ThoughtMCP

/** * Tests for OllamaEmbeddingModel * Requirements: 2.1, 2.2, 2.3, 2.4, 2.5 */ import { beforeEach, describe, expect, it, vi } from "vitest"; import { OllamaEmbeddingModel } from "../../../../embeddings/models/ollama-model"; import { EmbeddingError, NetworkError, TimeoutError, ValidationError, } from "../../../../embeddings/types"; const mockFetch = vi.fn(); vi.stubGlobal("fetch", mockFetch); describe("OllamaEmbeddingModel", () => { const defaultConfig = { host: "http://localhost:11434", modelName: "nomic-embed-text", dimension: 768, timeout: 30000, maxRetries: 3, }; beforeEach(() => { mockFetch.mockClear(); }); describe("Constructor", () => { it("should create model with valid configuration", () => { const model = new OllamaEmbeddingModel(defaultConfig); expect(model.getDimension()).toBe(768); expect(model.getModelName()).toBe("nomic-embed-text"); }); }); describe("Successful Generation", () => { it("should generate and normalize embedding", async () => { const rawEmbedding = new Array(768).fill(0).map((_, i) => (i === 0 ? 3 : i === 1 ? 4 : 0)); mockFetch.mockResolvedValueOnce({ ok: true, status: 200, json: async () => ({ embedding: rawEmbedding }), }); const model = new OllamaEmbeddingModel(defaultConfig); const result = await model.generate("test text"); expect(result).toHaveLength(768); expect(result[0]).toBeCloseTo(0.6, 5); expect(result[1]).toBeCloseTo(0.8, 5); }); it("should call API with correct parameters", async () => { mockFetch.mockResolvedValueOnce({ ok: true, status: 200, json: async () => ({ embedding: new Array(768).fill(0.1) }), }); const model = new OllamaEmbeddingModel(defaultConfig); await model.generate("test text"); expect(mockFetch).toHaveBeenCalledWith( "http://localhost:11434/api/embeddings", expect.objectContaining({ method: "POST", headers: { "Content-Type": "application/json" }, }) ); }); }); describe("Input Validation", () => { it("should reject empty string", async () => { const model = new OllamaEmbeddingModel(defaultConfig); await expect(model.generate("")).rejects.toThrow(ValidationError); }); it("should reject null input", async () => { const model = new OllamaEmbeddingModel(defaultConfig); await expect(model.generate(null as any)).rejects.toThrow(ValidationError); }); }); describe("HTTP Errors", () => { it("should handle 404 model not found", async () => { mockFetch.mockResolvedValue({ ok: false, status: 404, statusText: "Not Found", json: async () => ({}), } as Response); const model = new OllamaEmbeddingModel(defaultConfig); const error = await model.generate("test").catch((e) => e); expect(error).toBeInstanceOf(NetworkError); expect(error.message).toContain("Model 'nomic-embed-text' not found"); }); it("should handle 500 error", async () => { mockFetch.mockResolvedValue({ ok: false, status: 500, statusText: "Internal Server Error", json: async () => ({}), } as Response); const model = new OllamaEmbeddingModel(defaultConfig); const error = await model.generate("test").catch((e) => e); expect(error).toBeInstanceOf(NetworkError); expect(error.message).toContain("status 500"); }); it("should handle 503 service unavailable", async () => { mockFetch.mockResolvedValue({ ok: false, status: 503, statusText: "Service Unavailable", json: async () => ({}), } as Response); const model = new OllamaEmbeddingModel(defaultConfig); const error = await model.generate("test").catch((e) => e); expect(error).toBeInstanceOf(NetworkError); expect(error.message).toContain("status 503"); }); }); describe("Invalid Responses", () => { it("should handle missing embedding field", async () => { mockFetch.mockResolvedValue({ ok: true, status: 200, json: async () => ({}), } as Response); const model = new OllamaEmbeddingModel(defaultConfig); const error = await model.generate("test").catch((e) => e); expect(error).toBeInstanceOf(EmbeddingError); expect(error.message).toContain("Invalid response format"); }); it("should handle non-array embedding field", async () => { mockFetch.mockResolvedValue({ ok: true, status: 200, json: async () => ({ embedding: "not an array" }), } as Response); const model = new OllamaEmbeddingModel(defaultConfig); const error = await model.generate("test").catch((e) => e); expect(error).toBeInstanceOf(EmbeddingError); expect(error.message).toContain("Invalid response format"); }); it("should handle dimension mismatch", async () => { mockFetch.mockResolvedValue({ ok: true, status: 200, json: async () => ({ embedding: [0.1, 0.2] }), } as Response); const model = new OllamaEmbeddingModel(defaultConfig); const error = await model.generate("test").catch((e) => e); expect(error).toBeInstanceOf(EmbeddingError); expect(error.message).toContain("dimension mismatch"); }); }); describe("Network Errors", () => { it("should handle connection refused", async () => { mockFetch.mockRejectedValue(new TypeError("fetch failed")); const model = new OllamaEmbeddingModel(defaultConfig); const error = await model.generate("test").catch((e) => e); expect(error).toBeInstanceOf(NetworkError); expect(error.message).toContain("Failed to connect"); }); it("should handle timeout", async () => { mockFetch.mockImplementation( () => new Promise((resolve) => setTimeout( () => resolve({ ok: true, status: 200, json: async () => ({ embedding: new Array(768).fill(0.1) }), } as Response), 200 ) ) ); const model = new OllamaEmbeddingModel({ ...defaultConfig, timeout: 100, }); const error = await model.generate("test").catch((e) => e); expect(error).toBeInstanceOf(TimeoutError); expect(error.message).toContain("timed out"); }); it("should handle unknown errors", async () => { mockFetch.mockRejectedValue(new Error("Unknown error")); const model = new OllamaEmbeddingModel(defaultConfig); const error = await model.generate("test").catch((e) => e); expect(error).toBeInstanceOf(EmbeddingError); expect(error.message).toContain("Unexpected error"); }); }); describe("Retry Logic", () => { it("should retry on transient failures", async () => { let attempts = 0; mockFetch.mockImplementation(async () => { attempts++; if (attempts < 3) { throw new TypeError("Network error"); } return { ok: true, status: 200, json: async () => ({ embedding: new Array(768).fill(0.1) }), }; }); const model = new OllamaEmbeddingModel(defaultConfig); const result = await model.generate("test"); expect(result).toHaveLength(768); expect(attempts).toBe(3); }); it("should fail after max retries", async () => { mockFetch.mockRejectedValue(new TypeError("Persistent error")); const model = new OllamaEmbeddingModel({ ...defaultConfig, maxRetries: 2, }); await expect(model.generate("test")).rejects.toThrow(NetworkError); expect(mockFetch).toHaveBeenCalledTimes(3); }); }); describe("Normalization", () => { it("should normalize to unit vector", async () => { const rawEmbedding = [2, 2, 1]; mockFetch.mockResolvedValueOnce({ ok: true, status: 200, json: async () => ({ embedding: rawEmbedding }), }); const model = new OllamaEmbeddingModel({ ...defaultConfig, dimension: 3 }); const result = await model.generate("test"); const magnitude = Math.sqrt(result.reduce((sum, val) => sum + val * val, 0)); expect(magnitude).toBeCloseTo(1.0, 5); }); it("should handle zero-magnitude embedding", async () => { mockFetch.mockResolvedValueOnce({ ok: true, status: 200, json: async () => ({ embedding: [0, 0, 0] }), }); const model = new OllamaEmbeddingModel({ ...defaultConfig, dimension: 3 }); const result = await model.generate("test"); expect(result).toEqual([0, 0, 0]); }); }); });