diff --git a/README.md b/README.md index 56e7f9435..c8b158956 100644 --- a/README.md +++ b/README.md @@ -30,6 +30,8 @@ One-Click to get a well-designed cross-platform ChatGPT web UI, with GPT3, GPT4 [Deploy on Zeabur](https://vercel.com/new/clone?repository-url=https%3A%2F%2Fgithub.com%2FChatGPTNextWeb%2FChatGPT-Next-Web&env=OPENAI_API_KEY&env=CODE&project-name=nextchat&repository-name=NextChat) [Deploy on Zeabur](https://zeabur.com/templates/ZBUEFA) [Open in Gitpod](https://gitpod.io/#https://github.com/Yidadaa/ChatGPT-Next-Web) +[](https://monica.im/?utm=nxcrp) + ## Enterprise Edition @@ -89,12 +91,13 @@ For enterprise inquiries, please contact: **business@nextchat.dev** - [x] Desktop App with tauri - [x] Self-host Model: Fully compatible with [RWKV-Runner](https://github.com/josStorer/RWKV-Runner), as well as server deployment of [LocalAI](https://github.com/go-skynet/LocalAI): llama/gpt4all/rwkv/vicuna/koala/gpt4all-j/cerebras/falcon/dolly etc. - [x] Artifacts: Easily preview, copy and share generated content/webpages through a separate window [#5092](https://github.com/ChatGPTNextWeb/ChatGPT-Next-Web/pull/5092) -- [x] Plugins: support artifacts, network search, calculator, any other apis etc. [#165](https://github.com/Yidadaa/ChatGPT-Next-Web/issues/165) - - [x] artifacts - - [ ] network search, network search, calculator, any other apis etc. [#165](https://github.com/Yidadaa/ChatGPT-Next-Web/issues/165) +- [x] Plugins: support network search, calculator, any other apis etc. [#165](https://github.com/Yidadaa/ChatGPT-Next-Web/issues/165) [#5353](https://github.com/ChatGPTNextWeb/ChatGPT-Next-Web/issues/5353) + - [x] network search, calculator, any other apis etc. [#165](https://github.com/Yidadaa/ChatGPT-Next-Web/issues/165) [#5353](https://github.com/ChatGPTNextWeb/ChatGPT-Next-Web/issues/5353) +- [ ] local knowledge base ## What's New +- 🚀 v2.15.0 Now supports Plugins! Read this: [NextChat-Awesome-Plugins](https://github.com/ChatGPTNextWeb/NextChat-Awesome-Plugins) - 🚀 v2.14.0 Now supports Artifacts & SD - 🚀 v2.10.1 support Google Gemini Pro model. - 🚀 v2.9.11 you can use azure endpoint now. @@ -125,12 +128,13 @@ For enterprise inquiries, please contact: **business@nextchat.dev** - [x] 使用 tauri 打包桌面应用 - [x] 支持自部署的大语言模型：开箱即用 [RWKV-Runner](https://github.com/josStorer/RWKV-Runner) ，服务端部署 [LocalAI 项目](https://github.com/go-skynet/LocalAI) llama / gpt4all / rwkv / vicuna / koala / gpt4all-j / cerebras / falcon / dolly 等等，或者使用 [api-for-open-llm](https://github.com/xusenlinzy/api-for-open-llm) - [x] Artifacts: 通过独立窗口，轻松预览、复制和分享生成的内容/可交互网页 [#5092](https://github.com/ChatGPTNextWeb/ChatGPT-Next-Web/pull/5092) -- [x] 插件机制，支持 artifacts，联网搜索、计算器、调用其他平台 api [#165](https://github.com/Yidadaa/ChatGPT-Next-Web/issues/165) - - [x] artifacts - - [ ] 支持联网搜索、计算器、调用其他平台 api [#165](https://github.com/Yidadaa/ChatGPT-Next-Web/issues/165) +- [x] 插件机制，支持`联网搜索`、`计算器`、调用其他平台 api [#165](https://github.com/Yidadaa/ChatGPT-Next-Web/issues/165) [#5353](https://github.com/ChatGPTNextWeb/ChatGPT-Next-Web/issues/5353) + - [x] 支持联网搜索、计算器、调用其他平台 api [#165](https://github.com/Yidadaa/ChatGPT-Next-Web/issues/165) [#5353](https://github.com/ChatGPTNextWeb/ChatGPT-Next-Web/issues/5353) + - [ ] 本地知识库 ## 最新动态 +- 🚀 v2.15.0 现在支持插件功能了！了解更多：[NextChat-Awesome-Plugins](https://github.com/ChatGPTNextWeb/NextChat-Awesome-Plugins) - 🚀 v2.14.0 现在支持 Artifacts & SD 了。 - 🚀 v2.10.1 现在支持 Gemini Pro 模型。 - 🚀 v2.9.11 现在可以使用自定义 Azure 服务了。 @@ -280,6 +284,18 @@ Alibaba Cloud Api Key. Alibaba Cloud Api Url. +### `IFLYTEK_URL` (Optional) + +iflytek Api Url. + +### `IFLYTEK_API_KEY` (Optional) + +iflytek Api Key. + +### `IFLYTEK_API_SECRET` (Optional) + +iflytek Api Secret. + ### `HIDE_USER_API_KEY` (optional) > Default: Empty diff --git a/README_CN.md b/README_CN.md index 8c464dc09..beed396c5 100644 --- a/README_CN.md +++ b/README_CN.md @@ -172,6 +172,20 @@ ByteDance Api Url. 阿里云（千问）Api Url. +### `IFLYTEK_URL` (可选) + +讯飞星火Api Url. + +### `IFLYTEK_API_KEY` (可选) + +讯飞星火Api Key. + +### `IFLYTEK_API_SECRET` (可选) + +讯飞星火Api Secret. + + + ### `HIDE_USER_API_KEY` （可选） 如果你不想让用户自行填入 API Key，将此环境变量设置为 1 即可。 diff --git a/app/api/[provider]/[...path]/route.ts b/app/api/[provider]/[...path]/route.ts new file mode 100644 index 000000000..24aa5ec04 --- /dev/null +++ b/app/api/[provider]/[...path]/route.ts @@ -0,0 +1,70 @@ +import { ApiPath } from "@/app/constant"; +import { NextRequest, NextResponse } from "next/server"; +import { handle as openaiHandler } from "../../openai"; +import { handle as azureHandler } from "../../azure"; +import { handle as googleHandler } from "../../google"; +import { handle as anthropicHandler } from "../../anthropic"; +import { handle as baiduHandler } from "../../baidu"; +import { handle as bytedanceHandler } from "../../bytedance"; +import { handle as alibabaHandler } from "../../alibaba"; +import { handle as moonshotHandler } from "../../moonshot"; +import { handle as stabilityHandler } from "../../stability"; +import { handle as iflytekHandler } from "../../iflytek"; +import { handle as proxyHandler } from "../../proxy"; + +async function handle( + req: NextRequest, + { params }: { params: { provider: string; path: string[] } }, +) { + const apiPath = `/api/${params.provider}`; + console.log(`[${params.provider} Route] params `, params); + switch (apiPath) { + case ApiPath.Azure: + return azureHandler(req, { params }); + case ApiPath.Google: + return googleHandler(req, { params }); + case ApiPath.Anthropic: + return anthropicHandler(req, { params }); + case ApiPath.Baidu: + return baiduHandler(req, { params }); + case ApiPath.ByteDance: + return bytedanceHandler(req, { params }); + case ApiPath.Alibaba: + return alibabaHandler(req, { params }); + // case ApiPath.Tencent: using "/api/tencent" + case ApiPath.Moonshot: + return moonshotHandler(req, { params }); + case ApiPath.Stability: + return stabilityHandler(req, { params }); + case ApiPath.Iflytek: + return iflytekHandler(req, { params }); + case ApiPath.OpenAI: + return openaiHandler(req, { params }); + default: + return proxyHandler(req, { params }); + } +} + +export const GET = handle; +export const POST = handle; + +export const runtime = "edge"; +export const preferredRegion = [ + "arn1", + "bom1", + "cdg1", + "cle1", + "cpt1", + "dub1", + "fra1", + "gru1", + "hnd1", + "iad1", + "icn1", + "kix1", + "lhr1", + "pdx1", + "sfo1", + "sin1", + "syd1", +]; diff --git a/app/api/alibaba/[...path]/route.ts b/app/api/alibaba.ts similarity index 91% rename from app/api/alibaba/[...path]/route.ts rename to app/api/alibaba.ts index c97ce5934..675d9f301 100644 --- a/app/api/alibaba/[...path]/route.ts +++ b/app/api/alibaba.ts @@ -14,7 +14,7 @@ import type { RequestPayload } from "@/app/client/platforms/openai"; const serverConfig = getServerSideConfig(); -async function handle( +export async