+
db.close();
|
|
96
|
+
}
|
|
@@ -0,0 +1,43 @@
|
|
|
1
|
+
import path from 'path';
|
|
2
|
+
import { GraphDB } from '../db.js';
|
|
3
|
+
|
|
4
|
+
interface QueryOptions {
|
|
5
|
+
dbPath: string;
|
|
6
|
+
limit: number;
|
|
7
|
+
}
|
|
8
|
+
|
|
9
|
+
export async function queryCode(question: string, opts: QueryOptions): Promise<void> {
|
|
10
|
+
const dbPath = path.resolve(opts.dbPath);
|
|
11
|
+
const db = new GraphDB(dbPath);
|
|
12
|
+
|
|
13
|
+
const results = db.searchEntities(question, opts.limit);
|
|
14
|
+
|
|
15
|
+
if (results.length === 0) {
|
|
16
|
+
console.log('No results found.');
|
|
17
|
+
db.close();
|
|
18
|
+
return;
|
|
19
|
+
}
|
|
20
|
+
|
|
21
|
+
console.log(`Found ${results.length} results:\n`);
|
|
22
|
+
|
|
23
|
+
for (const entity of results) {
|
|
24
|
+
const id = (entity as any).id as number;
|
|
25
|
+
console.log(`━━━ ${entity.kind}: ${entity.name} ━━━`);
|
|
26
|
+
console.log(`File: ${entity.filePath}:${entity.startLine}-${entity.endLine}`);
|
|
27
|
+
|
|
28
|
+
// Get related entities for context
|
|
29
|
+
const related = db.getRelated(id);
|
|
30
|
+
if (related.length > 0) {
|
|
31
|
+
console.log(`Relations:`);
|
|
32
|
+
for (const r of related.slice(0, 5)) {
|
|
33
|
+
console.log(` ${r.relation} → ${r.entity.name} (${r.entity.kind})`);
|
|
34
|
+
}
|
|
35
|
+
}
|
|
36
|
+
|
|
37
|
+
console.log(`\n${entity.code.substring(0, 300)}`);
|
|
38
|
+
if (entity.code.length > 300) console.log(' ...');
|
|
39
|
+
console.log('');
|
|
40
|
+
}
|
|
41
|
+
|
|
42
|
+
db.close();
|
|
43
|
+
}
|
|
@@ -0,0 +1,90 @@
|
|
|
1
|
+
import { McpServer } from '@modelcontextprotocol/sdk/server/mcp.js';
|
|
2
|
+
import { StdioServerTransport } from '@modelcontextprotocol/sdk/server/stdio.js';
|
|
3
|
+
import { z } from 'zod';
|
|
4
|
+
import path from 'path';
|
|
5
|
+
import { GraphDB } from '../db.js';
|
|
6
|
+
|
|
7
|
+
interface ServeOptions {
|
|
8
|
+
dbPath: string;
|
|
9
|
+
port: number;
|
|
10
|
+
}
|
|
11
|
+
|
|
12
|
+
export async function serve(opts: ServeOptions): Promise<void> {
|
|
13
|
+
const dbPath = path.resolve(opts.dbPath);
|
|
14
|
+
const db = new GraphDB(dbPath);
|
|
15
|
+
|
|
16
|
+
const server = new McpServer({
|
|
17
|
+
name: 'graphrag',
|
|
18
|
+
version: '0.1.0',
|
|
19
|
+
});
|
|
20
|
+
|
|
21
|
+
server.tool(
|
|
22
|
+
'search_codebase',
|
|
23
|
+
'Search the indexed codebase for functions, classes, and code patterns',
|
|
24
|
+
{ query: z.string().describe('Search query'), limit: z.number().optional().describe('Max results (default 5)') },
|
|
25
|
+
async ({ query, limit }) => {
|
|
26
|
+
const results = db.searchEntities(query, limit ?? 5);
|
|
27
|
+
return {
|
|
28
|
+
content: [{
|
|
29
|
+
type: 'text' as const,
|
|
30
|
+
text: JSON.stringify(results.map(e => ({
|
|
31
|
+
name: e.name,
|
|
32
|
+
kind: e.kind,
|
|
33
|
+
file: `${e.filePath}:${e.startLine}-${e.endLine}`,
|
|
34
|
+
code: e.code.substring(0, 500),
|
|
35
|
+
})), null, 2),
|
|
36
|
+
}],
|
|
37
|
+
};
|
|
38
|
+
}
|
|
39
|
+
);
|
|
40
|
+
|
|
41
|
+
server.tool(
|
|
42
|
+
'get_function_context',
|
|
43
|
+
'Get a function/class with its full graph context (callers, callees, imports)',
|
|
44
|
+
{ name: z.string().describe('Function or class name') },
|
|
45
|
+
async ({ name }) => {
|
|
46
|
+
const entities = db.findEntitiesByName(name);
|
|
47
|
+
if (entities.length === 0) {
|
|
48
|
+
return { content: [{ type: 'text' as const, text: `No entity found: ${name}` }] };
|
|
49
|
+
}
|
|
50
|
+
const entity = entities[0];
|
|
51
|
+
const id = (entity as any).id as number;
|
|
52
|
+
const related = db.getRelated(id);
|
|
53
|
+
return {
|
|
54
|
+
content: [{
|
|
55
|
+
type: 'text' as const,
|
|
56
|
+
text: JSON.stringify({
|
|
57
|
+
entity: { name: entity.name, kind: entity.kind, file: entity.filePath, code: entity.code },
|
|
58
|
+
related: related.map(r => ({
|
|
59
|
+
name: r.entity.name, kind: r.entity.kind, relation: r.relation,
|
|
60
|
+
file: r.entity.filePath,
|
|
61
|
+
})),
|
|
62
|
+
}, null, 2),
|
|
63
|
+
}],
|
|
64
|
+
};
|
|
65
|
+
}
|
|
66
|
+
);
|
|
67
|
+
|
|
68
|
+
server.tool(
|
|
69
|
+
'find_usages',
|
|
70
|
+
'Find all usages of a function, class, or variable',
|
|
71
|
+
{ name: z.string().describe('Entity name to find usages of') },
|
|
72
|
+
async ({ name }) => {
|
|
73
|
+
const usages = db.getUsages(name);
|
|
74
|
+
return {
|
|
75
|
+
content: [{
|
|
76
|
+
type: 'text' as const,
|
|
77
|
+
text: JSON.stringify(usages.map(e => ({
|
|
78
|
+
name: e.name, kind: e.kind,
|
|
79
|
+
file: `${e.filePath}:${e.startLine}-${e.endLine}`,
|
|
80
|
+
code: e.code.substring(0, 200),
|
|
81
|
+
})), null, 2),
|
|
82
|
+
}],
|
|
83
|
+
};
|
|
84
|
+
}
|
|
85
|
+
);
|
|
86
|
+
|
|
87
|
+
const transport = new StdioServerTransport();
|
|
88
|
+
await server.connect(transport);
|
|
89
|
+
console.error('GraphRAG MCP server running on stdio');
|
|
90
|
+
}
|