@gracker/smartperfetto 1.0.14 → 1.0.16

package/dist/services/rag/appSourceIngester.js

@@ -0,0 +1,165 @@
+"use strict";
+// SPDX-License-Identifier: AGPL-3.0-or-later
+// Copyright (C) 2024-2026 Gracker (Chris)
+// This file is part of SmartPerfetto. See LICENSE for details.
+var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    var desc = Object.getOwnPropertyDescriptor(m, k);
+    if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
+      desc = { enumerable: true, get: function() { return m[k]; } };
+    }
+    Object.defineProperty(o, k2, desc);
+}) : (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    o[k2] = m[k];
+}));
+var __setModuleDefault = (this && this.__setModuleDefault) || (Object.create ? (function(o, v) {
+    Object.defineProperty(o, "default", { enumerable: true, value: v });
+}) : function(o, v) {
+    o["default"] = v;
+});
+var __importStar = (this && this.__importStar) || (function () {
+    var ownKeys = function(o) {
+        ownKeys = Object.getOwnPropertyNames || function (o) {
+            var ar = [];
+            for (var k in o) if (Object.prototype.hasOwnProperty.call(o, k)) ar[ar.length] = k;
+            return ar;
+        };
+        return ownKeys(o);
+    };
+    return function (mod) {
+        if (mod && mod.__esModule) return mod;
+        var result = {};
+        if (mod != null) for (var k = ownKeys(mod), i = 0; i < k.length; i++) if (k[i] !== "default") __createBinding(result, mod, k[i]);
+        __setModuleDefault(result, mod);
+        return result;
+    };
+})();
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.__TEST_ONLY__ = exports.AppSourceIngester = void 0;
+const crypto_1 = require("crypto");
+const fs = __importStar(require("fs"));
+const path = __importStar(require("path"));
+const secretPatterns_1 = require("../security/secretPatterns");
+const pathSecurityGate_1 = require("../codebase/pathSecurityGate");
+const baseIngester_1 = require("./baseIngester");
+const DEFAULT_MAX_CHUNK_CHARS = 2200;
+function makeChunkId(codebaseId, indexGeneration, relativePath, startLine) {
+    return (0, crypto_1.createHash)('sha256')
+        .update(`${codebaseId}|${indexGeneration}|${relativePath}|${startLine}`)
+        .digest('hex')
+        .slice(0, 16);
+}
+const chunkSource = baseIngester_1.chunkSourceBySymbols;
+function filterPreviewFiles(preview, ref, opts) {
+    const prefixes = [
+        ...(ref.pathFilters ?? []),
+        ...(opts.pathPrefix ? [opts.pathPrefix] : []),
+    ].filter(Boolean);
+    if (prefixes.length === 0)
+        return preview.acceptedFiles;
+    return preview.acceptedFiles.filter(file => prefixes.some(prefix => file.relativePath.startsWith(prefix.replace(/^\//, ''))));
+}
+class AppSourceIngester {
+    constructor(store, registry, gate = new pathSecurityGate_1.PathSecurityGate()) {
+        this.store = store;
+        this.registry = registry;
+        this.gate = gate;
+    }
+    ingest(codebaseId, opts = {}) {
+        const ref = this.registry.get(codebaseId);
+        if (!ref) {
+            throw new Error(`Codebase '${codebaseId}' not found`);
+        }
+        if (ref.kind !== 'app_source') {
+            throw new Error(`Codebase '${codebaseId}' is kind=${ref.kind}; app source ingestion requires app_source`);
+        }
+        const preview = this.gate.preview(ref.rootRealpath);
+        if (preview.blocked) {
+            this.registry.updateIngestStatus(codebaseId, {
+                lastIngestStatus: 'blocked_by_security',
+                lastIngestAt: Date.now(),
+                lastIngestError: preview.blockedReason,
+                blockedFileCount: preview.skippedFiles.length,
+            });
+            return {
+                codebaseId,
+                filesProcessed: 0,
+                chunksAdded: 0,
+                chunksSkipped: 0,
+                blockedFileCount: preview.skippedFiles.length,
+                redactionHitCount: 0,
+                errors: [{ filePath: ref.displayName, reason: preview.blockedReason ?? 'blocked' }],
+            };
+        }
+        const maxChars = opts.maxChunkChars ?? DEFAULT_MAX_CHUNK_CHARS;
+        const result = {
+            codebaseId,
+            filesProcessed: 0,
+            chunksAdded: 0,
+            chunksSkipped: 0,
+            blockedFileCount: preview.skippedFiles.length,
+            redactionHitCount: 0,
+            errors: [],
+        };
+        for (const file of filterPreviewFiles(preview, ref, opts)) {
+            result.filesProcessed++;
+            try {
+                const absolutePath = path.join(ref.rootRealpath, file.relativePath);
+                const content = fs.readFileSync(absolutePath, 'utf-8');
+                const chunks = chunkSource(content, maxChars);
+                if (chunks.length === 0) {
+                    result.chunksSkipped++;
+                    continue;
+                }
+                for (const chunk of chunks) {
+                    const redaction = (0, secretPatterns_1.redactSecrets)(chunk.text);
+                    result.redactionHitCount += redaction.redactedCount;
+                    this.store.addChunk({
+                        chunkId: makeChunkId(codebaseId, ref.indexGeneration, file.relativePath, chunk.startLine),
+                        kind: 'app_source',
+                        uri: `codebase://${codebaseId}/${file.relativePath}`,
+                        title: file.relativePath.split('/').pop(),
+                        snippet: chunk.text,
+                        tokenCount: (0, baseIngester_1.estimateTokenCount)(chunk.text),
+                        indexedAt: Date.now(),
+                        filePath: file.relativePath,
+                        lineRange: { start: chunk.startLine, end: chunk.endLine },
+                        ...(chunk.symbol ? { symbol: chunk.symbol } : {}),
+                        language: (0, baseIngester_1.languageForPath)(file.relativePath),
+                        ...(ref.commitHash ? { commitHash: ref.commitHash } : {}),
+                        ...(ref.vendor ? { vendor: ref.vendor } : {}),
+                        ...(ref.buildId ? { buildId: ref.buildId } : {}),
+                        codebaseId,
+                        registryOrigin: 'codebase_registry',
+                    });
+                    result.chunksAdded++;
+                }
+            }
+            catch (error) {
+                result.chunksSkipped++;
+                result.errors.push({
+                    filePath: file.relativePath,
+                    reason: error instanceof Error ? error.message : String(error),
+                });
+            }
+        }
+        this.registry.updateIngestStatus(codebaseId, {
+            lastIngestStatus: result.errors.length > 0 ? 'partial' : 'ok',
+            lastIngestAt: Date.now(),
+            lastIngestError: result.errors[0]?.reason,
+            chunkCount: result.chunksAdded,
+            blockedFileCount: result.blockedFileCount,
+            redactionHitCount: result.redactionHitCount,
+        });
+        return result;
+    }
+}
+exports.AppSourceIngester = AppSourceIngester;
+exports.__TEST_ONLY__ = {
+    chunkSource,
+    detectSymbol: baseIngester_1.detectSourceSymbol,
+    makeChunkId,
+    languageFor: baseIngester_1.languageForPath,
+};
+//# sourceMappingURL=appSourceIngester.js.map