function handle( req: NextRequest, { params }: { params: { path: string[] } }, ) { @@ -40,30 +40,6 @@ async function handle( } } -export const GET = handle; -export const POST = handle; - -export const runtime = "edge"; -export const preferredRegion = [ - "arn1", - "bom1", - "cdg1", - "cle1", - "cpt1", - "dub1", - "fra1", - "gru1", - "hnd1", - "iad1", - "icn1", - "kix1", - "lhr1", - "pdx1", - "sfo1", - "sin1", - "syd1", -]; - async function request(req: NextRequest) { const controller = new AbortController(); diff --git a/app/api/anthropic/[...path]/route.ts b/app/api/anthropic.ts similarity index 92% rename from app/api/anthropic/[...path]/route.ts rename to app/api/anthropic.ts index 20f8d52e0..3d49f4c88 100644 --- a/app/api/anthropic/[...path]/route.ts +++ b/app/api/anthropic.ts @@ -9,13 +9,13 @@ import { } from "@/app/constant"; import { prettyObject } from "@/app/utils/format"; import { NextRequest, NextResponse } from "next/server"; -import { auth } from "../../auth"; +import { auth } from "./auth"; import { isModelAvailableInServer } from "@/app/utils/model"; import { cloudflareAIGatewayUrl } from "@/app/utils/cloudflare"; const ALLOWD_PATH = new Set([Anthropic.ChatPath, Anthropic.ChatPath1]); -async function handle( +export async function handle( req: NextRequest, { params }: { params: { path: string[] } }, ) { @@ -56,30 +56,6 @@ async function handle( } } -export const GET = handle; -export const POST = handle; - -export const runtime = "edge"; -export const preferredRegion = [ - "arn1", - "bom1", - "cdg1", - "cle1", - "cpt1", - "dub1", - "fra1", - "gru1", - "hnd1", - "iad1", - "icn1", - "kix1", - "lhr1", - "pdx1", - "sfo1", - "sin1", - "syd1", -]; - const serverConfig = getServerSideConfig(); async function request(req: NextRequest) { diff --git a/app/api/auth.ts b/app/api/auth.ts index ff52dcd6e..95965ceec 100644 --- a/app/api/auth.ts +++ b/app/api/auth.ts @@ -88,6 +88,10 @@ export function auth(req: NextRequest, modelProvider: ModelProvider) { case ModelProvider.Moonshot: systemApiKey = serverConfig.moonshotApiKey; break; + case ModelProvider.Iflytek: + systemApiKey = + serverConfig.iflytekApiKey + ":" + serverConfig.iflytekApiSecret; + break; case ModelProvider.GPT: default: if (req.nextUrl.pathname.includes("azure/deployments")) { diff --git a/app/api/azure/[...path]/route.ts b/app/api/azure.ts similarity index 66% rename from app/api/azure/[...path]/route.ts rename to app/api/azure.ts index 4a17de0c8..e2cb0c7e6 100644 --- a/app/api/azure/[...path]/route.ts +++ b/app/api/azure.ts @@ -2,10 +2,10 @@ import { getServerSideConfig } from "@/app/config/server"; import { ModelProvider } from "@/app/constant"; import { prettyObject } from "@/app/utils/format"; import { NextRequest, NextResponse } from "next/server"; -import { auth } from "../../auth"; -import { requestOpenai } from "../../common"; +import { auth } from "./auth"; +import { requestOpenai } from "./common"; -async function handle( +export async function handle( req: NextRequest, { params }: { params: { path: string[] } }, ) { @@ -31,27 +31,3 @@ async function handle( return NextResponse.json(prettyObject(e)); } } - -export const GET = handle; -export const POST = handle; - -export const runtime = "edge"; -export const preferredRegion = [ - "arn1", - "bom1", - "cdg1", - "cle1", - "cpt1", - "dub1", - "fra1", - "gru1", - "hnd1", - "iad1", - "icn1", - "kix1", - "lhr1", - "pdx1", - "sfo1", - "sin1", - "syd1", -]; diff --git a/app/api/baidu/[...path]/route.ts b/app/api/baidu.ts similarity index 91% rename from app/api/baidu/[...path]/route.ts rename to app/api/baidu.ts index 94c9963c7..f4315d186 100644 --- a/app/api/baidu/[...path]/route.ts +++ b/app/api/baidu.ts @@ -14,7 +14,7 @@ import { getAccessToken } from "@/app/utils/baidu"; const serverConfig = getServerSideConfig(); -async function handle( +export async function handle( req: NextRequest, { params }: { params: { path: string[] } }, ) { @@ -52,30 +52,6 @@ async function handle( } } -export const GET = handle; -export const POST = handle; - -export const runtime = "edge"; -export const preferredRegion = [ - "arn1", - "bom1", - "cdg1", - "cle1", - "cpt1", - "dub1", - "fra1", - "gru1", - "hnd1", - "iad1", - "icn1", - "kix1", - "lhr1", - "pdx1", - "sfo1", - "sin1", - "syd1", -]; - async function request(req: NextRequest) { const controller = new AbortController(); diff --git a/app/api/bytedance/[...path]/route.ts b/app/api/bytedance.ts similarity index 90% rename from app/api/bytedance/[...path]/route.ts rename to app/api/bytedance.ts index 336c837f0..cb65b1061 100644 --- a/app/api/bytedance/[...path]/route.ts +++ b/app/api/bytedance.ts @@ -12,7 +12,7 @@ import { isModelAvailableInServer } from "@/app/utils/model"; const serverConfig = getServerSideConfig(); -async function handle( +export async function handle( req: NextRequest, { params }: { params: { path: string[] } }, ) { @@ -38,30 +38,6 @@ async function handle( } } -export const GET = handle; -export const POST = handle; - -export const runtime = "edge"; -export const preferredRegion = [ - "arn1", - "bom1", - "cdg1", - "cle1", - "cpt1", - "dub1", - "fra1", - "gru1", - "hnd1", - "iad1", - "icn1", - "kix1", - "lhr1", - "pdx1", - "sfo1", - "sin1", - "syd1", -]; - async function request(req: NextRequest) { const controller = new AbortController(); diff --git a/app/api/common.ts b/app/api/common.ts index 24453dd96..25decbf62 100644 --- a/app/api/common.ts +++ b/app/api/common.ts @@ -32,10 +32,7 @@ export async function requestOpenai(req: NextRequest) { authHeaderName = "Authorization"; } - let path = `${req.nextUrl.pathname}${req.nextUrl.search}`.replaceAll( - "/api/openai/", - "", - ); + let path = `${req.nextUrl.pathname}`.replaceAll("/api/openai/", ""); let baseUrl = (isAzure ? serverConfig.azureUrl : serverConfig.baseUrl) || OPENAI_BASE_URL; diff --git a/app/api/google/[...path]/route.ts b/app/api/google.ts similarity index 96% rename from app/api/google/[...path]/route.ts rename to app/api/google.ts index 83a7ce794..98fe469bf 100644 --- a/app/api/google/[...path]/route.ts +++ b/app/api/google.ts @@ -1,5 +1,5 @@ import { NextRequest, NextResponse } from "next/server"; -import { auth } from "../../auth"; +import { auth } from "./auth"; import { getServerSideConfig } from "@/app/config/server"; import { ApiPath, @@ -11,9 +11,9 @@ import { prettyObject } from "@/app/utils/format"; const serverConfig = getServerSideConfig(); -async function handle( +export async function handle( req: NextRequest, - { params }: { params: { path: string[] } }, + { params }: { params: { provider: string; path: string[] } }, ) { console.log("[Google Route] params ", params); diff --git a/app/api/iflytek.ts b/app/api/iflytek.ts new file mode 100644 index 000000000..eabdd9f4c --- /dev/null +++ b/app/api/iflytek.ts @@ -0,0 +1,131 @@ +import { getServerSideConfig } from "@/app/config/server"; +import { + Iflytek, + IFLYTEK_BASE_URL, + ApiPath, + ModelProvider, + ServiceProvider, +} from "@/app/constant"; +import { prettyObject } from "@/app/utils/format"; +import { NextRequest, NextResponse } from "next/server"; +import { auth } from "@/app/api/auth"; +import { isModelAvailableInServer } from "@/app/utils/model"; +import type { RequestPayload } from "@/app/client/platforms/openai"; +// iflytek + +const serverConfig = getServerSideConfig(); + +export async function handle( + req: NextRequest, + { params }: { params: { path: string[] } }, +) { + console.log("[Iflytek Route] params ", params); + + if (req.method === "OPTIONS") { + return NextResponse.json({ body: "OK" }, { status: 200 }); + } + + const authResult = auth(req, ModelProvider.Iflytek); + if (authResult.error) { + return NextResponse.json(authResult, { + status: 401, + }); + } + + try { + const response = await request(req); + return response; + } catch (e) { + console.error("[Iflytek] ", e); + return NextResponse.json(prettyObject(e)); + } +} + +async function request(req: NextRequest) { + const controller = new AbortController(); + + // iflytek use base url or just remove the path + let path = `${req.nextUrl.pathname}`.replaceAll(ApiPath.Iflytek, ""); + + let baseUrl = serverConfig.iflytekUrl || IFLYTEK_BASE_URL; + + if (!baseUrl.startsWith("http")) { + baseUrl = `https://${baseUrl}`; + } + + if (baseUrl.endsWith("/")) { + baseUrl = baseUrl.slice(0, -1); + } + + console.log("[Proxy] ", path); + console.log("[Base Url]", baseUrl); + + const timeoutId = setTimeout( + () => { + controller.abort(); + }, + 10 * 60 * 1000, + ); + + const fetchUrl = `${baseUrl}${path}`; + const fetchOptions: RequestInit = { + headers: { + "Content-Type": "application/json", + Authorization: req.headers.get("Authorization") ?? "", + }, + method: req.method, + body: req.body, + redirect: "manual", + // @ts-ignore + duplex: "half", + signal: controller.signal, + }; + + // try to refuse some request to some models + if (serverConfig.customModels && req.body) { + try { + const clonedBody = await req.text(); + fetchOptions.body = clonedBody; + + const jsonBody = JSON.parse(clonedBody) as { model?: string }; + + // not undefined and is false + if ( + isModelAvailableInServer( + serverConfig.customModels, + jsonBody?.model as string, + ServiceProvider.Iflytek as string, + ) + ) { + return NextResponse.json( + { + error: true, + message: `you are not allowed to use ${jsonBody?.model} model`, + }, + { + status: 403, + }, + ); + } + } catch (e) { + console.error(`[Iflytek] filter`, e); + } + } + try { + const res = await fetch(fetchUrl, fetchOptions); + + // to prevent browser prompt for credentials + const newHeaders = new Headers(res.headers); + newHeaders.delete("www-authenticate"); + // to disable nginx buffering + newHeaders.set("X-Accel-Buffering", "no"); + + return new Response(res.body, { + status: res.status, + statusText: res.statusText, + headers: newHeaders, + }); + } finally { + clearTimeout(timeoutId); + } +} diff --git a/app/api/moonshot/[...path]/route.ts b/app/api/moonshot.ts similarity index 91% rename from app/api/moonshot/[...path]/route.ts rename to app/api/moonshot.ts index 14bc0a40d..247dd6183 100644 --- a/app/api/moonshot/[...path]/route.ts +++ b/app/api/moonshot.ts @@ -14,7 +14,7 @@ import type { RequestPayload } from "@/app/client/platforms/openai"; const serverConfig = getServerSideConfig(); -async function handle( +export async function handle( req: NextRequest, { params }: { params: { path: string[] } }, ) { @@ -40,30 +40,6 @@ async function handle( } } -export const GET = handle; -export const POST = handle; - -export const runtime = "edge"; -export const preferredRegion = [ - "arn1", - "bom1", - "cdg1", - "cle1", - "cpt1", - "dub1", - "fra1", - "gru1", - "hnd1", - "iad1", - "icn1", - "kix1", - "lhr1", - "pdx1", - "sfo1", - "sin1", - "syd1", -]; - async function request(req: NextRequest) { const controller = new AbortController(); diff --git a/app/api/openai/[...path]/route.ts b/app/api/openai.ts similarity index 80% rename from app/api/openai/[...path]/route.ts rename to app/api/openai.ts index 77059c151..7dfd84e17 100644 --- a/app/api/openai/[...path]/route.ts +++ b/app/api/openai.ts @@ -3,8 +3,8 @@ import { getServerSideConfig } from "@/app/config/server"; import { ModelProvider, OpenaiPath } from "@/app/constant"; import { prettyObject } from "@/app/utils/format"; import { NextRequest, NextResponse } from "next/server"; -import { auth } from "../../auth"; -import { requestOpenai } from "../../common"; +import { auth } from "./auth"; +import { requestOpenai } from "./common"; const ALLOWD_PATH = new Set(Object.values(OpenaiPath)); @@ -13,14 +13,16 @@ function getModels(remoteModelRes: OpenAIListModelResponse) { if (config.disableGPT4) { remoteModelRes.data = remoteModelRes.data.filter( - (m) => !m.id.startsWith("gpt-4"), + (m) => + !(m.id.startsWith("gpt-4") || m.id.startsWith("chatgpt-4o")) || + m.id.startsWith("gpt-4o-mini"), ); } return remoteModelRes; } -async function handle( +export async function handle( req: NextRequest, { params }: { params: { path: string[] } }, ) { @@ -70,27 +72,3 @@ async function handle( return NextResponse.json(prettyObject(e)); } } - -export const GET = handle; -export const POST = handle; - -export const runtime = "edge"; -export const preferredRegion = [ - "arn1", - "bom1", - "cdg1", - "cle1", - "cpt1", - "dub1", - "fra1", - "gru1", - "hnd1", - "iad1", - "icn1", - "kix1", - "lhr1", - "pdx1", - "sfo1", - "sin1", - "syd1", -]; diff --git a/app/api/proxy.ts b/app/api/proxy.ts new file mode 100644 index 000000000..731003aa1 --- /dev/null +++ b/app/api/proxy.ts @@ -0,0 +1,75 @@ +import { NextRequest, NextResponse } from "next/server"; + +export async function handle( + req: NextRequest, + { params }: { params: { path: string[] } }, +) { + console.log("[Proxy Route] params ", params); + + if (req.method === "OPTIONS") { + return NextResponse.json({ body: "OK" }, { status: 200 }); + } + + // remove path params from searchParams + req.nextUrl.searchParams.delete("path"); + req.nextUrl.searchParams.delete("provider"); + + const subpath = params.path.join("/"); + const fetchUrl = `${req.headers.get( + "x-base-url", + )}/${subpath}?