@ank1015/providers 0.0.1 → 0.0.3

package/test/providers/google-stream.test.ts

@@ -1,369 +0,0 @@
-import { describe, it, expect } from 'vitest';
-import { Context } from '../../src/types';
-import { MODELS } from '../../src/models.generated';
-import { transformSchemaForGoogle } from '../../src/providers/google';
-import { buildGoogleMessages } from '../../src/providers/convert';
-
-describe('Google Provider Streaming Tests', () => {
-	const model = MODELS.google['gemini-2.5-flash'];
-
-	describe('Schema transformation', () => {
-		it('should transform const to enum', () => {
-
-			const schema = {
-				type: 'string',
-				const: 'active',
-			};
-
-			const result = transformSchemaForGoogle(schema);
-
-			expect(result).toHaveProperty('enum');
-			expect(result.enum).toEqual(['active']);
-			expect(result).not.toHaveProperty('const');
-		});
-
-		it('should transform anyOf with const values to enum', () => {
-
-			const schema = {
-				anyOf: [
-					{ const: 'option1' },
-					{ const: 'option2' },
-					{ const: 'option3' },
-				],
-			};
-
-			const result = transformSchemaForGoogle(schema);
-
-			expect(result).toHaveProperty('enum');
-			expect(result.enum).toEqual(['option1', 'option2', 'option3']);
-			expect(result).not.toHaveProperty('anyOf');
-		});
-
-		it('should recursively transform nested schemas', () => {
-
-			const schema = {
-				type: 'object',
-				properties: {
-					status: { const: 'active' },
-					priority: {
-						anyOf: [
-							{ const: 'high' },
-							{ const: 'low' },
-						],
-					},
-				},
-			};
-
-			const result = transformSchemaForGoogle(schema);
-
-			expect(result.properties.status.enum).toEqual(['active']);
-			expect(result.properties.priority.enum).toEqual(['high', 'low']);
-		});
-
-		it('should not transform anyOf without all const values', () => {
-
-			const schema = {
-				anyOf: [
-					{ const: 'value' },
-					{ type: 'number' },
-				],
-			};
-
-			const result = transformSchemaForGoogle(schema);
-
-			expect(result).toHaveProperty('anyOf');
-			expect(result).not.toHaveProperty('enum');
-		});
-
-		it('should preserve other schema properties', () => {
-
-			const schema = {
-				type: 'string',
-				description: 'A status field',
-				const: 'active',
-			};
-
-			const result = transformSchemaForGoogle(schema);
-
-			expect(result.description).toBe('A status field');
-			expect(result.type).toBe('string');
-		});
-	});
-
-	describe('Message building', () => {
-		it('should convert user text message correctly', () => {
-
-			const context: Context = {
-				messages: [
-					{
-						role: 'user',
-						content: [{ type: 'text', content: 'Hello' }],
-						timestamp: Date.now(),
-					},
-				],
-			};
-
-			const result = buildGoogleMessages(model, context);
-
-			expect(result).toHaveLength(1);
-			expect(result[0].role).toBe('user');
-			expect(result[0].parts).toBeDefined();
-		});
-
-		it('should convert system prompt correctly', () => {
-
-			const context: Context = {
-				systemPrompt: 'You are a helpful assistant.',
-				messages: [],
-			};
-
-			const result = buildGoogleMessages(model, context);
-
-			// Google uses systemInstruction separately, not in messages
-			expect(result).toEqual([]);
-		});
-
-		it('should handle user message with image', () => {
-
-			const context: Context = {
-				messages: [
-					{
-						role: 'user',
-						content: [
-							{ type: 'text', content: 'Describe:' },
-							{ type: 'image', data: 'base64data', mimeType: 'image/png' },
-						],
-						timestamp: Date.now(),
-					},
-				],
-			};
-
-			const result = buildGoogleMessages(model, context);
-
-			expect(result).toHaveLength(1);
-			expect(result[0].parts.length).toBeGreaterThan(1);
-			// Should have both text and inlineData parts
-		});
-
-		it('should handle user message with file', () => {
-
-			const context: Context = {
-				messages: [
-					{
-						role: 'user',
-						content: [
-							{ type: 'text', content: 'Analyze:' },
-							{ type: 'file', data: 'base64pdf', mimeType: 'application/pdf' },
-						],
-						timestamp: Date.now(),
