Spaces:
Sleeping
Sleeping
File size: 9,619 Bytes
0bcc252 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 |
import request from 'supertest';
import { EventEmitter } from 'events';
import type { Express } from 'express';
const TEST_SECRET = 'test-secret';
let app: Express;
describe('/v1/chat/completions', () => {
jest.setTimeout(120000); // Increase timeout for all tests in this suite
beforeEach(async () => {
// Set up test environment
process.env.NODE_ENV = 'test';
process.env.LLM_PROVIDER = 'openai'; // Use OpenAI provider for tests
process.env.OPENAI_API_KEY = 'test-key';
process.env.JINA_API_KEY = 'test-key';
// Clean up any existing secret
const existingSecretIndex = process.argv.findIndex(arg => arg.startsWith('--secret='));
if (existingSecretIndex !== -1) {
process.argv.splice(existingSecretIndex, 1);
}
// Set up test secret and import server module
process.argv.push(`--secret=${TEST_SECRET}`);
// Import server module (jest.resetModules() is called automatically before each test)
const { default: serverModule } = await require('../app');
app = serverModule;
});
afterEach(async () => {
// Clean up environment variables
delete process.env.OPENAI_API_KEY;
delete process.env.JINA_API_KEY;
// Clean up any remaining event listeners
const emitter = EventEmitter.prototype;
emitter.removeAllListeners();
emitter.setMaxListeners(emitter.getMaxListeners() + 1);
// Clean up test secret
const secretIndex = process.argv.findIndex(arg => arg.startsWith('--secret='));
if (secretIndex !== -1) {
process.argv.splice(secretIndex, 1);
}
// Wait for any pending promises to settle
await new Promise(resolve => setTimeout(resolve, 500));
// Reset module cache to ensure clean state
jest.resetModules();
});
it('should require authentication when secret is set', async () => {
// Note: secret is already set in beforeEach
const response = await request(app)
.post('/v1/chat/completions')
.send({
model: 'test-model',
messages: [{ role: 'user', content: 'test' }]
});
expect(response.status).toBe(401);
});
it('should allow requests without auth when no secret is set', async () => {
// Remove secret for this test
const secretIndex = process.argv.findIndex(arg => arg.startsWith('--secret='));
if (secretIndex !== -1) {
process.argv.splice(secretIndex, 1);
}
// Reset module cache to ensure clean state
jest.resetModules();
// Reload server module without secret
const { default: serverModule } = await require('../app');
app = serverModule;
const response = await request(app)
.post('/v1/chat/completions')
.send({
model: 'test-model',
messages: [{ role: 'user', content: 'test' }]
});
expect(response.status).toBe(200);
});
it('should reject requests without user message', async () => {
const response = await request(app)
.post('/v1/chat/completions')
.set('Authorization', `Bearer ${TEST_SECRET}`)
.send({
model: 'test-model',
messages: [{ role: 'developer', content: 'test' }]
});
expect(response.status).toBe(400);
expect(response.body.error).toBe('Last message must be from user');
});
it('should handle non-streaming request', async () => {
const response = await request(app)
.post('/v1/chat/completions')
.set('Authorization', `Bearer ${TEST_SECRET}`)
.send({
model: 'test-model',
messages: [{ role: 'user', content: 'test' }]
});
expect(response.status).toBe(200);
expect(response.body).toMatchObject({
object: 'chat.completion',
choices: [{
message: {
role: 'assistant'
}
}]
});
});
it('should handle streaming request and track tokens correctly', async () => {
return new Promise<void>((resolve, reject) => {
let isDone = false;
let totalCompletionTokens = 0;
const cleanup = () => {
clearTimeout(timeoutHandle);
isDone = true;
resolve();
};
const timeoutHandle = setTimeout(() => {
if (!isDone) {
cleanup();
reject(new Error('Test timed out'));
}
}, 30000);
request(app)
.post('/v1/chat/completions')
.set('Authorization', `Bearer ${TEST_SECRET}`)
.send({