${req.nextUrl.searchParams.toString()}`; + const skipHeaders = ["connection", "host", "origin", "referer", "cookie"]; + const headers = new Headers( + Array.from(req.headers.entries()).filter((item) => { + if ( + item[0].indexOf("x-") > -1 || + item[0].indexOf("sec-") > -1 || + skipHeaders.includes(item[0]) + ) { + return false; + } + return true; + }), + ); + const controller = new AbortController(); + const fetchOptions: RequestInit = { + headers, + method: req.method, + body: req.body, + // to fix #2485: https://stackoverflow.com/questions/55920957/cloudflare-worker-typeerror-one-time-use-body + redirect: "manual", + // @ts-ignore + duplex: "half", + signal: controller.signal, + }; + + const timeoutId = setTimeout( + () => { + controller.abort(); + }, + 10 * 60 * 1000, + ); + + try { + const res = await fetch(fetchUrl, fetchOptions); + // to prevent browser prompt for credentials + const newHeaders = new Headers(res.headers); + newHeaders.delete("www-authenticate"); + // to disable nginx buffering + newHeaders.set("X-Accel-Buffering", "no"); + + // The latest version of the OpenAI API forced the content-encoding to be "br" in json response + // So if the streaming is disabled, we need to remove the content-encoding header + // Because Vercel uses gzip to compress the response, if we don't remove the content-encoding header + // The browser will try to decode the response with brotli and fail + newHeaders.delete("content-encoding"); + + return new Response(res.body, { + status: res.status, + statusText: res.statusText, + headers: newHeaders, + }); + } finally { + clearTimeout(timeoutId); + } +} diff --git a/app/api/stability/[...path]/route.ts b/app/api/stability.ts similarity index 95% rename from app/api/stability/[...path]/route.ts rename to app/api/stability.ts index 4b2bcc305..2646ace85 100644 --- a/app/api/stability/[...path]/route.ts +++ b/app/api/stability.ts @@ -3,7 +3,7 @@ import { getServerSideConfig } from "@/app/config/server"; import { ModelProvider, STABILITY_BASE_URL } from "@/app/constant"; import { auth } from "@/app/api/auth"; -async function handle( +export async function handle( req: NextRequest, { params }: { params: { path: string[] } }, ) { @@ -97,8 +97,3 @@ async function handle( clearTimeout(timeoutId); } } - -export const GET = handle; -export const POST = handle; - -export const runtime = "edge"; diff --git a/app/api/tencent/route.ts b/app/api/tencent/route.ts index d506d1016..885909e7a 100644 --- a/app/api/tencent/route.ts +++ b/app/api/tencent/route.ts @@ -43,7 +43,7 @@ async function handle( export const GET = handle; export const POST = handle; -export const runtime = "nodejs"; +export const runtime = "edge"; export const preferredRegion = [ "arn1", "bom1", diff --git a/app/api/webdav/[...path]/route.ts b/app/api/webdav/[...path]/route.ts index 1f58a884f..9f96cbfcf 100644 --- a/app/api/webdav/[...path]/route.ts +++ b/app/api/webdav/[...path]/route.ts @@ -29,6 +29,7 @@ async function handle( const requestUrl = new URL(req.url); let endpoint = requestUrl.searchParams.get("endpoint"); + let proxy_method = requestUrl.searchParams.get("proxy_method") || req.method; // Validate the endpoint to prevent potential SSRF attacks if ( @@ -65,7 +66,11 @@ async function handle( const targetPath = `${endpoint}${endpointPath}`; // only allow MKCOL, GET, PUT - if (req.method !== "MKCOL" && req.method !== "GET" && req.method !== "PUT") { + if ( + proxy_method !== "MKCOL" && + proxy_method !== "GET" && + proxy_method !== "PUT" + ) { return NextResponse.json( { error: true, @@ -78,7 +83,7 @@ async function handle( } // for MKCOL request, only allow request ${folder} - if (req.method === "MKCOL" && !targetPath.endsWith(folder)) { + if (proxy_method === "MKCOL" && !targetPath.endsWith(folder)) { return NextResponse.json( { error: true, @@ -91,7 +96,7 @@ async function handle( } // for GET request, only allow request ending with fileName - if (req.method === "GET" && !targetPath.endsWith(fileName)) { + if (proxy_method === "GET" && !targetPath.endsWith(fileName)) { return NextResponse.json( { error: true, @@ -104,7 +109,7 @@ async function handle( } // for PUT request, only allow request ending with fileName - if (req.method === "PUT" && !targetPath.endsWith(fileName)) { + if (proxy_method === "PUT" && !targetPath.endsWith(fileName)) { return NextResponse.json( { error: true, @@ -118,7 +123,7 @@ async function handle( const targetUrl = targetPath; - const method = req.method; + const method = proxy_method || req.method; const shouldNotHaveBody = ["get", "head"].includes( method?.toLowerCase() ?? "", ); @@ -143,7 +148,7 @@ async function handle( "[Any Proxy]", targetUrl, { - method: req.method, + method: method, }, { status: fetchResult?.status, diff --git a/app/client/api.ts b/app/client/api.ts index f10e47618..cecc453ba 100644 --- a/app/client/api.ts +++ b/app/client/api.ts @@ -5,8 +5,14 @@ import { ModelProvider, ServiceProvider, } from "../constant"; -import { ChatMessage, ModelType, useAccessStore, useChatStore } from "../store"; -import { ChatGPTApi } from "./platforms/openai"; +import { + ChatMessageTool, + ChatMessage, + ModelType, + useAccessStore, + useChatStore, +} from "../store"; +import { ChatGPTApi, DalleRequestPayload } from "./platforms/openai"; import { GeminiProApi } from "./platforms/google"; import { ClaudeApi } from "./platforms/anthropic"; import { ErnieApi } from "./platforms/baidu"; @@ -14,6 +20,7 @@ import { DoubaoApi } from "./platforms/bytedance"; import { QwenApi } from "./platforms/alibaba"; import { HunyuanApi } from "./platforms/tencent"; import { MoonshotApi } from "./platforms/moonshot"; +import { SparkApi } from "./platforms/iflytek"; export const ROLES = ["system", "user", "assistant"] as const; export type MessageRole = (typeof ROLES)[number]; @@ -42,6 +49,9 @@ export interface LLMConfig { stream?: boolean; presence_penalty?: number; frequency_penalty?: number; + size?: DalleRequestPayload["size"]; + quality?: DalleRequestPayload["quality"]; + style?: DalleRequestPayload["style"]; } export interface ChatOptions { @@ -52,6 +62,8 @@ export interface ChatOptions { onFinish: (message: string) => void; onError?: (err: Error) => void; onController?: (controller: AbortController) => void; + onBeforeTool?: (tool: ChatMessageTool) => void; + onAfterTool?: (tool: ChatMessageTool) => void; } export interface LLMUsage { @@ -64,12 +76,14 @@ export interface LLMModel { displayName?: string; available: boolean; provider: LLMModelProvider; + sorted: number; } export interface LLMModelProvider { id: string; providerName: string; providerType: string; + sorted: number; } export abstract class LLMApi { @@ -118,12 +132,16 @@ export class ClientApi { break; case ModelProvider.Qwen: this.llm = new QwenApi(); + break; case ModelProvider.Hunyuan: this.llm = new HunyuanApi(); break; case ModelProvider.Moonshot: this.llm = new MoonshotApi(); break; + case ModelProvider.Iflytek: + this.llm = new SparkApi(); + break; default: this.llm = new ChatGPTApi(); } @@ -207,6 +225,7 @@ export function getHeaders() { const isByteDance = modelConfig.providerName === ServiceProvider.ByteDance; const isAlibaba = modelConfig.providerName === ServiceProvider.Alibaba; const isMoonshot = modelConfig.providerName === ServiceProvider.Moonshot; + const isIflytek = modelConfig.providerName === ServiceProvider.Iflytek; const