-					},
-				],
-			};
-
-			const result = buildGoogleMessages(model, context);
-
-			expect(result).toHaveLength(1);
-			expect(result[0].parts.length).toBeGreaterThan(1);
-		});
-
-		it('should convert tool result messages', () => {
-
-			const context: Context = {
-				messages: [
-					{
-						role: 'toolResult',
-						toolName: 'calculator',
-						toolCallId: 'call_123',
-						content: [{ type: 'text', content: '42' }],
-						isError: false,
-						timestamp: Date.now(),
-					},
-				],
-			};
-
-			const result = buildGoogleMessages(model, context);
-
-			expect(result).toHaveLength(1);
-			expect(result[0].role).toBe('user');
-		});
-
-		it('should handle native Google assistant messages', () => {
-
-			const nativeMessage = {
-				role: 'assistant' as const,
-				_provider: 'google' as const,
-				message: {
-					candidates: [
-						{
-							content: {
-								parts: [{ text: 'Hello' }],
-								role: 'model',
-							},
-							finishReason: 'STOP',
-						},
-					],
-					usageMetadata: {
-						promptTokenCount: 10,
-						candidatesTokenCount: 5,
-						totalTokenCount: 15,
-					},
-				},
-			};
-
-			const context: Context = {
-				messages: [nativeMessage as any],
-			};
-
-			const result = buildGoogleMessages(model, context);
-
-			expect(result).toHaveLength(1);
-			expect(result[0].role).toBe('model');
-		});
-
-		it('should throw on cross-provider assistant messages', () => {
-
-			const openaiMessage = {
-				role: 'assistant' as const,
-				_provider: 'openai' as const,
-				_native: {},
-			};
-
-			const context: Context = {
-				messages: [openaiMessage as any],
-			};
-
-			expect(() => buildGoogleMessages(model, context)).toThrow();
-		});
-
-		it('should filter out non-supported content types', () => {
-
-			const textOnlyModel = {
-				...model,
-				input: ['text' as const],
-			};
-
-			const context: Context = {
-				messages: [
-					{
-						role: 'user',
-						content: [
-							{ type: 'text', content: 'Text' },
-							{ type: 'image', data: 'base64', mimeType: 'image/png' },
-						],
-						timestamp: Date.now(),
-					},
-				],
-			};
-
-			const result = buildGoogleMessages(textOnlyModel, context);
-
-			expect(result).toHaveLength(1);
-			// Should only have text parts
-			expect(result[0].parts.length).toBe(1);
-			expect(result[0].parts[0]).toHaveProperty('text');
-		});
-	});
-
-	describe('Event streaming', () => {
-		it('should emit start event', () => {
-			// This would require mocking the Google SDK
-			expect(true).toBe(true);
-		});
-
-		it('should emit text delta events', () => {
-			// This would require mocking the Google SDK
-			expect(true).toBe(true);
-		});
-
-		it('should detect thinking mode from part.thought flag', () => {
-			// This would require mocking the Google SDK
-			expect(true).toBe(true);
-		});
-
-		it('should emit done event with complete message', () => {
-			// This would require mocking the Google SDK
-			expect(true).toBe(true);
-		});
-
-		it('should handle function call streaming', () => {
-			// This would require mocking the Google SDK
-			expect(true).toBe(true);
-		});
-
-		it('should auto-generate function call IDs', () => {
-			// This would require mocking the Google SDK
-			expect(true).toBe(true);
-		});
-
-		it('should calculate token usage including thinking tokens', () => {
-			// This would require mocking the Google SDK
-			expect(true).toBe(true);
-		});
-
-		it('should calculate costs correctly', () => {
-			// This would require mocking the Google SDK
-			expect(true).toBe(true);
-		});
-	});
-
-	describe('Options handling', () => {
-		it('should pass through temperature parameter', () => {
-			// Would test that temperature is passed to Google SDK
-			expect(true).toBe(true);
-		});
-
-		it('should pass through maxOutputTokens parameter', () => {
-			// Would test that maxOutputTokens is passed to Google SDK
-			expect(true).toBe(true);
-		});
-
-		it('should handle extended thinking configuration', () => {
-			// Would test extended thinking parameters
-			expect(true).toBe(true);
-		});
-