package/dist/services/rag/appSourceIngester.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"appSourceIngester.js","sourceRoot":"","sources":["../../../src/services/rag/appSourceIngester.ts"],"names":[],"mappings":";AAAA,6CAA6C;AAC7C,0CAA0C;AAC1C,+DAA+D;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AAE/D,mCAAkC;AAClC,uCAAyB;AACzB,2CAA6B;AAG7B,+DAAyD;AAEzD,mEAAsF;AACtF,iDAKwB;AAExB,MAAM,uBAAuB,GAAG,IAAI,CAAC;AAsBrC,SAAS,WAAW,CAAC,UAAkB,EAAE,eAAuB,EAAE,YAAoB,EAAE,SAAiB;IACvG,OAAO,IAAA,mBAAU,EAAC,QAAQ,CAAC;SACxB,MAAM,CAAC,GAAG,UAAU,IAAI,eAAe,IAAI,YAAY,IAAI,SAAS,EAAE,CAAC;SACvE,MAAM,CAAC,KAAK,CAAC;SACb,KAAK,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC;AAClB,CAAC;AAED,MAAM,WAAW,GAAG,mCAAoB,CAAC;AAEzC,SAAS,kBAAkB,CAAC,OAA0B,EAAE,GAAgB,EAAE,IAA4B;IACpG,MAAM,QAAQ,GAAG;QACf,GAAG,CAAC,GAAG,CAAC,WAAW,IAAI,EAAE,CAAC;QAC1B,GAAG,CAAC,IAAI,CAAC,UAAU,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,UAAU,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC;KAC9C,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC;IAClB,IAAI,QAAQ,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,OAAO,CAAC,aAAa,CAAC;IACxD,OAAO,OAAO,CAAC,aAAa,CAAC,MAAM,CAAC,IAAI,CAAC,EAAE,CACzC,QAAQ,CAAC,IAAI,CAAC,MAAM,CAAC,EAAE,CAAC,IAAI,CAAC,YAAY,CAAC,UAAU,CAAC,MAAM,CAAC,OAAO,CAAC,KAAK,EAAE,EAAE,CAAC,CAAC,CAAC,CACjF,CAAC;AACJ,CAAC;AAED,MAAa,iBAAiB;IAC5B,YACmB,KAAe,EACf,QAA0B,EAC1B,OAAyB,IAAI,mCAAgB,EAAE;QAF/C,UAAK,GAAL,KAAK,CAAU;QACf,aAAQ,GAAR,QAAQ,CAAkB;QAC1B,SAAI,GAAJ,IAAI,CAA2C;IAC/D,CAAC;IAEJ,MAAM,CAAC,UAAkB,EAAE,OAA+B,EAAE;QAC1D,MAAM,GAAG,GAAG,IAAI,CAAC,QAAQ,CAAC,GAAG,CAAC,UAAU,CAAC,CAAC;QAC1C,IAAI,CAAC,GAAG,EAAE,CAAC;YACT,MAAM,IAAI,KAAK,CAAC,aAAa,UAAU,aAAa,CAAC,CAAC;QACxD,CAAC;QACD,IAAI,GAAG,CAAC,IAAI,KAAK,YAAY,EAAE,CAAC;YAC9B,MAAM,IAAI,KAAK,CAAC,aAAa,UAAU,aAAa,GAAG,CAAC,IAAI,4CAA4C,CAAC,CAAC;QAC5G,CAAC;QAED,MAAM,OAAO,GAAG,IAAI,CAAC,IAAI,CAAC,OAAO,CAAC,GAAG,CAAC,YAAY,CAAC,CAAC;QACpD,IAAI,OAAO,CAAC,OAAO,EAAE,CAAC;YACpB,IAAI,CAAC,QAAQ,CAAC,kBAAkB,CAAC,UAAU,EAAE;gBAC3C,gBAAgB,EAAE,qBAAqB;gBACvC,YAAY,EAAE,IAAI,CAAC,GAAG,EAAE;gBACxB,eAAe,EAAE,OAAO,CAAC,aAAa;gBACtC,gBAAgB,EAAE,OAAO,CAAC,YAAY,CAAC,MAAM;aAC9C,CAAC,CAAC;YACH,OAAO;gBACL,UAAU;gBACV,cAAc,EAAE,CAAC;gBACjB,WAAW,EAAE,CAAC;gBACd,aAAa,EAAE,CAAC;gBAChB,gBAAgB,EAAE,OAAO,CAAC,YAAY,CAAC,MAAM;gBAC7C,iBAAiB,EAAE,CAAC;gBACpB,MAAM,EAAE,CAAC,EAAC,QAAQ,EAAE,GAAG,CAAC,WAAW,EAAE,MAAM,EAAE,OAAO,CAAC,aAAa,IAAI,SAAS,EAAC,CAAC;aAClF,CAAC;QACJ,CAAC;QAED,MAAM,QAAQ,GAAG,IAAI,CAAC,aAAa,IAAI,uBAAuB,CAAC;QAC/D,MAAM,MAAM,GAA0B;YACpC,UAAU;YACV,cAAc,EAAE,CAAC;YACjB,WAAW,EAAE,CAAC;YACd,aAAa,EAAE,CAAC;YAChB,gBAAgB,EAAE,OAAO,CAAC,YAAY,CAAC,MAAM;YAC7C,iBAAiB,EAAE,CAAC;YACpB,MAAM,EAAE,EAAE;SACX,CAAC;QAEF,KAAK,MAAM,IAAI,IAAI,kBAAkB,CAAC,OAAO,EAAE,GAAG,EAAE,IAAI,CAAC,EAAE,CAAC;YAC1D,MAAM,CAAC,cAAc,EAAE,CAAC;YACxB,IAAI,CAAC;gBACH,MAAM,YAAY,GAAG,IAAI,CAAC,IAAI,CAAC,GAAG,CAAC,YAAY,EAAE,IAAI,CAAC,YAAY,CAAC,CAAC;gBACpE,MAAM,OAAO,GAAG,EAAE,CAAC,YAAY,CAAC,YAAY,EAAE,OAAO,CAAC,CAAC;gBACvD,MAAM,MAAM,GAAG,WAAW,CAAC,OAAO,EAAE,QAAQ,CAAC,CAAC;gBAC9C,IAAI,MAAM,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;oBACxB,MAAM,CAAC,aAAa,EAAE,CAAC;oBACvB,SAAS;gBACX,CAAC;gBACD,KAAK,MAAM,KAAK,IAAI,MAAM,EAAE,CAAC;oBAC3B,MAAM,SAAS,GAAG,IAAA,8BAAa,EAAC,KAAK,CAAC,IAAI,CAAC,CAAC;oBAC5C,MAAM,CAAC,iBAAiB,IAAI,SAAS,CAAC,aAAa,CAAC;oBACpD,IAAI,CAAC,KAAK,CAAC,QAAQ,CAAC;wBAClB,OAAO,EAAE,WAAW,CAAC,UAAU,EAAE,GAAG,CAAC,eAAe,EAAE,IAAI,CAAC,YAAY,EAAE,KAAK,CAAC,SAAS,CAAC;wBACzF,IAAI,EAAE,YAAY;wBAClB,GAAG,EAAE,cAAc,UAAU,IAAI,IAAI,CAAC,YAAY,EAAE;wBACpD,KAAK,EAAE,IAAI,CAAC,YAAY,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,GAAG,EAAE;wBACzC,OAAO,EAAE,KAAK,CAAC,IAAI;wBACnB,UAAU,EAAE,IAAA,iCAAkB,EAAC,KAAK,CAAC,IAAI,CAAC;wBAC1C,SAAS,EAAE,IAAI,CAAC,GAAG,EAAE;wBACrB,QAAQ,EAAE,IAAI,CAAC,YAAY;wBAC3B,SAAS,EAAE,EAAC,KAAK,EAAE,KAAK,CAAC,SAAS,EAAE,GAAG,EAAE,KAAK,CAAC,OAAO,EAAC;wBACvD,GAAG,CAAC,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,EAAC,MAAM,EAAE,KAAK,CAAC,MAAM,EAAC,CAAC,CAAC,CAAC,EAAE,CAAC;wBAC/C,QAAQ,EAAE,IAAA,8BAAe,EAAC,IAAI,CAAC,YAAY,CAAC;wBAC5C,GAAG,CAAC,GAAG,CAAC,UAAU,CAAC,CAAC,CAAC,EAAC,UAAU,EAAE,GAAG,CAAC,UAAU,EAAC,CAAC,CAAC,CAAC,EAAE,CAAC;wBACvD,GAAG,CAAC,GAAG,CAAC,MAAM,CAAC,CAAC,CAAC,EAAC,MAAM,EAAE,GAAG,CAAC,MAAM,EAAC,CAAC,CAAC,CAAC,EAAE,CAAC;wBAC3C,GAAG,CAAC,GAAG,CAAC,OAAO,CAAC,CAAC,CAAC,EAAC,OAAO,EAAE,GAAG,CAAC,OAAO,EAAC,CAAC,CAAC,CAAC,EAAE,CAAC;wBAC9C,UAAU;wBACV,cAAc,EAAE,mBAAmB;qBACpC,CAAC,CAAC;oBACH,MAAM,CAAC,WAAW,EAAE,CAAC;gBACvB,CAAC;YACH,CAAC;YAAC,OAAO,KAAK,EAAE,CAAC;gBACf,MAAM,CAAC,aAAa,EAAE,CAAC;gBACvB,MAAM,CAAC,MAAM,CAAC,IAAI,CAAC;oBACjB,QAAQ,EAAE,IAAI,CAAC,YAAY;oBAC3B,MAAM,EAAE,KAAK,YAAY,KAAK,CAAC,CAAC,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,MAAM,CAAC,KAAK,CAAC;iBAC/D,CAAC,CAAC;YACL,CAAC;QACH,CAAC;QAED,IAAI,CAAC,QAAQ,CAAC,kBAAkB,CAAC,UAAU,EAAE;YAC3C,gBAAgB,EAAE,MAAM,CAAC,MAAM,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,IAAI;YAC7D,YAAY,EAAE,IAAI,CAAC,GAAG,EAAE;YACxB,eAAe,EAAE,MAAM,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,MAAM;YACzC,UAAU,EAAE,MAAM,CAAC,WAAW;YAC9B,gBAAgB,EAAE,MAAM,CAAC,gBAAgB;YACzC,iBAAiB,EAAE,MAAM,CAAC,iBAAiB;SAC5C,CAAC,CAAC;QACH,OAAO,MAAM,CAAC;IAChB,CAAC;CACF;AAlGD,8CAkGC;AAEY,QAAA,aAAa,GAAG;IAC3B,WAAW;IACX,YAAY,EAAE,iCAAkB;IAChC,WAAW;IACX,WAAW,EAAE,8BAAe;CAC7B,CAAC"}