isEnabledAccessControl = accessStore.enabledAccessControl(); const apiKey = isGoogle ? accessStore.googleApiKey @@ -220,6 +239,10 @@ export function getHeaders() { ? accessStore.alibabaApiKey : isMoonshot ? accessStore.moonshotApiKey + : isIflytek + ? accessStore.iflytekApiKey && accessStore.iflytekApiSecret + ? accessStore.iflytekApiKey + ":" + accessStore.iflytekApiSecret + : "" : accessStore.openaiApiKey; return { isGoogle, @@ -229,6 +252,7 @@ export function getHeaders() { isByteDance, isAlibaba, isMoonshot, + isIflytek, apiKey, isEnabledAccessControl, }; @@ -282,6 +306,8 @@ export function getClientApi(provider: ServiceProvider): ClientApi { return new ClientApi(ModelProvider.Hunyuan); case ServiceProvider.Moonshot: return new ClientApi(ModelProvider.Moonshot); + case ServiceProvider.Iflytek: + return new ClientApi(ModelProvider.Iflytek); default: return new ClientApi(ModelProvider.GPT); } diff --git a/app/client/platforms/anthropic.ts b/app/client/platforms/anthropic.ts index b079ba1ad..7dd39c9cd 100644 --- a/app/client/platforms/anthropic.ts +++ b/app/client/platforms/anthropic.ts @@ -1,6 +1,12 @@ import { ACCESS_CODE_PREFIX, Anthropic, ApiPath } from "@/app/constant"; import { ChatOptions, getHeaders, LLMApi, MultimodalContent } from "../api"; -import { useAccessStore, useAppConfig, useChatStore } from "@/app/store"; +import { + useAccessStore, + useAppConfig, + useChatStore, + usePluginStore, + ChatMessageTool, +} from "@/app/store"; import { getClientConfig } from "@/app/config/client"; import { DEFAULT_API_HOST } from "@/app/constant"; import { @@ -11,8 +17,9 @@ import { import Locale from "../../locales"; import { prettyObject } from "@/app/utils/format"; import { getMessageTextContent, isVisionModel } from "@/app/utils"; -import { preProcessImageContent } from "@/app/utils/chat"; +import { preProcessImageContent, stream } from "@/app/utils/chat"; import { cloudflareAIGatewayUrl } from "@/app/utils/cloudflare"; +import { RequestPayload } from "./openai"; export type MultiBlockContent = { type: "image" | "text"; @@ -191,112 +198,126 @@ export class ClaudeApi implements LLMApi { const controller = new AbortController(); options.onController?.(controller); - const payload = { - method: "POST", - body: JSON.stringify(requestBody), - signal: controller.signal, - headers: { - ...getHeaders(), // get common headers - "anthropic-version": accessStore.anthropicApiVersion, - // do not send `anthropicApiKey` in browser!!! - // Authorization: getAuthKey(accessStore.anthropicApiKey), - }, - }; - if (shouldStream) { - try { - const context = { - text: "", - finished: false, - }; - - const finish = () => { - if (!context.finished) { - options.onFinish(context.text); - context.finished = true; - } - }; - - controller.signal.onabort = finish; - fetchEventSource(path, { - ...payload, - async onopen(res) { - const contentType = res.headers.get("content-type"); - console.log("response content type: ", contentType); - - if (contentType?.startsWith("text/plain")) { - context.text = await res.clone().text(); - return finish(); - } - - if ( - !res.ok || - !res.headers - .get("content-type") - ?.startsWith(EventStreamContentType) || - res.status !== 200 - ) { - const responseTexts = [context.text]; - let extraInfo = await res.clone().text(); - try { - const resJson = await res.clone().json(); - extraInfo = prettyObject(resJson); - } catch {} - - if (res.status === 401) { - responseTexts.push(Locale.Error.Unauthorized); - } - - if (extraInfo) { - responseTexts.push(extraInfo); - } - - context.text = responseTexts.join("\n\n"); - - return finish(); - } - }, - onmessage(msg) { - let chunkJson: - | undefined - | { - type: "content_block_delta" | "content_block_stop"; - delta?: { - type: "text_delta"; - text: string; - }; - index: number; + let index = -1; + const [tools, funcs] = usePluginStore + .getState() + .getAsTools( + useChatStore.getState().currentSession().mask?.plugin || [], + ); + return stream( + path, + requestBody, + { + ...getHeaders(), + "anthropic-version": accessStore.anthropicApiVersion, + }, + // @ts-ignore + tools.map((tool) => ({ + name: tool?.function?.name, + description: tool?.function?.description, + input_schema: tool?.function?.parameters, + })), + funcs, + controller, + // parseSSE + (text: string, runTools: ChatMessageTool[]) => { + // console.log("parseSSE", text, runTools); + let chunkJson: + | undefined + | { + type: "content_block_delta" | "content_block_stop"; + content_block?: { + type: "tool_use"; + id: string; + name: string; }; - try { - chunkJson = JSON.parse(msg.data); - } catch (e) { - console.error("[Response] parse error", msg.data); - } + delta?: { + type: "text_delta" | "input_json_delta"; + text?: string; + partial_json?: string; + }; + index: number; + }; + chunkJson = JSON.parse(text); - if (!chunkJson || chunkJson.type === "content_block_stop") { - return finish(); - } - - const { delta } = chunkJson; - if (delta?.text) { - context.text += delta.text; - options.onUpdate?.(context.text, delta.text); - } - }, - onclose() { - finish(); - }, - onerror(e) { - options.onError?.(e); - throw e; - }, - openWhenHidden: true, - }); - } catch (e) { - console.error("failed to chat", e); - options.onError?.(e as Error); - } + if (chunkJson?.content_block?.type == "tool_use") { + index += 1; + const id = chunkJson?.content_block.id; + const name = chunkJson?.content_block.name; + runTools.push({ + id, + type: "function", + function: { + name, + arguments: "", + }, + }); + } + if ( + chunkJson?.delta?.type == "input_json_delta" && + chunkJson?.delta?.partial_json + ) { + // @ts-ignore + runTools[index]["function"]["arguments"] += + chunkJson?.delta?.partial_json; + } + return chunkJson?.delta?.text; + }, + // processToolMessage, include tool_calls message and tool call results + ( + requestPayload: RequestPayload, + toolCallMessage: any, + toolCallResult: any[], + ) => { + // reset index value + index = -1; + // @ts-ignore + requestPayload?.messages?.splice( + // @ts-ignore + requestPayload?.messages?.length, + 0, + { + role: "assistant", + content: toolCallMessage.tool_calls.map( + (tool: ChatMessageTool) => ({ + type: "tool_use", + id: tool.id, + name: tool?.function?.name, + input: tool?.function?.arguments + ? JSON.parse(tool?.function?.arguments) + : {}, + }), + ), + }, + // @ts-ignore + ...toolCallResult.map((result) => ({ + role: "user", + content: [ + { + type: "tool_result", + tool_use_id: result.tool_call_id, + content: result.content, + }, + ], + })), + ); + }, + options, + ); } else { + const payload = { + method: "POST", + body: JSON.stringify(requestBody), + signal: controller.signal, + headers: { + ...getHeaders(), // get common headers + "anthropic-version": accessStore.anthropicApiVersion, + // do not send `anthropicApiKey` in browser!!! + // Authorization: getAuthKey(accessStore.anthropicApiKey), + }, + }; + try { controller.signal.onabort = () => options.onFinish(""); diff --git a/app/client/platforms/baidu.ts b/app/client/platforms/baidu.ts index 188b78bf9..3be147f49 100644 --- a/app/client/platforms/baidu.ts +++ b/app/client/platforms/baidu.ts @@ -77,16 +77,24 @@ export class ErnieApi implements LLMApi { async chat(options: ChatOptions) { const messages = options.messages.map((v) => ({ - role: v.role, + // "error_code": 336006, "error_msg": "the role of message with even index in the messages must be user or function", + role: v.role === "system" ? "user" : v.role, content: getMessageTextContent(v), })); // "error_code": 336006, "error_msg": "the length of messages must be an odd number", if (messages.length % 2 === 0) { - messages.unshift({ - role: "user", - content: " ", - }); + if (messages.at(0)?.role === "user") { + messages.splice(1, 0, { + role: "assistant", + content: " ", + }); + } else { + messages.unshift({ + role: "user", + content: " ", + }); + } } const modelConfig = { diff --git a/app/client/platforms/iflytek.ts b/app/client/platforms/iflytek.ts new file mode 100644 index 000000000..73cea5ba0 --- /dev/null +++ b/app/client/platforms/iflytek.ts @@ -0,0 +1,240 @@ +"use client"; +import { + ApiPath, + DEFAULT_API_HOST, + Iflytek, + REQUEST_TIMEOUT_MS, +} from "@/app/constant"; +import { useAccessStore, useAppConfig, useChatStore } from "@/app/store"; + +import { ChatOptions, getHeaders, LLMApi, LLMModel } from "../api"; +import Locale from "../../locales"; +import { + EventStreamContentType, + fetchEventSource, +} from "@fortaine/fetch-event-source"; +import { prettyObject } from "@/app/utils/format"; +import { getClientConfig } from "@/app/config/client"; +import { getMessageTextContent } from "@/app/utils"; + +import { OpenAIListModelResponse, RequestPayload } from "./openai"; + +export class SparkApi implements LLMApi { + private disableListModels = true; + + path(path: string): string { + const accessStore = useAccessStore.getState(); + + let baseUrl = ""; + + if (accessStore.useCustomConfig) { + baseUrl = accessStore.iflytekUrl; + } + + if (baseUrl.length === 0) { + const isApp = !!getClientConfig()?.isApp; + const apiPath = ApiPath.Iflytek; + baseUrl = isApp ? DEFAULT_API_HOST + "/proxy" + apiPath : apiPath; + } + + if (baseUrl.endsWith("/")) { + baseUrl = baseUrl.slice(0, baseUrl.length - 1); + } + if (!baseUrl.startsWith("http") && !baseUrl.startsWith(ApiPath.Iflytek)) { + baseUrl = "https://" + baseUrl; + } + + console.log("[Proxy Endpoint] ", baseUrl, path); + + return [baseUrl, path].join("/"); + } + + extractMessage(res: any) { + return res.choices?.at(0)?.message?.content ?? ""; + } + + async chat(options: ChatOptions) { + const messages: ChatOptions["messages"] = []; + for (const v of options.messages) { + const content = getMessageTextContent(v); + messages.push({ role: v.role, content }); + } + + const modelConfig = { + ...useAppConfig.getState().modelConfig, + ...useChatStore.getState().currentSession().mask.modelConfig, + ...{ + model: options.config.model, + providerName: options.config.providerName, + }, + }; + + const requestPayload: RequestPayload = { + messages, + stream: options.config.stream, + model: modelConfig.model, + temperature: modelConfig.temperature, + presence_penalty: modelConfig.presence_penalty, + frequency_penalty: modelConfig.frequency_penalty, + top_p: modelConfig.top_p, + // max_tokens: Math.max(modelConfig.max_tokens, 1024), + // Please do not ask me why not send max_tokens, no reason, this param is just shit, I dont want to explain anymore. + }; + + console.log("[Request] Spark payload: ", requestPayload); + + const shouldStream = !!options.config.stream; + const controller = new AbortController(); + options.onController?.(controller); + + try { + const chatPath = this.path(Iflytek.ChatPath); + const chatPayload = { + method: "POST", + body: JSON.stringify(requestPayload), + signal: controller.signal, + headers: getHeaders(), + }; + + // Make a fetch request + const requestTimeoutId = setTimeout( + () => controller.abort(), + REQUEST_TIMEOUT_MS, + ); + + if (shouldStream) { + let responseText = ""; + let remainText = ""; + let finished = false; + + // Animate response text to make it look smooth + function animateResponseText() { + if (finished || controller.signal.aborted) { + responseText += remainText; + console.log("[Response Animation] finished"); + return; + } + + if (remainText.length > 0) { + const fetchCount = Math.max(1, Math.round(remainText.length / 60)); + const fetchText = remainText.slice(0, fetchCount); + responseText += fetchText; + remainText = remainText.slice(fetchCount); + options.onUpdate?.(responseText, fetchText); + } + + requestAnimationFrame(animateResponseText); + } + + // Start animation + animateResponseText(); + + const finish = () => { + if (!finished) { + finished = true; + options.onFinish(responseText + remainText); + } + }; + + controller.signal.onabort = finish; + + fetchEventSource(chatPath, { + ...chatPayload, + async onopen(res) { + clearTimeout(requestTimeoutId); + const contentType = res.headers.get("content-type"); + console.log("[Spark] request response content type: ", contentType); + + if (contentType?.startsWith("text/plain")) { + responseText = await res.clone().text(); + return finish(); + } + + // Handle different error scenarios + if ( + !res.ok || + !res.headers + .get("content-type") + ?.startsWith(EventStreamContentType) || + res.status !== 200 + ) { + let extraInfo = await res.clone().text(); + try { + const resJson = await res.clone().json(); + extraInfo = prettyObject(resJson); + } catch {} + + if (res.status === 401) { + extraInfo = Locale.Error.Unauthorized; + } + + options.onError?.( + new Error( + `Request failed with status ${res.status}: ${extraInfo}`, + ), + ); + return finish(); + } + }, + onmessage(msg) { + if (msg.data === "[DONE]" || finished) { + return finish(); + } + const text = msg.data; + try { + const json = JSON.parse(text); + const choices = json.choices as Array<{ + delta: { content: string }; + }>; + const delta = choices[0]?.delta?.content; + + if (delta) { + remainText += delta; + } + } catch (e) { + console.error("[Request] parse error", text); + options.onError?.(new Error(`Failed to parse response: ${text}`)); + } + }, + onclose() { + finish(); + }, + onerror(e) { + options.onError?.(e); + throw e; + }, + openWhenHidden: true, + }); + } else { + const res = await fetch(chatPath, chatPayload); + clearTimeout(requestTimeoutId); + + if (!res.ok) { + const errorText = await res.text(); + options.onError?.( + new Error(`Request failed with status ${res.status}: ${errorText}`), + ); + return; + } + + const resJson = await res.json(); + const message = this.extractMessage(resJson); + options.onFinish(message); + } + } catch (e) { + console.log("[Request] failed to make a chat request", e); + options.onError?.