-		it('should handle response MIME type setting', () => {
-			// Would test response MIME type configuration
-			expect(true).toBe(true);
-		});
-
-		it('should handle abort signal', () => {
-			// Would test abort signal is passed through
-			expect(true).toBe(true);
-		});
-	});
-
-	describe('Error handling', () => {
-		it('should handle invalid API key error', () => {
-			// Would test API key validation
-			expect(true).toBe(true);
-		});
-
-		it('should handle network errors', () => {
-			// Would test network error handling
-			expect(true).toBe(true);
-		});
-
-		it('should handle rate limit errors', () => {
-			// Would test rate limit handling
-			expect(true).toBe(true);
-		});
-
-		it('should handle finish reason mapping', () => {
-			// Would test FinishReason to StopReason mapping
-			expect(true).toBe(true);
-		});
-	});
-
-	describe('Parts accumulation', () => {
-		it('should accumulate consecutive text parts', () => {
-			// Would test that consecutive text parts are merged
-			expect(true).toBe(true);
-		});
-
-		it('should not merge parts with different types', () => {
-			// Would test that different part types are kept separate
-			expect(true).toBe(true);
-		});
-	});
-});

package/test/providers/openai-stream.test.ts

@@ -1,251 +0,0 @@
-import { describe, it, expect, vi, beforeEach } from 'vitest';
-import { streamOpenAI } from '../../src/providers/openai';
-import { Context } from '../../src/types';
-import { MODELS } from '../../src/models.generated';
-import { buildOpenAIMessages } from '../../src/providers/convert';
-import { ResponseInputMessageItem } from 'openai/resources/responses/responses.js';
-
-describe('OpenAI Provider Streaming Tests', () => {
-	const model = MODELS.openai['gpt-5-mini'];
-
-	describe('Message building', () => {
-		it('should convert user text message correctly', () => {
-
-			const context: Context = {
-				messages: [
-					{
-						role: 'user',
-						content: [{ type: 'text', content: 'Hello' }],
-						timestamp: Date.now(),
-					},
-				],
-			};
-
-			const result = buildOpenAIMessages(model, context);
-
-			expect(result).toHaveLength(1);
-			expect((result[0] as ResponseInputMessageItem).role).toBe('user');
-			expect((result[0] as ResponseInputMessageItem).content ).toBeDefined();
-		});
-
-		it('should convert system prompt to developer role', () => {
-
-			const context: Context = {
-				systemPrompt: 'You are a helpful assistant.',
-				messages: [],
-			};
-
-			const result = buildOpenAIMessages(model, context);
-
-			expect(result).toHaveLength(1);
-			expect((result[0] as ResponseInputMessageItem).role).toBe('developer');
-		});
-
-		it('should handle user message with mixed content', () => {
-
-			const context: Context = {
-				messages: [
-					{
-						role: 'user',
-						content: [
-							{ type: 'text', content: 'Describe this:' },
-							{ type: 'image', data: 'base64data', mimeType: 'image/png' },
-						],
-						timestamp: Date.now(),
-					},
-				],
-			};
-
-			const result = buildOpenAIMessages(model, context);
-
-			expect(result).toHaveLength(1);
-			expect(Array.isArray((result[0] as ResponseInputMessageItem).content)).toBe(true);
-			expect(((result[0] as ResponseInputMessageItem).content as any[]).length).toBeGreaterThan(1);
-		});
-
-		it('should filter out non-supported content types', () => {
-
-			const textOnlyModel = {
-				...model,
-				input: ['text' as const],
-			};
-
-			const context: Context = {
-				messages: [
-					{
-						role: 'user',
-						content: [
-							{ type: 'text', content: 'Text' },
-							{ type: 'image', data: 'base64', mimeType: 'image/png' },
-						],
-						timestamp: Date.now(),
-					},
-				],
-			};
-
-			const result = buildOpenAIMessages(textOnlyModel, context);
-
-			expect(result).toHaveLength(1);
-			// Should only have text content
-			if (Array.isArray((result[0] as ResponseInputMessageItem).content)) {
-				expect((result[0] as ResponseInputMessageItem).content.length).toBe(1);
-				expect((result[0] as ResponseInputMessageItem).content[0].type).toBe('input_text');
-			}
-		});
-