package/dist/services/rag/baseIngester.d.ts

@@ -0,0 +1,18 @@
+import type { RagChunk } from '../../types/sparkContracts';
+export interface SourceChunk {
+    text: string;
+    startLine: number;
+    endLine: number;
+    symbol?: string;
+}
+export interface OffsetChunk {
+    text: string;
+    offset: number;
+}
+export declare function estimateTokenCount(text: string): number;
+export declare function stableChunkId(parts: readonly unknown[], length?: number): string;
+export declare function languageForPath(relativePath: string): NonNullable<RagChunk['language']>;
+export declare function detectSourceSymbol(line: string): string | undefined;
+export declare function chunkSourceBySymbols(content: string, maxChars: number): SourceChunk[];
+export declare function chunkTextByParagraphs(text: string, maxChars: number): OffsetChunk[];
+//# sourceMappingURL=baseIngester.d.ts.map

package/dist/services/rag/baseIngester.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"baseIngester.d.ts","sourceRoot":"","sources":["../../../src/services/rag/baseIngester.ts"],"names":[],"mappings":"AAOA,OAAO,KAAK,EAAC,QAAQ,EAAC,MAAM,4BAA4B,CAAC;AAEzD,MAAM,WAAW,WAAW;IAC1B,IAAI,EAAE,MAAM,CAAC;IACb,SAAS,EAAE,MAAM,CAAC;IAClB,OAAO,EAAE,MAAM,CAAC;IAChB,MAAM,CAAC,EAAE,MAAM,CAAC;CACjB;AAED,MAAM,WAAW,WAAW;IAC1B,IAAI,EAAE,MAAM,CAAC;IACb,MAAM,EAAE,MAAM,CAAC;CAChB;AAED,wBAAgB,kBAAkB,CAAC,IAAI,EAAE,MAAM,GAAG,MAAM,CAEvD;AAED,wBAAgB,aAAa,CAAC,KAAK,EAAE,SAAS,OAAO,EAAE,EAAE,MAAM,SAAK,GAAG,MAAM,CAK5E;AAED,wBAAgB,eAAe,CAAC,YAAY,EAAE,MAAM,GAAG,WAAW,CAAC,QAAQ,CAAC,UAAU,CAAC,CAAC,CAUvF;AAED,wBAAgB,kBAAkB,CAAC,IAAI,EAAE,MAAM,GAAG,MAAM,GAAG,SAAS,CASnE;AAED,wBAAgB,oBAAoB,CAAC,OAAO,EAAE,MAAM,EAAE,QAAQ,EAAE,MAAM,GAAG,WAAW,EAAE,CA2ErF;AAED,wBAAgB,qBAAqB,CAAC,IAAI,EAAE,MAAM,EAAE,QAAQ,EAAE,MAAM,GAAG,WAAW,EAAE,CA0BnF"}

