Fixes #4779: Add comprehensive validation and error handling for codebase indexing

src/services/code-index/processors/scanner.ts

@@ -325,10 +325,24 @@ export class DirectoryScanner implements IDirectoryScanner {
 				success = true
 			} catch (error) {
 				lastError = error as Error
-				console.error(`[DirectoryScanner] Error processing batch (attempt ${attempts}):`, error)
+				const errorMessage = error instanceof Error ? error.message : String(error)
+				const batchInfo = {
+					batchSize: batchBlocks.length,
+					attempt: attempts,
+					maxRetries: MAX_BATCH_RETRIES,
+					fileCount: batchFileInfos.length,
+					sampleFiles: batchFileInfos.slice(0, 3).map((info) => info.filePath),
+				}
+
+				console.error(`[DirectoryScanner] Error processing batch (attempt ${attempts}/${MAX_BATCH_RETRIES}):`, {
+					error: errorMessage,
+					batchInfo,
+					originalError: error,
+				})
 
 				if (attempts < MAX_BATCH_RETRIES) {
 					const delay = INITIAL_RETRY_DELAY_MS * Math.pow(2, attempts - 1)
+					console.log(`[DirectoryScanner] Retrying batch in ${delay}ms...`)
 					await new Promise((resolve) => setTimeout(resolve, delay))
 				}
 			}

src/services/code-index/vector-store/__tests__/qdrant-client.spec.ts

@@ -385,10 +385,14 @@ describe("QdrantVectorStore", () => {
 
 	describe("upsertPoints", () => {
 		it("should correctly call qdrantClient.upsert with processed points", async () => {
+			// Create vectors with correct dimensions (1536)
+			const mockVector1 = new Array(1536).fill(0).map((_, i) => i / 1536)
+			const mockVector2 = new Array(1536).fill(0).map((_, i) => (i + 100) / 1536)
+
 			const mockPoints = [
 				{
 					id: "test-id-1",
-					vector: [0.1, 0.2, 0.3],
+					vector: mockVector1,
 					payload: {
 						filePath: "src/components/Button.tsx",
 						content: "export const Button = () => {}",
@@ -398,7 +402,7 @@ describe("QdrantVectorStore", () => {
 				},
 				{
 					id: "test-id-2",
-					vector: [0.4, 0.5, 0.6],
+					vector: mockVector2,
 					payload: {
 						filePath: "src/utils/helpers.ts",
 						content: "export function helper() {}",
@@ -417,7 +421,7 @@ describe("QdrantVectorStore", () => {
 				points: [
 					{
 						id: "test-id-1",
-						vector: [0.1, 0.2, 0.3],
+						vector: mockVector1,
 						payload: {
 							filePath: "src/components/Button.tsx",
 							content: "export const Button = () => {}",
@@ -432,7 +436,7 @@ describe("QdrantVectorStore", () => {
 					},
 					{
 						id: "test-id-2",
-						vector: [0.4, 0.5, 0.6],
+						vector: mockVector2,
 						payload: {
 							filePath: "src/utils/helpers.ts",
 							content: "export function helper() {}",
@@ -451,10 +455,11 @@ describe("QdrantVectorStore", () => {
 		})
 
 		it("should handle points without filePath in payload", async () => {
+			const mockVector = new Array(1536).fill(0).map((_, i) => i / 1536)
 			const mockPoints = [
 				{
 					id: "test-id-1",
-					vector: [0.1, 0.2, 0.3],
+					vector: mockVector,
 					payload: {
 						content: "some content without filePath",
 						startLine: 1,
@@ -471,7 +476,7 @@ describe("QdrantVectorStore", () => {
 				points: [
 					{
 						id: "test-id-1",
-						vector: [0.1, 0.2, 0.3],
+						vector: mockVector,
 						payload: {
 							content: "some content without filePath",
 							startLine: 1,
@@ -488,17 +493,16 @@ describe("QdrantVectorStore", () => {
 
 			await vectorStore.upsertPoints([])
 
-			expect(mockQdrantClientInstance.upsert).toHaveBeenCalledWith(expectedCollectionName, {
-				points: [],
-				wait: true,
-			})
+			// Should not call upsert for empty arrays (early return)
+			expect(mockQdrantClientInstance.upsert).not.toHaveBeenCalled()
 		})
 
 		it("should correctly process pathSegments for nested file paths", async () => {
+			const mockVector = new Array(1536).fill(0).map((_, i) => i / 1536)
 			const mockPoints = [
 				{
 					id: "test-id-1",
-					vector: [0.1, 0.2, 0.3],
+					vector: mockVector,
 					payload: {
 						filePath: "src/components/ui/forms/InputField.tsx",
 						content: "export const InputField = () => {}",
@@ -516,7 +520,7 @@ describe("QdrantVectorStore", () => {
 				points: [
 					{
 						id: "test-id-1",
-						vector: [0.1, 0.2, 0.3],
+						vector: mockVector,
 						payload: {
 							filePath: "src/components/ui/forms/InputField.tsx",
 							content: "export const InputField = () => {}",
@@ -537,10 +541,11 @@ describe("QdrantVectorStore", () => {
 		})
 
 		it("should handle error scenarios when qdrantClient.upsert fails", async () => {
+			const mockVector = new Array(1536).fill(0).map((_, i) => i / 1536)
 			const mockPoints = [
 				{
 					id: "test-id-1",
-					vector: [0.1, 0.2, 0.3],
+					vector: mockVector,
 					payload: {
 						filePath: "src/test.ts",
 						content: "test content",
@@ -554,10 +559,19 @@ describe("QdrantVectorStore", () => {
 			mockQdrantClientInstance.upsert.mockRejectedValue(upsertError)
 			vitest.spyOn(console, "error").mockImplementation(() => {})
 