-		it('should convert tool result messages', () => {
-
-			const context: Context = {
-				messages: [
-					{
-						role: 'toolResult',
-						toolName: 'calculator',
-						toolCallId: 'call_123',
-						content: [{ type: 'text', content: '42' }],
-						isError: false,
-						timestamp: Date.now(),
-					},
-				],
-			};
-
-			const result = buildOpenAIMessages(model, context);
-
-			expect(result).toHaveLength(1);
-			expect((result[0] as ResponseInputMessageItem).role).toBe(undefined);
-		});
-
-		it('should handle native OpenAI assistant messages', () => {
-
-			const nativeMessage = {
-				role: 'assistant' as const,
-				_provider: 'openai' as const,
-				message: {
-					output: [
-						{
-							type: 'function_call',
-							call_id: 'call_123',
-							name: 'calculator',
-							arguments: '{"expression": "2 + 2"}',
-						},
-					],
-					id: 'resp_123',
-					object: "response",
-					created_at: 1740855869,
-					output_text: '',
-					status: "completed",
-					incomplete_details: null,
-					parallel_tool_calls: false,
-					error: null,
-					instructions: null,
-					max_output_tokens: null,
-					model: "gpt-4o-mini-2024-07-18",
-					user: undefined,
-					metadata: {},
-					previous_response_id: null,
-					temperature: 1,
-					text: {},
-					tool_choice: "auto",
-					tools: [],
-					top_p: 1,
-					truncation: "disabled",
-					usage: { input_tokens: 10, output_tokens: 5, total_tokens: 15, input_tokens_details: {cached_tokens: 0}, output_tokens_details: {reasoning_tokens: 0} },
-				},
-			};
-
-			const context: Context = {
-				messages: [nativeMessage as any],
-			};
-
-			const result = buildOpenAIMessages(model, context);
-
-			expect(result).toHaveLength(1);
-			expect((result[0] as ResponseInputMessageItem).role).toBe(undefined);
-		});
-
-		it('should throw on cross-provider assistant messages', () => {
-
-			const googleMessage = {
-				role: 'assistant' as const,
-				_provider: 'google' as const,
-				_native: {},
-			};
-
-			const context: Context = {
-				messages: [googleMessage as any],
-			};
-
-			expect(() => buildOpenAIMessages(model, context)).toThrow();
-		});
-	});
-
-	describe('Event streaming', () => {
-		it('should emit start event', () => {
-			// This would require mocking the OpenAI SDK
-			expect(true).toBe(true);
-		});
-
-		it('should emit text delta events', () => {
-			// This would require mocking the OpenAI SDK
-			expect(true).toBe(true);
-		});
-
-		it('should emit done event with complete message', () => {
-			// This would require mocking the OpenAI SDK
-			expect(true).toBe(true);
-		});
-
-		it('should handle tool call streaming', () => {
-			// This would require mocking the OpenAI SDK
-			expect(true).toBe(true);
-		});
-
-		it('should calculate token usage correctly', () => {
-			// This would require mocking the OpenAI SDK
-			expect(true).toBe(true);
-		});
-
-		it('should calculate costs correctly', () => {
-			// This would require mocking the OpenAI SDK
-			expect(true).toBe(true);
-		});
-	});
-
-	describe('Options handling', () => {
-		it('should pass through temperature parameter', () => {
-			// Would test that temperature is passed to OpenAI SDK
-			expect(true).toBe(true);
-		});
-
-		it('should pass through maxOutputTokens parameter', () => {
-			// Would test that max_tokens is passed to OpenAI SDK
-			expect(true).toBe(true);
-		});
-
-		it('should handle reasoning mode configuration', () => {
-			// Would test reasoning mode parameters
-			expect(true).toBe(true);
-		});
-
-		it('should handle abort signal', () => {
-			// Would test abort signal is passed through
-			expect(true).toBe(true);
-		});
-	});
-
-	describe('Error handling', () => {
-		it('should handle invalid API key error', () => {
-			// Would test API key validation
-			expect(true).toBe(true);
-		});
-
-		it('should handle network errors', () => {
-			// Would test network error handling
-			expect(true).toBe(true);
-		});
-
-		it('should handle rate limit errors', () => {
-			// Would test rate limit handling
-			expect(true).toBe(true);
-		});
-	});
-});