package/dist/services/rag/baseIngester.js

@@ -0,0 +1,197 @@
+"use strict";
+// SPDX-License-Identifier: AGPL-3.0-or-later
+// Copyright (C) 2024-2026 Gracker (Chris)
+// This file is part of SmartPerfetto. See LICENSE for details.
+var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    var desc = Object.getOwnPropertyDescriptor(m, k);
+    if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
+      desc = { enumerable: true, get: function() { return m[k]; } };
+    }
+    Object.defineProperty(o, k2, desc);
+}) : (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    o[k2] = m[k];
+}));
+var __setModuleDefault = (this && this.__setModuleDefault) || (Object.create ? (function(o, v) {
+    Object.defineProperty(o, "default", { enumerable: true, value: v });
+}) : function(o, v) {
+    o["default"] = v;
+});
+var __importStar = (this && this.__importStar) || (function () {
+    var ownKeys = function(o) {
+        ownKeys = Object.getOwnPropertyNames || function (o) {
+            var ar = [];
+            for (var k in o) if (Object.prototype.hasOwnProperty.call(o, k)) ar[ar.length] = k;
+            return ar;
+        };
+        return ownKeys(o);
+    };
+    return function (mod) {
+        if (mod && mod.__esModule) return mod;
+        var result = {};
+        if (mod != null) for (var k = ownKeys(mod), i = 0; i < k.length; i++) if (k[i] !== "default") __createBinding(result, mod, k[i]);
+        __setModuleDefault(result, mod);
+        return result;
+    };
+})();
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.estimateTokenCount = estimateTokenCount;
+exports.stableChunkId = stableChunkId;
+exports.languageForPath = languageForPath;
+exports.detectSourceSymbol = detectSourceSymbol;
+exports.chunkSourceBySymbols = chunkSourceBySymbols;
+exports.chunkTextByParagraphs = chunkTextByParagraphs;
+const crypto_1 = require("crypto");
+const path = __importStar(require("path"));
+function estimateTokenCount(text) {
+    return Math.max(1, Math.round(text.length / 4));
+}
+function stableChunkId(parts, length = 16) {
+    return (0, crypto_1.createHash)('sha256')
+        .update(parts.map(part => String(part)).join('|'))
+        .digest('hex')
+        .slice(0, length);
+}
+function languageForPath(relativePath) {
+    const ext = path.extname(relativePath);
+    if (ext === '.kt' || ext === '.kts')
+        return 'kotlin';
+    if (ext === '.java')
+        return 'java';
+    if (ext === '.c')
+        return 'c';
+    if (ext === '.cc' || ext === '.cpp' || ext === '.cxx' || ext === '.h' || ext === '.hpp')
+        return 'cpp';
+    if (ext === '.rs')
+        return 'rust';
+    if (ext === '.go')
+        return 'go';
+    if (ext === '.py')
+        return 'py';
+    return 'unknown';
+}
+function detectSourceSymbol(line) {
+    const classMatch = line.match(/\b(?:class|object|interface|enum\s+class|struct)\s+([A-Za-z_][A-Za-z0-9_]*)/);
+    if (classMatch?.[1])
+        return classMatch[1];
+    const kotlinFun = line.match(/\bfun\s+([A-Za-z_][A-Za-z0-9_]*)\s*\(/);
+    if (kotlinFun?.[1])
+        return kotlinFun[1];
+    const cLikeFunction = line.match(/\b(?:[A-Za-z_][A-Za-z0-9_:<>~*&\s]+)\s+([A-Za-z_~][A-Za-z0-9_:~]*)\s*\([^;]*\)\s*(?:const\s*)?(?:noexcept\s*)?(?:override\s*)?\{/);
+    if (cLikeFunction?.[1])
+        return cLikeFunction[1].split('::').pop();
+    const javaMethod = line.match(/\b(?:public|private|protected|internal|static|final|suspend|override|\s)+[A-Za-z0-9_<>,.?[\]\s]+\s+([A-Za-z_][A-Za-z0-9_]*)\s*\([^;]*\)\s*\{/);
+    return javaMethod?.[1];
+}
+function chunkSourceBySymbols(content, maxChars) {
+    const normalized = content.replace(/\r\n?/g, '\n');
+    if (normalized.trim().length === 0)
+        return [];
+    const lines = normalized.split('\n');
+    const boundaries = new Set([0]);
+    for (let i = 0; i < lines.length; i++) {
+        if (detectSourceSymbol(lines[i]))
+            boundaries.add(i);
+    }
+    boundaries.add(lines.length);
+    const sorted = Array.from(boundaries).sort((a, b) => a - b);
+    const sections = [];
+    for (let i = 0; i < sorted.length - 1; i++) {
+        const start = sorted[i];
+        const end = sorted[i + 1];
+        if (end <= start)
+            continue;
+        const text = lines.slice(start, end).join('\n').trim();
+        if (!text)
+            continue;
+        sections.push({
+            text,
+            startLine: start + 1,
+            endLine: end,
+            symbol: detectSourceSymbol(lines[start]),
+        });
+    }
+    const out = [];
+    for (const section of sections) {
+        if (section.text.length <= maxChars * 2) {
+            out.push(section);
+            continue;
+        }
+        const sectionLines = section.text.split('\n');
+        let startLine = section.startLine;
+        let buf = [];
+        let bufChars = 0;
+        for (const line of sectionLines) {
+            if (bufChars === 0 && line.length > maxChars) {
+                let cursor = 0;
+                while (cursor < line.length) {
+                    const piece = line.slice(cursor, cursor + maxChars);
+                    out.push({
+                        text: piece,
+                        startLine,
+                        endLine: startLine,
+                        symbol: section.symbol,
+                    });
+                    cursor += maxChars;
+                }
+                startLine += 1;
+                continue;
+            }
+            if (bufChars > 0 && bufChars + line.length + 1 > maxChars) {
+                out.push({
+                    text: buf.join('\n'),
+                    startLine,
+                    endLine: startLine + buf.length - 1,
+                    symbol: section.symbol,
+                });
+                startLine += buf.length;
+                buf = [];
+                bufChars = 0;
+            }
+            buf.push(line);
+            bufChars += line.length + 1;
+        }
+        if (buf.length > 0) {
+            out.push({
+                text: buf.join('\n'),
+                startLine,
+                endLine: startLine + buf.length - 1,
+                symbol: section.symbol,
+            });
+        }
+    }
+    return out;
+}
+function chunkTextByParagraphs(text, maxChars) {
+    const trimmed = text.trim();
+    if (trimmed.length === 0)
+        return [];
+    const paragraphs = trimmed
+        .split(/\n\s*\n+/)
+        .map(p => p.trim())
+        .filter(p => p.length > 0);
+    const out = [];
+    let cursor = 0;
+    let buf = '';
+    let bufStart = 0;
+    for (const p of paragraphs) {
+        if (buf.length === 0) {
+            buf = p;
+            bufStart = cursor;
+        }
+        else if (buf.length + 2 + p.length <= maxChars) {
+            buf += '\n\n' + p;
+        }
+        else {
+            out.push({ text: buf, offset: bufStart });
+            buf = p;
+            bufStart = cursor;
+        }
+        cursor += p.length + 2;
+    }
+    if (buf.length > 0)
+        out.push({ text: buf, offset: bufStart });
+    return out;
+}
+//# sourceMappingURL=baseIngester.js.map