model: 'test-model',
messages: [{ role: 'user', content: 'test' }],
stream: true
})
.buffer(true)
.parse((res, callback) => {
const response = res as unknown as {
on(event: 'data', listener: (chunk: Buffer) => void): void;
on(event: 'end', listener: () => void): void;
on(event: 'error', listener: (err: Error) => void): void;
};
let responseData = '';
response.on('error', (err) => {
cleanup();
callback(err, null);
});
response.on('data', (chunk) => {
responseData += chunk.toString();
});
response.on('end', () => {
try {
callback(null, responseData);
} catch (err) {
cleanup();
callback(err instanceof Error ? err : new Error(String(err)), null);
}
});
})
.end((err, res) => {
if (err) return reject(err);
expect(res.status).toBe(200);
expect(res.headers['content-type']).toBe('text/event-stream');
// Verify stream format and content
if (isDone) return; // Prevent multiple resolves
const responseText = res.body as string;
const chunks = responseText
.split('\n\n')
.filter((line: string) => line.startsWith('data: '))
.map((line: string) => JSON.parse(line.replace('data: ', '')));
// Process all chunks
expect(chunks.length).toBeGreaterThan(0);
// Verify initial chunk format
expect(chunks[0]).toMatchObject({
id: expect.any(String),
object: 'chat.completion.chunk',
choices: [{
index: 0,
delta: { role: 'assistant' },
logprobs: null,
finish_reason: null
}]
});
// Verify content chunks have content
chunks.slice(1).forEach(chunk => {
const content = chunk.choices[0].delta.content;
if (content && content.trim()) {
totalCompletionTokens += 1; // Count 1 token per chunk as per Vercel convention
}
expect(chunk).toMatchObject({
object: 'chat.completion.chunk',
choices: [{
delta: expect.objectContaining({
content: expect.any(String)
})
}]
});
});
// Verify final chunk format if present
const lastChunk = chunks[chunks.length - 1];
if (lastChunk?.choices?.[0]?.finish_reason === 'stop') {
expect(lastChunk).toMatchObject({
object: 'chat.completion.chunk',
choices: [{
delta: {},
finish_reason: 'stop'
}]
});
}
// Verify we tracked some completion tokens
expect(totalCompletionTokens).toBeGreaterThan(0);
// Clean up and resolve
if (!isDone) {
cleanup();
}
});
});
});
it('should track tokens correctly in error response', async () => {
const response = await request(app)
.post('/v1/chat/completions')
.set('Authorization', `Bearer ${TEST_SECRET}`)
.send({
model: 'test-model',
messages: [] // Invalid messages array
});
expect(response.status).toBe(400);
expect(response.body).toHaveProperty('error');
expect(response.body.error).toBe('Messages array is required and must not be empty');
// Make another request to verify token tracking after error
const validResponse = await request(app)
.post('/v1/chat/completions')
.set('Authorization', `Bearer ${TEST_SECRET}`)
.send({
model: 'test-model',
messages: [{ role: 'user', content: 'test' }]
});
// Verify token tracking still works after error
expect(validResponse.body.usage).toMatchObject({
prompt_tokens: expect.any(Number),
completion_tokens: expect.any(Number),
total_tokens: expect.any(Number)
});
// Basic token tracking structure should be present
expect(validResponse.body.usage.total_tokens).toBe(
validResponse.body.usage.prompt_tokens + validResponse.body.usage.completion_tokens
);
});
it('should provide token usage in Vercel AI SDK format', async () => {
const response = await request(app)
.post('/v1/chat/completions')
.set('Authorization', `Bearer ${TEST_SECRET}`)
.send({
model: 'test-model',
messages: [{ role: 'user', content: 'test' }]
});
expect(response.status).toBe(200);
const usage = response.body.usage;
expect(usage).toMatchObject({
prompt_tokens: expect.any(Number),
completion_tokens: expect.any(Number),
total_tokens: expect.any(Number)
});
// Basic token tracking structure should be present
expect(usage.total_tokens).toBe(
usage.prompt_tokens + usage.completion_tokens
);
});
});
|