-			await expect(vectorStore.upsertPoints(mockPoints)).rejects.toThrow(upsertError)
+			await expect(vectorStore.upsertPoints(mockPoints)).rejects.toThrow(
+				"Failed to upsert 1 points to collection ws-a1b2c3d4e5f6g7h8: Upsert failed",
+			)
 
 			expect(mockQdrantClientInstance.upsert).toHaveBeenCalledTimes(1)
-			expect(console.error).toHaveBeenCalledWith("Failed to upsert points:", upsertError)
+			expect(console.error).toHaveBeenCalledWith(
+				"Failed to upsert points:",
+				expect.objectContaining({
+					message: "Upsert failed",
+					pointsCount: 1,
+					collectionName: "ws-a1b2c3d4e5f6g7h8",
+				}),
+			)
 			;(console.error as any).mockRestore()
 		})
 	})

src/services/code-index/vector-store/qdrant-client.ts

@@ -118,6 +118,32 @@ export class QdrantVectorStore implements IVectorStore {
 		}
 	}
 
+	/**
+	 * Validates vector dimensions against the expected collection vector size
+	 * @param vectors Array of vectors to validate
+	 * @returns true if all vectors have correct dimensions
+	 */
+	private validateVectorDimensions(vectors: number[][]): void {
+		for (let i = 0; i < vectors.length; i++) {
+			const vector = vectors[i]
+			if (!Array.isArray(vector)) {
+				throw new Error(`Vector at index ${i} is not an array`)
+			}
+			if (vector.length !== this.vectorSize) {
+				throw new Error(
+					`Vector at index ${i} has incorrect dimensions: expected ${this.vectorSize}, got ${vector.length}`,
+				)
+			}
+			// Check for invalid values (NaN, Infinity)
+			for (let j = 0; j < vector.length; j++) {
+				const value = vector[j]
+				if (typeof value !== "number" || !isFinite(value)) {
+					throw new Error(`Vector at index ${i} contains invalid value at position ${j}: ${value}`)
+				}
+			}
+		}
+	}
+
 	/**
 	 * Upserts points into the vector store
 	 * @param points Array of points to upsert
@@ -130,6 +156,34 @@ export class QdrantVectorStore implements IVectorStore {
 		}>,
 	): Promise<void> {
 		try {
+			// Validate input points
+			if (!Array.isArray(points)) {
+				throw new Error("Points must be an array")
+			}
+
+			if (points.length === 0) {
+				return // Nothing to upsert
+			}
+
+			// Validate each point structure and collect vectors for dimension validation
+			const vectors: number[][] = []
+			for (let i = 0; i < points.length; i++) {
+				const point = points[i]
+				if (!point.id || typeof point.id !== "string") {
+					throw new Error(`Point at index ${i} must have a valid string id`)
+				}
+				if (!Array.isArray(point.vector) || point.vector.length === 0) {
+					throw new Error(`Point at index ${i} must have a valid vector array`)
+				}
+				if (!point.payload || typeof point.payload !== "object") {
+					throw new Error(`Point at index ${i} must have a valid payload object`)
+				}
+				vectors.push(point.vector)
+			}
+
+			// Validate vector dimensions
+			this.validateVectorDimensions(vectors)
+
 			const processedPoints = points.map((point) => {
 				if (point.payload?.filePath) {
 					const segments = point.payload.filePath.split(path.sep).filter(Boolean)
@@ -141,23 +195,50 @@ export class QdrantVectorStore implements IVectorStore {
 						{},
 					)
 					return {
-						...point,
+						id: point.id,
+						vector: point.vector,
 						payload: {
 							...point.payload,
 							pathSegments,
 						},
 					}
 				}
-				return point
+				return {
+					id: point.id,
+					vector: point.vector,
+					payload: point.payload,
+				}
 			})
 
-			await this.client.upsert(this.collectionName, {
+			// Use the batch upsert operation with proper error handling
+			const upsertRequest = {
 				points: processedPoints,
 				wait: true,
-			})
-		} catch (error) {
-			console.error("Failed to upsert points:", error)
-			throw error
+			}
+
+			await this.client.upsert(this.collectionName, upsertRequest)
+		} catch (error: any) {
+			// Enhanced error logging to help debug the "Bad Request" issue
+			const errorMessage = error?.message || error?.toString() || "Unknown error"
+			const errorDetails = {
+				message: errorMessage,
+				status: error?.status || error?.response?.status,
+				statusText: error?.statusText || error?.response?.statusText,
+				data: error?.data || error?.response?.data,
+				pointsCount: points.length,
+				collectionName: this.collectionName,
+				vectorSize: this.vectorSize,
+				sampleVectorLengths: points.slice(0, 3).map((p) => p.vector?.length || "undefined"),
+			}
+
+			console.error("Failed to upsert points:", errorDetails)
+
+			// Re-throw with more context
+			const enhancedError = new Error(
+				`Failed to upsert ${points.length} points to collection ${this.collectionName}: ${errorMessage}`,
+			)
+			enhancedError.cause = error
+			throw enhancedError
 		}
 	}