package/dist/services/rag/baseIngester.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"baseIngester.js","sourceRoot":"","sources":["../../../src/services/rag/baseIngester.ts"],"names":[],"mappings":";AAAA,6CAA6C;AAC7C,0CAA0C;AAC1C,+DAA+D;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AAmB/D,gDAEC;AAED,sCAKC;AAED,0CAUC;AAED,gDASC;AAED,oDA2EC;AAED,sDA0BC;AA1JD,mCAAkC;AAClC,2CAA6B;AAgB7B,SAAgB,kBAAkB,CAAC,IAAY;IAC7C,OAAO,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC;AAClD,CAAC;AAED,SAAgB,aAAa,CAAC,KAAyB,EAAE,MAAM,GAAG,EAAE;IAClE,OAAO,IAAA,mBAAU,EAAC,QAAQ,CAAC;SACxB,MAAM,CAAC,KAAK,CAAC,GAAG,CAAC,IAAI,CAAC,EAAE,CAAC,MAAM,CAAC,IAAI,CAAC,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;SACjD,MAAM,CAAC,KAAK,CAAC;SACb,KAAK,CAAC,CAAC,EAAE,MAAM,CAAC,CAAC;AACtB,CAAC;AAED,SAAgB,eAAe,CAAC,YAAoB;IAClD,MAAM,GAAG,GAAG,IAAI,CAAC,OAAO,CAAC,YAAY,CAAC,CAAC;IACvC,IAAI,GAAG,KAAK,KAAK,IAAI,GAAG,KAAK,MAAM;QAAE,OAAO,QAAQ,CAAC;IACrD,IAAI,GAAG,KAAK,OAAO;QAAE,OAAO,MAAM,CAAC;IACnC,IAAI,GAAG,KAAK,IAAI;QAAE,OAAO,GAAG,CAAC;IAC7B,IAAI,GAAG,KAAK,KAAK,IAAI,GAAG,KAAK,MAAM,IAAI,GAAG,KAAK,MAAM,IAAI,GAAG,KAAK,IAAI,IAAI,GAAG,KAAK,MAAM;QAAE,OAAO,KAAK,CAAC;IACtG,IAAI,GAAG,KAAK,KAAK;QAAE,OAAO,MAAM,CAAC;IACjC,IAAI,GAAG,KAAK,KAAK;QAAE,OAAO,IAAI,CAAC;IAC/B,IAAI,GAAG,KAAK,KAAK;QAAE,OAAO,IAAI,CAAC;IAC/B,OAAO,SAAS,CAAC;AACnB,CAAC;AAED,SAAgB,kBAAkB,CAAC,IAAY;IAC7C,MAAM,UAAU,GAAG,IAAI,CAAC,KAAK,CAAC,6EAA6E,CAAC,CAAC;IAC7G,IAAI,UAAU,EAAE,CAAC,CAAC,CAAC;QAAE,OAAO,UAAU,CAAC,CAAC,CAAC,CAAC;IAC1C,MAAM,SAAS,GAAG,IAAI,CAAC,KAAK,CAAC,uCAAuC,CAAC,CAAC;IACtE,IAAI,SAAS,EAAE,CAAC,CAAC,CAAC;QAAE,OAAO,SAAS,CAAC,CAAC,CAAC,CAAC;IACxC,MAAM,aAAa,GAAG,IAAI,CAAC,KAAK,CAAC,kIAAkI,CAAC,CAAC;IACrK,IAAI,aAAa,EAAE,CAAC,CAAC,CAAC;QAAE,OAAO,aAAa,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC,GAAG,EAAE,CAAC;IAClE,MAAM,UAAU,GAAG,IAAI,CAAC,KAAK,CAAC,8IAA8I,CAAC,CAAC;IAC9K,OAAO,UAAU,EAAE,CAAC,CAAC,CAAC,CAAC;AACzB,CAAC;AAED,SAAgB,oBAAoB,CAAC,OAAe,EAAE,QAAgB;IACpE,MAAM,UAAU,GAAG,OAAO,CAAC,OAAO,CAAC,QAAQ,EAAE,IAAI,CAAC,CAAC;IACnD,IAAI,UAAU,CAAC,IAAI,EAAE,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,EAAE,CAAC;IAC9C,MAAM,KAAK,GAAG,UAAU,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC;IACrC,MAAM,UAAU,GAAG,IAAI,GAAG,CAAS,CAAC,CAAC,CAAC,CAAC,CAAC;IACxC,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,KAAK,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;QACtC,IAAI,kBAAkB,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC;YAAE,UAAU,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC;IACtD,CAAC;IACD,UAAU,CAAC,GAAG,CAAC,KAAK,CAAC,MAAM,CAAC,CAAC;IAC7B,MAAM,MAAM,GAAG,KAAK,CAAC,IAAI,CAAC,UAAU,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC;IAC5D,MAAM,QAAQ,GAAkB,EAAE,CAAC;IACnC,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,MAAM,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC;QAC3C,MAAM,KAAK,GAAG,MAAM,CAAC,CAAC,CAAC,CAAC;QACxB,MAAM,GAAG,GAAG,MAAM,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC;QAC1B,IAAI,GAAG,IAAI,KAAK;YAAE,SAAS;QAC3B,MAAM,IAAI,GAAG,KAAK,CAAC,KAAK,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,IAAI,EAAE,CAAC;QACvD,IAAI,CAAC,IAAI;YAAE,SAAS;QACpB,QAAQ,CAAC,IAAI,CAAC;YACZ,IAAI;YACJ,SAAS,EAAE,KAAK,GAAG,CAAC;YACpB,OAAO,EAAE,GAAG;YACZ,MAAM,EAAE,kBAAkB,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC;SACzC,CAAC,CAAC;IACL,CAAC;IAED,MAAM,GAAG,GAAkB,EAAE,CAAC;IAC9B,KAAK,MAAM,OAAO,IAAI,QAAQ,EAAE,CAAC;QAC/B,IAAI,OAAO,CAAC,IAAI,CAAC,MAAM,IAAI,QAAQ,GAAG,CAAC,EAAE,CAAC;YACxC,GAAG,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC;YAClB,SAAS;QACX,CAAC;QACD,MAAM,YAAY,GAAG,OAAO,CAAC,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC;QAC9C,IAAI,SAAS,GAAG,OAAO,CAAC,SAAS,CAAC;QAClC,IAAI,GAAG,GAAa,EAAE,CAAC;QACvB,IAAI,QAAQ,GAAG,CAAC,CAAC;QACjB,KAAK,MAAM,IAAI,IAAI,YAAY,EAAE,CAAC;YAChC,IAAI,QAAQ,KAAK,CAAC,IAAI,IAAI,CAAC,MAAM,GAAG,QAAQ,EAAE,CAAC;gBAC7C,IAAI,MAAM,GAAG,CAAC,CAAC;gBACf,OAAO,MAAM,GAAG,IAAI,CAAC,MAAM,EAAE,CAAC;oBAC5B,MAAM,KAAK,GAAG,IAAI,CAAC,KAAK,CAAC,MAAM,EAAE,MAAM,GAAG,QAAQ,CAAC,CAAC;oBACpD,GAAG,CAAC,IAAI,CAAC;wBACP,IAAI,EAAE,KAAK;wBACX,SAAS;wBACT,OAAO,EAAE,SAAS;wBAClB,MAAM,EAAE,OAAO,CAAC,MAAM;qBACvB,CAAC,CAAC;oBACH,MAAM,IAAI,QAAQ,CAAC;gBACrB,CAAC;gBACD,SAAS,IAAI,CAAC,CAAC;gBACf,SAAS;YACX,CAAC;YACD,IAAI,QAAQ,GAAG,CAAC,IAAI,QAAQ,GAAG,IAAI,CAAC,MAAM,GAAG,CAAC,GAAG,QAAQ,EAAE,CAAC;gBAC1D,GAAG,CAAC,IAAI,CAAC;oBACP,IAAI,EAAE,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC;oBACpB,SAAS;oBACT,OAAO,EAAE,SAAS,GAAG,GAAG,CAAC,MAAM,GAAG,CAAC;oBACnC,MAAM,EAAE,OAAO,CAAC,MAAM;iBACvB,CAAC,CAAC;gBACH,SAAS,IAAI,GAAG,CAAC,MAAM,CAAC;gBACxB,GAAG,GAAG,EAAE,CAAC;gBACT,QAAQ,GAAG,CAAC,CAAC;YACf,CAAC;YACD,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;YACf,QAAQ,IAAI,IAAI,CAAC,MAAM,GAAG,CAAC,CAAC;QAC9B,CAAC;QACD,IAAI,GAAG,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;YACnB,GAAG,CAAC,IAAI,CAAC;gBACP,IAAI,EAAE,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC;gBACpB,SAAS;gBACT,OAAO,EAAE,SAAS,GAAG,GAAG,CAAC,MAAM,GAAG,CAAC;gBACnC,MAAM,EAAE,OAAO,CAAC,MAAM;aACvB,CAAC,CAAC;QACL,CAAC;IACH,CAAC;IACD,OAAO,GAAG,CAAC;AACb,CAAC;AAED,SAAgB,qBAAqB,CAAC,IAAY,EAAE,QAAgB;IAClE,MAAM,OAAO,GAAG,IAAI,CAAC,IAAI,EAAE,CAAC;IAC5B,IAAI,OAAO,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,EAAE,CAAC;IACpC,MAAM,UAAU,GAAG,OAAO;SACvB,KAAK,CAAC,UAAU,CAAC;SACjB,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;SAClB,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC;IAC7B,MAAM,GAAG,GAAkB,EAAE,CAAC;IAC9B,IAAI,MAAM,GAAG,CAAC,CAAC;IACf,IAAI,GAAG,GAAG,EAAE,CAAC;IACb,IAAI,QAAQ,GAAG,CAAC,CAAC;IACjB,KAAK,MAAM,CAAC,IAAI,UAAU,EAAE,CAAC;QAC3B,IAAI,GAAG,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;YACrB,GAAG,GAAG,CAAC,CAAC;YACR,QAAQ,GAAG,MAAM,CAAC;QACpB,CAAC;aAAM,IAAI,GAAG,CAAC,MAAM,GAAG,CAAC,GAAG,CAAC,CAAC,MAAM,IAAI,QAAQ,EAAE,CAAC;YACjD,GAAG,IAAI,MAAM,GAAG,CAAC,CAAC;QACpB,CAAC;aAAM,CAAC;YACN,GAAG,CAAC,IAAI,CAAC,EAAC,IAAI,EAAE,GAAG,EAAE,MAAM,EAAE,QAAQ,EAAC,CAAC,CAAC;YACxC,GAAG,GAAG,CAAC,CAAC;YACR,QAAQ,GAAG,MAAM,CAAC;QACpB,CAAC;QACD,MAAM,IAAI,CAAC,CAAC,MAAM,GAAG,CAAC,CAAC;IACzB,CAAC;IACD,IAAI,GAAG,CAAC,MAAM,GAAG,CAAC;QAAE,GAAG,CAAC,IAAI,CAAC,EAAC,IAAI,EAAE,GAAG,EAAE,MAAM,EAAE,QAAQ,EAAC,CAAC,CAAC;IAC5D,OAAO,GAAG,CAAC;AACb,CAAC"}

