Spaces:
Running
Running
File size: 8,166 Bytes
98c9143 | 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 | import { afterEach, beforeEach, describe, expect, mock, test } from "bun:test"
import { Hono } from "hono"
import type { ResolvedProviderConfig } from "../src/lib/config"
const actualConfigModule = await import("../src/lib/config")
const actualRateLimitModule = await import("../src/lib/rate-limit")
const actualTokenUsageModule = await import("../src/lib/token-usage")
let providerConfig: ResolvedProviderConfig | null = null
let modelMappings: Record<string, string> = {}
interface TokenCountPayload {
model: string
}
interface TokenCountModel {
capabilities: {
tokenizer: string
}
id: string
}
const getTokenCount = mock(
(_payload: TokenCountPayload, _model: TokenCountModel) =>
Promise.resolve({ input: 40, output: 2 }),
)
const checkRateLimit = mock(() => {
throw new Error("Copilot rate limit should not run for provider aliases")
})
const noopTokenUsageRecorder = () => {}
await mock.module("~/lib/config", () => ({
...actualConfigModule,
getProviderConfig: () => providerConfig,
resolveMappedModel: (model: string) => modelMappings[model] ?? model,
}))
await mock.module("~/lib/rate-limit", () => ({
...actualRateLimitModule,
checkRateLimit,
}))
await mock.module("~/lib/tokenizer", () => ({
getTokenCount,
}))
await mock.module("~/lib/token-usage", () => ({
...actualTokenUsageModule,
createProviderTokenUsageRecorder: () => noopTokenUsageRecorder,
}))
const { messageRoutes } = await import("../src/routes/messages/route")
const { resolveCountTokensModel } = await import(
"../src/routes/messages/count-tokens-handler"
)
const originalFetch = globalThis.fetch
const fetchMock = mock((_url: string | URL | Request, _init?: RequestInit) =>
Promise.resolve(
new Response(
JSON.stringify({
choices: [
{
finish_reason: "stop",
index: 0,
logprobs: null,
message: {
content: "answer text",
role: "assistant",
},
},
],
created: 0,
id: "chatcmpl-test",
model: "qwen-plus",
object: "chat.completion",
usage: {
completion_tokens: 2,
prompt_tokens: 8,
total_tokens: 10,
},
}),
{
headers: {
"content-type": "application/json",
},
},
),
),
)
const createApp = () => {
const app = new Hono()
app.route("/v1/messages", messageRoutes)
return app
}
beforeEach(() => {
providerConfig = {
apiKey: "provider-key",
authType: "authorization",
baseUrl: "https://dashscope.example/compatible-mode",
models: {
"qwen-plus": {
temperature: 0.2,
toolContentSupportType: [],
},
},
name: "dash",
type: "openai-compatible",
}
modelMappings = {}
checkRateLimit.mockClear()
fetchMock.mockClear()
getTokenCount.mockClear()
;(globalThis as unknown as { fetch: typeof fetch }).fetch =
fetchMock as unknown as typeof fetch
})
afterEach(() => {
;(globalThis as unknown as { fetch: typeof fetch }).fetch = originalFetch
providerConfig = null
})
describe("provider/model aliases on top-level messages routes", () => {
test("routes mapped /v1/messages models to the provider before rate limiting", async () => {
modelMappings = {
"claude-opus-4-7": "dash/qwen-plus",
}
const app = createApp()
const response = await app.request("/v1/messages", {
body: JSON.stringify({
max_tokens: 128,
messages: [{ content: "hello", role: "user" }],
model: "claude-opus-4-7",
}),
headers: {
"content-type": "application/json",
},
method: "POST",
})
expect(response.status).toBe(200)
expect(checkRateLimit).not.toHaveBeenCalled()
expect(fetchMock).toHaveBeenCalledTimes(1)
const [url, init] = fetchMock.mock.calls[0]
expect(url).toBe(
"https://dashscope.example/compatible-mode/v1/chat/completions",
)