package/dist/services/rag/kernelSourceIngester.d.ts

@@ -0,0 +1,35 @@
+import type { RagStore } from '../ragStore';
+import type { CodebaseRegistry, CodebaseRef } from '../codebase/codebaseRegistry';
+import { PathSecurityGate, type PathPreviewResult } from '../codebase/pathSecurityGate';
+export interface KernelSourceIngestOptions {
+    maxChunkChars?: number;
+    pathPrefix?: string;
+}
+export interface KernelSourceIngestError {
+    filePath: string;
+    reason: string;
+}
+export interface KernelSourceIngestResult {
+    codebaseId: string;
+    filesProcessed: number;
+    chunksAdded: number;
+    chunksSkipped: number;
+    blockedFileCount: number;
+    redactionHitCount: number;
+    errors: KernelSourceIngestError[];
+}
+declare function spdxLicense(content: string): string | undefined;
+declare function filterPreviewFiles(preview: PathPreviewResult, ref: CodebaseRef, opts: KernelSourceIngestOptions): import("../codebase/pathSecurityGate").PathPreviewFile[];
+export declare class KernelSourceIngester {
+    private readonly store;
+    private readonly registry;
+    private readonly gate;
+    constructor(store: RagStore, registry: CodebaseRegistry, gate?: PathSecurityGate);
+    ingest(codebaseId: string, opts?: KernelSourceIngestOptions): KernelSourceIngestResult;
+}
+export declare const __TEST_ONLY__: {
+    spdxLicense: typeof spdxLicense;
+    filterPreviewFiles: typeof filterPreviewFiles;
+};
+export {};
+//# sourceMappingURL=kernelSourceIngester.d.ts.map

package/dist/services/rag/kernelSourceIngester.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"kernelSourceIngester.d.ts","sourceRoot":"","sources":["../../../src/services/rag/kernelSourceIngester.ts"],"names":[],"mappings":"AAOA,OAAO,KAAK,EAAC,QAAQ,EAAC,MAAM,aAAa,CAAC;AAC1C,OAAO,KAAK,EAAC,gBAAgB,EAAE,WAAW,EAAC,MAAM,8BAA8B,CAAC;AAChF,OAAO,EAAC,gBAAgB,EAAE,KAAK,iBAAiB,EAAC,MAAM,8BAA8B,CAAC;AAYtF,MAAM,WAAW,yBAAyB;IACxC,aAAa,CAAC,EAAE,MAAM,CAAC;IACvB,UAAU,CAAC,EAAE,MAAM,CAAC;CACrB;AAED,MAAM,WAAW,uBAAuB;IACtC,QAAQ,EAAE,MAAM,CAAC;IACjB,MAAM,EAAE,MAAM,CAAC;CAChB;AAED,MAAM,WAAW,wBAAwB;IACvC,UAAU,EAAE,MAAM,CAAC;IACnB,cAAc,EAAE,MAAM,CAAC;IACvB,WAAW,EAAE,MAAM,CAAC;IACpB,aAAa,EAAE,MAAM,CAAC;IACtB,gBAAgB,EAAE,MAAM,CAAC;IACzB,iBAAiB,EAAE,MAAM,CAAC;IAC1B,MAAM,EAAE,uBAAuB,EAAE,CAAC;CACnC;AAED,iBAAS,WAAW,CAAC,OAAO,EAAE,MAAM,GAAG,MAAM,GAAG,SAAS,CAIxD;AAED,iBAAS,kBAAkB,CAAC,OAAO,EAAE,iBAAiB,EAAE,GAAG,EAAE,WAAW,EAAE,IAAI,EAAE,yBAAyB,4DASxG;AAED,qBAAa,oBAAoB;IAE7B,OAAO,CAAC,QAAQ,CAAC,KAAK;IACtB,OAAO,CAAC,QAAQ,CAAC,QAAQ;IACzB,OAAO,CAAC,QAAQ,CAAC,IAAI;gBAFJ,KAAK,EAAE,QAAQ,EACf,QAAQ,EAAE,gBAAgB,EAC1B,IAAI,GAAE,gBAAyC;IAGlE,MAAM,CAAC,UAAU,EAAE,MAAM,EAAE,IAAI,GAAE,yBAA8B,GAAG,wBAAwB;CAoG3F;AAED,eAAO,MAAM,aAAa;;;CAGzB,CAAC"}