const upstreamBody = JSON.parse((init as RequestInit).body as string) as {
model: string
}
expect(upstreamBody.model).toBe("qwen-plus")
})
test("routes /v1/messages to the provider and strips the provider prefix", async () => {
const app = createApp()
const response = await app.request("/v1/messages", {
body: JSON.stringify({
max_tokens: 128,
messages: [{ content: "hello", role: "user" }],
model: "dash/qwen-plus",
}),
headers: {
"content-type": "application/json",
},
method: "POST",
})
expect(response.status).toBe(200)
expect(checkRateLimit).not.toHaveBeenCalled()
expect(fetchMock).toHaveBeenCalledTimes(1)
const [url, init] = fetchMock.mock.calls[0]
expect(url).toBe(
"https://dashscope.example/compatible-mode/v1/chat/completions",
)
const upstreamBody = JSON.parse((init as RequestInit).body as string) as {
model: string
}
expect(upstreamBody.model).toBe("qwen-plus")
const json = (await response.json()) as { model: string }
expect(json.model).toBe("qwen-plus")
})
test("routes /v1/messages/count_tokens to provider token counting with the stripped model", async () => {
const app = createApp()
const response = await app.request("/v1/messages/count_tokens", {
body: JSON.stringify({
max_tokens: 128,
messages: [{ content: "hello", role: "user" }],
model: "dash/qwen-plus",
}),
headers: {
"content-type": "application/json",
},
method: "POST",
})
expect(response.status).toBe(200)
expect(await response.json()).toEqual({
input_tokens: 42,
})
expect(getTokenCount).toHaveBeenCalledTimes(1)
const [openAIPayload, selectedModel] = getTokenCount.mock.calls[0] as [
TokenCountPayload,
TokenCountModel,
]
expect(openAIPayload.model).toBe("qwen-plus")
expect(selectedModel.id).toBe("qwen-plus")
expect(selectedModel.capabilities.tokenizer).toBe("o200k_base")
})
test("routes mapped /v1/messages/count_tokens models to provider token counting", async () => {
modelMappings = {
"claude-opus-4-7": "dash/qwen-plus",
}
const app = createApp()
const response = await app.request("/v1/messages/count_tokens", {
body: JSON.stringify({
max_tokens: 128,
messages: [{ content: "hello", role: "user" }],
model: "claude-opus-4-7",
}),
headers: {
"content-type": "application/json",
},
method: "POST",
})
expect(response.status).toBe(200)
expect(await response.json()).toEqual({
input_tokens: 42,
})
expect(getTokenCount).toHaveBeenCalledTimes(1)
const [openAIPayload, selectedModel] = getTokenCount.mock.calls[0] as [
TokenCountPayload,
TokenCountModel,
]
expect(openAIPayload.model).toBe("qwen-plus")
expect(selectedModel.id).toBe("qwen-plus")
expect(selectedModel.capabilities.tokenizer).toBe("o200k_base")
})
test("resolves missing top-level count_tokens models to the o200k_base fallback model", () => {
const resolved = resolveCountTokensModel("missing-model", () => undefined)
expect(resolved.fallback).toBe(true)
expect(resolved.model.id).toBe("missing-model")
expect(resolved.model.capabilities.tokenizer).toBe("o200k_base")
})
test("does not return a fake count when provider token counting fails", async () => {
getTokenCount.mockImplementationOnce(
(_payload: TokenCountPayload, _model: TokenCountModel) =>
Promise.reject(new Error("tokenizer failed")),
)
const app = createApp()
const response = await app.request("/v1/messages/count_tokens", {
body: JSON.stringify({
max_tokens: 128,
messages: [{ content: "hello", role: "user" }],
model: "dash/qwen-plus",
}),
headers: {
"content-type": "application/json",
},
method: "POST",
})
expect(response.status).toBe(500)
expect(await response.json()).toEqual({
error: {
message: "tokenizer failed",
type: "error",
},
})
})
})
|