package/dist/services/rag/kernelSourceIngester.js

@@ -0,0 +1,170 @@
+"use strict";
+// SPDX-License-Identifier: AGPL-3.0-or-later
+// Copyright (C) 2024-2026 Gracker (Chris)
+// This file is part of SmartPerfetto. See LICENSE for details.
+var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    var desc = Object.getOwnPropertyDescriptor(m, k);
+    if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
+      desc = { enumerable: true, get: function() { return m[k]; } };
+    }
+    Object.defineProperty(o, k2, desc);
+}) : (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    o[k2] = m[k];
+}));
+var __setModuleDefault = (this && this.__setModuleDefault) || (Object.create ? (function(o, v) {
+    Object.defineProperty(o, "default", { enumerable: true, value: v });
+}) : function(o, v) {
+    o["default"] = v;
+});
+var __importStar = (this && this.__importStar) || (function () {
+    var ownKeys = function(o) {
+        ownKeys = Object.getOwnPropertyNames || function (o) {
+            var ar = [];
+            for (var k in o) if (Object.prototype.hasOwnProperty.call(o, k)) ar[ar.length] = k;
+            return ar;
+        };
+        return ownKeys(o);
+    };
+    return function (mod) {
+        if (mod && mod.__esModule) return mod;
+        var result = {};
+        if (mod != null) for (var k = ownKeys(mod), i = 0; i < k.length; i++) if (k[i] !== "default") __createBinding(result, mod, k[i]);
+        __setModuleDefault(result, mod);
+        return result;
+    };
+})();
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.__TEST_ONLY__ = exports.KernelSourceIngester = void 0;
+const fs = __importStar(require("fs"));
+const path = __importStar(require("path"));
+const pathSecurityGate_1 = require("../codebase/pathSecurityGate");
+const secretPatterns_1 = require("../security/secretPatterns");
+const baseIngester_1 = require("./baseIngester");
+const DEFAULT_MAX_CHUNK_CHARS = 1800;
+const UNKNOWN_LICENSE = 'UNKNOWN';
+function spdxLicense(content) {
+    const header = content.split(/\r?\n/).slice(0, 12).join('\n');
+    const match = header.match(/SPDX-License-Identifier:\s*([A-Za-z0-9+_.()-]+(?:\s+(?:WITH|OR|AND)\s+[A-Za-z0-9+_.()-]+)*)/);
+    return match?.[1]?.trim();
+}
+function filterPreviewFiles(preview, ref, opts) {
+    const prefixes = [
+        ...(ref.pathFilters ?? []),
+        ...(opts.pathPrefix ? [opts.pathPrefix] : []),
+    ].filter(Boolean).map(prefix => prefix.replace(/^\//, ''));
+    if (prefixes.length === 0)
+        return [];
+    return preview.acceptedFiles.filter(file => prefixes.some(prefix => file.relativePath.startsWith(prefix)));
+}
+class KernelSourceIngester {
+    constructor(store, registry, gate = new pathSecurityGate_1.PathSecurityGate()) {
+        this.store = store;
+        this.registry = registry;
+        this.gate = gate;
+    }
+    ingest(codebaseId, opts = {}) {
+        const ref = this.registry.get(codebaseId);
+        if (!ref) {
+            throw new Error(`Codebase '${codebaseId}' not found`);
+        }
+        if (ref.kind !== 'kernel_source') {
+            throw new Error(`Codebase '${codebaseId}' is kind=${ref.kind}; kernel ingestion requires kernel_source`);
+        }
+        if (!ref.vendor) {
+            throw new Error(`Kernel codebase '${codebaseId}' requires vendor`);
+        }
+        if ((ref.pathFilters ?? []).length === 0 && !opts.pathPrefix) {
+            throw new Error(`Kernel codebase '${codebaseId}' requires pathFilters or pathPrefix`);
+        }
+        const preview = this.gate.preview(ref.rootRealpath);
+        if (preview.blocked) {
+            this.registry.updateIngestStatus(codebaseId, {
+                lastIngestStatus: 'blocked_by_security',
+                lastIngestAt: Date.now(),
+                lastIngestError: preview.blockedReason,
+                blockedFileCount: preview.skippedFiles.length,
+            });
+            return {
+                codebaseId,
+                filesProcessed: 0,
+                chunksAdded: 0,
+                chunksSkipped: 0,
+                blockedFileCount: preview.skippedFiles.length,
+                redactionHitCount: 0,
+                errors: [{ filePath: ref.displayName, reason: preview.blockedReason ?? 'blocked' }],
+            };
+        }
+        const result = {
+            codebaseId,
+            filesProcessed: 0,
+            chunksAdded: 0,
+            chunksSkipped: 0,
+            blockedFileCount: preview.skippedFiles.length,
+            redactionHitCount: 0,
+            errors: [],
+        };
+        const maxChars = opts.maxChunkChars ?? DEFAULT_MAX_CHUNK_CHARS;
+        for (const file of filterPreviewFiles(preview, ref, opts)) {
+            result.filesProcessed++;
+            try {
+                const absolutePath = path.join(ref.rootRealpath, file.relativePath);
+                const content = fs.readFileSync(absolutePath, 'utf-8');
+                const license = ref.licenseTag ?? spdxLicense(content);
+                const chunks = (0, baseIngester_1.chunkSourceBySymbols)(content, maxChars);
+                if (chunks.length === 0) {
+                    result.chunksSkipped++;
+                    continue;
+                }
+                for (const chunk of chunks) {
+                    const redaction = (0, secretPatterns_1.redactSecrets)(chunk.text);
+                    result.redactionHitCount += redaction.redactedCount;
+                    this.store.addChunk({
+                        chunkId: (0, baseIngester_1.stableChunkId)([codebaseId, ref.indexGeneration, file.relativePath, chunk.startLine]),
+                        kind: 'kernel_source',
+                        uri: `codebase://${codebaseId}/${file.relativePath}`,
+                        title: file.relativePath.split('/').pop(),
+                        snippet: redaction.text,
+                        tokenCount: (0, baseIngester_1.estimateTokenCount)(redaction.text),
+                        license: license ?? UNKNOWN_LICENSE,
+                        indexedAt: Date.now(),
+                        filePath: file.relativePath,
+                        lineRange: { start: chunk.startLine, end: chunk.endLine },
+                        ...(chunk.symbol ? { symbol: chunk.symbol } : {}),
+                        language: (0, baseIngester_1.languageForPath)(file.relativePath),
+                        ...(ref.commitHash ? { commitHash: ref.commitHash } : {}),
+                        vendor: ref.vendor,
+                        ...(ref.buildId ? { buildId: ref.buildId } : {}),
+                        codebaseId,
+                        registryOrigin: 'codebase_registry',
+                        ...(!license ? { unsupportedReason: 'license_unknown' } : {}),
+                    });
+                    result.chunksAdded++;
+                }
+            }
+            catch (error) {
+                result.chunksSkipped++;
+                result.errors.push({
+                    filePath: file.relativePath,
+                    reason: error instanceof Error ? error.message : String(error),
+                });
+            }
+        }
+        this.registry.updateIngestStatus(codebaseId, {
+            lastIngestStatus: result.errors.length > 0 ? 'partial' : 'ok',
+            lastIngestAt: Date.now(),
+            lastIngestError: result.errors[0]?.reason,
+            chunkCount: result.chunksAdded,
+            blockedFileCount: result.blockedFileCount,
+            redactionHitCount: result.redactionHitCount,
+        });
+        return result;
+    }
+}
+exports.KernelSourceIngester = KernelSourceIngester;
+exports.__TEST_ONLY__ = {
+    spdxLicense,
+    filterPreviewFiles,
+};
+//# sourceMappingURL=kernelSourceIngester.js.map

package/dist/services/rag/kernelSourceIngester.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"kernelSourceIngester.js","sourceRoot":"","sources":["../../../src/services/rag/kernelSourceIngester.ts"],"names":[],"mappings":";AAAA,6CAA6C;AAC7C,0CAA0C;AAC1C,+DAA+D;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AAE/D,uCAAyB;AACzB,2CAA6B;AAI7B,mEAAsF;AACtF,+DAAyD;AACzD,iDAKwB;AAExB,MAAM,uBAAuB,GAAG,IAAI,CAAC;AACrC,MAAM,eAAe,GAAG,SAAS,CAAC;AAsBlC,SAAS,WAAW,CAAC,OAAe;IAClC,MAAM,MAAM,GAAG,OAAO,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;IAC9D,MAAM,KAAK,GAAG,MAAM,CAAC,KAAK,CAAC,6FAA6F,CAAC,CAAC;IAC1H,OAAO,KAAK,EAAE,CAAC,CAAC,CAAC,EAAE,IAAI,EAAE,CAAC;AAC5B,CAAC;AAED,SAAS,kBAAkB,CAAC,OAA0B,EAAE,GAAgB,EAAE,IAA+B;IACvG,MAAM,QAAQ,GAAG;QACf,GAAG,CAAC,GAAG,CAAC,WAAW,IAAI,EAAE,CAAC;QAC1B,GAAG,CAAC,IAAI,CAAC,UAAU,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,UAAU,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC;KAC9C,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,GAAG,CAAC,MAAM,CAAC,EAAE,CAAC,MAAM,CAAC,OAAO,CAAC,KAAK,EAAE,EAAE,CAAC,CAAC,CAAC;IAC3D,IAAI,QAAQ,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,EAAE,CAAC;IACrC,OAAO,OAAO,CAAC,aAAa,CAAC,MAAM,CAAC,IAAI,CAAC,EAAE,CACzC,QAAQ,CAAC,IAAI,CAAC,MAAM,CAAC,EAAE,CAAC,IAAI,CAAC,YAAY,CAAC,UAAU,CAAC,MAAM,CAAC,CAAC,CAC9D,CAAC;AACJ,CAAC;AAED,MAAa,oBAAoB;IAC/B,YACmB,KAAe,EACf,QAA0B,EAC1B,OAAyB,IAAI,mCAAgB,EAAE;QAF/C,UAAK,GAAL,KAAK,CAAU;QACf,aAAQ,GAAR,QAAQ,CAAkB;QAC1B,SAAI,GAAJ,IAAI,CAA2C;IAC/D,CAAC;IAEJ,MAAM,CAAC,UAAkB,EAAE,OAAkC,EAAE;QAC7D,MAAM,GAAG,GAAG,IAAI,CAAC,QAAQ,CAAC,GAAG,CAAC,UAAU,CAAC,CAAC;QAC1C,IAAI,CAAC,GAAG,EAAE,CAAC;YACT,MAAM,IAAI,KAAK,CAAC,aAAa,UAAU,aAAa,CAAC,CAAC;QACxD,CAAC;QACD,IAAI,GAAG,CAAC,IAAI,KAAK,eAAe,EAAE,CAAC;YACjC,MAAM,IAAI,KAAK,CAAC,aAAa,UAAU,aAAa,GAAG,CAAC,IAAI,2CAA2C,CAAC,CAAC;QAC3G,CAAC;QACD,IAAI,CAAC,GAAG,CAAC,MAAM,EAAE,CAAC;YAChB,MAAM,IAAI,KAAK,CAAC,oBAAoB,UAAU,mBAAmB,CAAC,CAAC;QACrE,CAAC;QACD,IAAI,CAAC,GAAG,CAAC,WAAW,IAAI,EAAE,CAAC,CAAC,MAAM,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,UAAU,EAAE,CAAC;YAC7D,MAAM,IAAI,KAAK,CAAC,oBAAoB,UAAU,sCAAsC,CAAC,CAAC;QACxF,CAAC;QAED,MAAM,OAAO,GAAG,IAAI,CAAC,IAAI,CAAC,OAAO,CAAC,GAAG,CAAC,YAAY,CAAC,CAAC;QACpD,IAAI,OAAO,CAAC,OAAO,EAAE,CAAC;YACpB,IAAI,CAAC,QAAQ,CAAC,kBAAkB,CAAC,UAAU,EAAE;gBAC3C,gBAAgB,EAAE,qBAAqB;gBACvC,YAAY,EAAE,IAAI,CAAC,GAAG,EAAE;gBACxB,eAAe,EAAE,OAAO,CAAC,aAAa;gBACtC,gBAAgB,EAAE,OAAO,CAAC,YAAY,CAAC,MAAM;aAC9C,CAAC,CAAC;YACH,OAAO;gBACL,UAAU;gBACV,cAAc,EAAE,CAAC;gBACjB,WAAW,EAAE,CAAC;gBACd,aAAa,EAAE,CAAC;gBAChB,gBAAgB,EAAE,OAAO,CAAC,YAAY,CAAC,MAAM;gBAC7C,iBAAiB,EAAE,CAAC;gBACpB,MAAM,EAAE,CAAC,EAAC,QAAQ,EAAE,GAAG,CAAC,WAAW,EAAE,MAAM,EAAE,OAAO,CAAC,aAAa,IAAI,SAAS,EAAC,CAAC;aAClF,CAAC;QACJ,CAAC;QAED,MAAM,MAAM,GAA6B;YACvC,UAAU;YACV,cAAc,EAAE,CAAC;YACjB,WAAW,EAAE,CAAC;YACd,aAAa,EAAE,CAAC;YAChB,gBAAgB,EAAE,OAAO,CAAC,YAAY,CAAC,MAAM;YAC7C,iBAAiB,EAAE,CAAC;YACpB,MAAM,EAAE,EAAE;SACX,CAAC;QACF,MAAM,QAAQ,GAAG,IAAI,CAAC,aAAa,IAAI,uBAAuB,CAAC;QAE/D,KAAK,MAAM,IAAI,IAAI,kBAAkB,CAAC,OAAO,EAAE,GAAG,EAAE,IAAI,CAAC,EAAE,CAAC;YAC1D,MAAM,CAAC,cAAc,EAAE,CAAC;YACxB,IAAI,CAAC;gBACH,MAAM,YAAY,GAAG,IAAI,CAAC,IAAI,CAAC,GAAG,CAAC,YAAY,EAAE,IAAI,CAAC,YAAY,CAAC,CAAC;gBACpE,MAAM,OAAO,GAAG,EAAE,CAAC,YAAY,CAAC,YAAY,EAAE,OAAO,CAAC,CAAC;gBACvD,MAAM,OAAO,GAAG,GAAG,CAAC,UAAU,IAAI,WAAW,CAAC,OAAO,CAAC,CAAC;gBACvD,MAAM,MAAM,GAAG,IAAA,mCAAoB,EAAC,OAAO,EAAE,QAAQ,CAAC,CAAC;gBACvD,IAAI,MAAM,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;oBACxB,MAAM,CAAC,aAAa,EAAE,CAAC;oBACvB,SAAS;gBACX,CAAC;gBACD,KAAK,MAAM,KAAK,IAAI,MAAM,EAAE,CAAC;oBAC3B,MAAM,SAAS,GAAG,IAAA,8BAAa,EAAC,KAAK,CAAC,IAAI,CAAC,CAAC;oBAC5C,MAAM,CAAC,iBAAiB,IAAI,SAAS,CAAC,aAAa,CAAC;oBACpD,IAAI,CAAC,KAAK,CAAC,QAAQ,CAAC;wBAClB,OAAO,EAAE,IAAA,4BAAa,EAAC,CAAC,UAAU,EAAE,GAAG,CAAC,eAAe,EAAE,IAAI,CAAC,YAAY,EAAE,KAAK,CAAC,SAAS,CAAC,CAAC;wBAC7F,IAAI,EAAE,eAAe;wBACrB,GAAG,EAAE,cAAc,UAAU,IAAI,IAAI,CAAC,YAAY,EAAE;wBACpD,KAAK,EAAE,IAAI,CAAC,YAAY,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,GAAG,EAAE;wBACzC,OAAO,EAAE,SAAS,CAAC,IAAI;wBACvB,UAAU,EAAE,IAAA,iCAAkB,EAAC,SAAS,CAAC,IAAI,CAAC;wBAC9C,OAAO,EAAE,OAAO,IAAI,eAAe;wBACnC,SAAS,EAAE,IAAI,CAAC,GAAG,EAAE;wBACrB,QAAQ,EAAE,IAAI,CAAC,YAAY;wBAC3B,SAAS,EAAE,EAAC,KAAK,EAAE,KAAK,CAAC,SAAS,EAAE,GAAG,EAAE,KAAK,CAAC,OAAO,EAAC;wBACvD,GAAG,CAAC,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,EAAC,MAAM,EAAE,KAAK,CAAC,MAAM,EAAC,CAAC,CAAC,CAAC,EAAE,CAAC;wBAC/C,QAAQ,EAAE,IAAA,8BAAe,EAAC,IAAI,CAAC,YAAY,CAAC;wBAC5C,GAAG,CAAC,GAAG,CAAC,UAAU,CAAC,CAAC,CAAC,EAAC,UAAU,EAAE,GAAG,CAAC,UAAU,EAAC,CAAC,CAAC,CAAC,EAAE,CAAC;wBACvD,MAAM,EAAE,GAAG,CAAC,MAAM;wBAClB,GAAG,CAAC,GAAG,CAAC,OAAO,CAAC,CAAC,CAAC,EAAC,OAAO,EAAE,GAAG,CAAC,OAAO,EAAC,CAAC,CAAC,CAAC,EAAE,CAAC;wBAC9C,UAAU;wBACV,cAAc,EAAE,mBAAmB;wBACnC,GAAG,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,EAAC,iBAAiB,EAAE,iBAAiB,EAAC,CAAC,CAAC,CAAC,EAAE,CAAC;qBAC5D,CAAC,CAAC;oBACH,MAAM,CAAC,WAAW,EAAE,CAAC;gBACvB,CAAC;YACH,CAAC;YAAC,OAAO,KAAK,EAAE,CAAC;gBACf,MAAM,CAAC,aAAa,EAAE,CAAC;gBACvB,MAAM,CAAC,MAAM,CAAC,IAAI,CAAC;oBACjB,QAAQ,EAAE,IAAI,CAAC,YAAY;oBAC3B,MAAM,EAAE,KAAK,YAAY,KAAK,CAAC,CAAC,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,MAAM,CAAC,KAAK,CAAC;iBAC/D,CAAC,CAAC;YACL,CAAC;QACH,CAAC;QAED,IAAI,CAAC,QAAQ,CAAC,kBAAkB,CAAC,UAAU,EAAE;YAC3C,gBAAgB,EAAE,MAAM,CAAC,MAAM,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,IAAI;YAC7D,YAAY,EAAE,IAAI,CAAC,GAAG,EAAE;YACxB,eAAe,EAAE,MAAM,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,MAAM;YACzC,UAAU,EAAE,MAAM,CAAC,WAAW;YAC9B,gBAAgB,EAAE,MAAM,CAAC,gBAAgB;YACzC,iBAAiB,EAAE,MAAM,CAAC,iBAAiB;SAC5C,CAAC,CAAC;QACH,OAAO,MAAM,CAAC;IAChB,CAAC;CACF;AA3GD,oDA2GC;AAEY,QAAA,aAAa,GAAG;IAC3B,WAAW;IACX,kBAAkB;CACnB,CAAC"}