(e as Error); + } + } + + async usage() { + return { + used: 0, + total: 0, + }; + } + + async models(): Promise { + return []; + } +} diff --git a/app/client/platforms/moonshot.ts b/app/client/platforms/moonshot.ts index 7d257ccb2..cd10d2f6c 100644 --- a/app/client/platforms/moonshot.ts +++ b/app/client/platforms/moonshot.ts @@ -8,9 +8,15 @@ import { REQUEST_TIMEOUT_MS, ServiceProvider, } from "@/app/constant"; -import { useAccessStore, useAppConfig, useChatStore } from "@/app/store"; +import { + useAccessStore, + useAppConfig, + useChatStore, + ChatMessageTool, + usePluginStore, +} from "@/app/store"; import { collectModelsWithDefaultModel } from "@/app/utils/model"; -import { preProcessImageContent } from "@/app/utils/chat"; +import { preProcessImageContent, stream } from "@/app/utils/chat"; import { cloudflareAIGatewayUrl } from "@/app/utils/cloudflare"; import { @@ -116,115 +122,66 @@ export class MoonshotApi implements LLMApi { ); if (shouldStream) { - let responseText = ""; - let remainText = ""; - let finished = false; - - // animate response to make it looks smooth - function animateResponseText() { - if (finished || controller.signal.aborted) { - responseText += remainText; - console.log("[Response Animation] finished"); - if (responseText?.length === 0) { - options.onError?.(new Error("empty response from server")); + const [tools, funcs] = usePluginStore + .getState() + .getAsTools( + useChatStore.getState().currentSession().mask?.plugin || [], + ); + return stream( + chatPath, + requestPayload, + getHeaders(), + tools as any, + funcs, + controller, + // parseSSE + (text: string, runTools: ChatMessageTool[]) => { + // console.log("parseSSE", text, runTools); + const json = JSON.parse(text); + const choices = json.choices as Array<{ + delta: { + content: string; + tool_calls: ChatMessageTool[]; + }; + }>; + const tool_calls = choices[0]?.delta?.tool_calls; + if (tool_calls?.length > 0) { + const index = tool_calls[0]?.index; + const id = tool_calls[0]?.id; + const args = tool_calls[0]?.function?.arguments; + if (id) { + runTools.push({ + id, + type: tool_calls[0]?.type, + function: { + name: tool_calls[0]?.function?.name as string, + arguments: args, + }, + }); + } else { + // @ts-ignore + runTools[index]["function"]["arguments"] += args; + } } - return; - } - - if (remainText.length > 0) { - const fetchCount = Math.max(1, Math.round(remainText.length / 60)); - const fetchText = remainText.slice(0, fetchCount); - responseText += fetchText; - remainText = remainText.slice(fetchCount); - options.onUpdate?.(responseText, fetchText); - } - - requestAnimationFrame(animateResponseText); - } - - // start animaion - animateResponseText(); - - const finish = () => { - if (!finished) { - finished = true; - options.onFinish(responseText + remainText); - } - }; - - controller.signal.onabort = finish; - - fetchEventSource(chatPath, { - ...chatPayload, - async onopen(res) { - clearTimeout(requestTimeoutId); - const contentType = res.headers.get("content-type"); - console.log( - "[OpenAI] request response content type: ", - contentType, + return choices[0]?.delta?.content; + }, + // processToolMessage, include tool_calls message and tool call results + ( + requestPayload: RequestPayload, + toolCallMessage: any, + toolCallResult: any[], + ) => { + // @ts-ignore + requestPayload?.messages?.splice( + // @ts-ignore + requestPayload?.messages?.length, + 0, + toolCallMessage, + ...toolCallResult, ); - - if (contentType?.startsWith("text/plain")) { - responseText = await res.clone().text(); - return finish(); - } - - if ( - !res.ok || - !res.headers - .get("content-type") - ?.startsWith(EventStreamContentType) || - res.status !== 200 - ) { - const responseTexts = [responseText]; - let extraInfo = await res.clone().text(); - try { - const resJson = await res.clone().json(); - extraInfo = prettyObject(resJson); - } catch {} - - if (res.status === 401) { - responseTexts.push(Locale.Error.Unauthorized); - } - - if (extraInfo) { - responseTexts.push(extraInfo); - } - - responseText = responseTexts.join("\n\n"); - - return finish(); - } }, - onmessage(msg) { - if (msg.data === "[DONE]" || finished) { - return finish(); - } - const text = msg.data; - try { - const json = JSON.parse(text); - const choices = json.choices as Array<{ - delta: { content: string }; - }>; - const delta = choices[0]?.delta?.content; - const textmoderation = json?.prompt_filter_results; - - if (delta) { - remainText += delta; - } - } catch (e) { - console.error("[Request] parse error", text, msg); - } - }, - onclose() { - finish(); - }, - onerror(e) { - options.onError?.(e); - throw e; - }, - openWhenHidden: true, - }); + options, + ); } else { const res = await fetch(chatPath, chatPayload); clearTimeout(requestTimeoutId); diff --git a/app/client/platforms/openai.ts b/app/client/platforms/openai.ts index 680125fe6..664ff872b 100644 --- a/app/client/platforms/openai.ts +++ b/app/client/platforms/openai.ts @@ -9,10 +9,22 @@ import { REQUEST_TIMEOUT_MS, ServiceProvider, } from "@/app/constant"; -import { useAccessStore, useAppConfig, useChatStore } from "@/app/store"; +import { + ChatMessageTool, + useAccessStore, + useAppConfig, + useChatStore, + usePluginStore, +} from "@/app/store"; import { collectModelsWithDefaultModel } from "@/app/utils/model"; -import { preProcessImageContent } from "@/app/utils/chat"; +import { + preProcessImageContent, + uploadImage, + base64Image2Blob, + stream, +} from "@/app/utils/chat"; import { cloudflareAIGatewayUrl } from "@/app/utils/cloudflare"; +import { DalleSize, DalleQuality, DalleStyle } from "@/app/typing"; import { ChatOptions, @@ -33,6 +45,7 @@ import { getMessageTextContent, getMessageImages, isVisionModel, + isDalle3 as _isDalle3, } from "@/app/utils"; export interface OpenAIListModelResponse { @@ -58,6 +71,16 @@ export interface RequestPayload { max_tokens?: number; } +export interface DalleRequestPayload { + model: string; + prompt: string; + response_format: "url" | "b64_json"; + n: number; + size: DalleSize; + quality: DalleQuality; + style: DalleStyle; +} + export class ChatGPTApi implements LLMApi { private disableListModels = true; @@ -100,20 +123,31 @@ export class ChatGPTApi implements LLMApi { return cloudflareAIGatewayUrl([baseUrl, path].join("/")); } - extractMessage(res: any) { - return res.choices?.at(0)?.message?.content ?? ""; + async extractMessage(res: any) { + if (res.error) { + return "```\n" + JSON.stringify(res, null, 4) + "\n```"; + } + // dalle3 model return url, using url create image message + if (res.data) { + let url = res.data?.at(0)?.url ?? ""; + const b64_json = res.data?.at(0)?.b64_json ?? ""; + if (!url && b64_json) { + // uploadImage + url = await uploadImage(base64Image2Blob(b64_json, "image/png")); + } + return [ + { + type: "image_url", + image_url: { + url, + }, + }, + ]; + } + return res.choices?.at(0)?.message?.content ?? res; } async chat(options: ChatOptions) { - const visionModel = isVisionModel(options.config.model); - const messages: ChatOptions["messages"] = []; - for (const v of options.messages) { - const content = visionModel - ? await preProcessImageContent(v.content) - : getMessageTextContent(v); - messages.push({ role: v.role, content }); - } - const modelConfig = { ...useAppConfig.getState().modelConfig, ...useChatStore.getState().currentSession().mask.modelConfig, @@ -123,26 +157,57 @@ export class ChatGPTApi implements LLMApi { }, }; - const requestPayload: RequestPayload = { - messages, - stream: options.config.stream, - model: modelConfig.model, - temperature: modelConfig.temperature, - presence_penalty: modelConfig.presence_penalty, - frequency_penalty: modelConfig.frequency_penalty, - top_p: modelConfig.top_p, - // max_tokens: Math.max(modelConfig.max_tokens, 1024), - // Please do not ask me why not send max_tokens, no reason, this param is just shit, I dont want to explain anymore. - }; + let requestPayload: RequestPayload | DalleRequestPayload; - // add max_tokens to vision model - if (visionModel && modelConfig.model.includes("preview")) { - requestPayload["max_tokens"] = Math.max(modelConfig.max_tokens, 4000); + const isDalle3 = _isDalle3(options.config.model); + const isO1 = options.config.model.startsWith("o1"); + if (isDalle3) { + const prompt = getMessageTextContent( + options.messages.slice(-1)?.pop() as any, + ); + requestPayload = { + model: options.config.model, + prompt, + // URLs are only valid for 60 minutes after the image has been generated. + response_format: "b64_json", // using b64_json, and save image in CacheStorage + n: 1, + size: options.config?.size ?? "1024x1024", + quality: options.config?.quality ?? "standard", + style: options.config?.style ?? "vivid", + }; + } else { + const visionModel = isVisionModel(options.config.model); + const messages: ChatOptions["messages"] = []; + for (const v of options.messages) { + const content = visionModel + ? await preProcessImageContent(v.content) + : getMessageTextContent(v); + if (!(isO1 && v.role === "system")) + messages.push({ role: v.role, content }); + } + + // O1 not support image, tools (plugin in ChatGPTNextWeb) and system, stream, logprobs, temperature, top_p, n, presence_penalty, frequency_penalty yet. + requestPayload = { + messages, + stream: !isO1 ? options.config.stream : false, + model: modelConfig.model, + temperature: !isO1 ? modelConfig.temperature : 1, + presence_penalty: !isO1 ? modelConfig.presence_penalty : 0, + frequency_penalty: !isO1 ? modelConfig.frequency_penalty : 0, + top_p: !isO1 ? modelConfig.top_p : 1, + // max_tokens: Math.max(modelConfig.max_tokens, 1024), + // Please do not ask me why not send max_tokens, no reason, this param is just shit, I dont want to explain anymore. + }; + + // add max_tokens to vision model + if (visionModel) { + requestPayload["max_tokens"] = Math.max(modelConfig.max_tokens, 4000); + } } console.log("[Request] openai payload: ", requestPayload); - const shouldStream = !!options.config.stream; + const shouldStream = !isDalle3 && !!options.config.stream && !isO1; const controller = new AbortController(); options.onController?.(controller); @@ -168,156 +233,97 @@ export class ChatGPTApi implements LLMApi { model?.provider?.providerName === ServiceProvider.Azure, ); chatPath = this.path( - Azure.ChatPath( + (isDalle3 ? Azure.ImagePath : Azure.ChatPath)( (model?.displayName ?? model?.name) as string, useCustomConfig ? useAccessStore.getState().azureApiVersion : "", ), ); } else { - chatPath = this.path(OpenaiPath.ChatPath); + chatPath = this.path( + isDalle3 ? OpenaiPath.ImagePath : OpenaiPath.ChatPath, + ); } - const chatPayload = { - method: "POST", - body: JSON.stringify(requestPayload), - signal: controller.signal, - headers: getHeaders(), - }; - - // make a fetch request - const requestTimeoutId = setTimeout( - () => controller.abort(), - REQUEST_TIMEOUT_MS, - ); - if (shouldStream) { - let responseText = ""; - let remainText = ""; - let finished = false; - - // animate response to make it looks smooth - function animateResponseText() { - if (finished || controller.signal.aborted) { - responseText += remainText; - console.log("[Response Animation] finished"); - if (responseText?.length === 0) { - options.onError?.(new Error("empty response from server")); + const [tools, funcs] = usePluginStore + .getState() + .getAsTools( + useChatStore.getState().currentSession().mask?.plugin || [], + ); + // console.log("getAsTools", tools, funcs); + stream( + chatPath, + requestPayload, + getHeaders(), + tools as any, + funcs, + controller, + // parseSSE + (text: string, runTools: ChatMessageTool[]) => { + // console.log("parseSSE", text, runTools); + const json = JSON.parse(text); + const choices = json.choices as Array<{ + delta: { + content: string; + tool_calls: ChatMessageTool[]; + }; + }>; + const tool_calls = choices[0]?.delta?.tool_calls; + if (tool_calls?.length > 0) { + const index = tool_calls[0]?.index; + const id = tool_calls[0]?.id; + const args = tool_calls[0]?.function?.arguments; + if (id) { + runTools.push({ + id, + type: tool_calls[0]?.type, + function: { + name: tool_calls[0]?.function?.name as string, + arguments: args, + }, + }); + } else { + // @ts-ignore + runTools[index]["function"]["arguments"] += args; + } } - return; - } - - if (remainText.length > 0) { - const fetchCount = Math.max(1, Math.round(remainText.length / 60)); - const fetchText = remainText.slice(0, fetchCount); - responseText += fetchText; - remainText = remainText.slice(fetchCount); - options.onUpdate?.(responseText, fetchText); - } - - requestAnimationFrame(animateResponseText); - } - - // start animaion - animateResponseText(); - - const finish = () => { - if (!finished) { - finished = true; - options.onFinish(responseText + remainText); - } + return choices[0]?.delta?.content; + }, + // processToolMessage, include tool_calls message and tool call results + ( + requestPayload: RequestPayload, + toolCallMessage: any, + toolCallResult: any[], + ) => { + // @ts-ignore + requestPayload?.messages?.splice( + // @ts-ignore + requestPayload?.messages?.length, + 0, + toolCallMessage, + ...toolCallResult, + ); + }, + options, + ); + } else { + const chatPayload = { + method: "POST", + body: JSON.stringify(requestPayload), + signal: controller.signal, + headers: getHeaders(), }; - controller.signal.onabort = finish; + // make a fetch request + const requestTimeoutId = setTimeout( + () => controller.abort(), + isDalle3 || isO1 ? REQUEST_TIMEOUT_MS * 2 : REQUEST_TIMEOUT_MS, // dalle3 using b64_json is slow. + ); - fetchEventSource(chatPath, { - ...chatPayload, - async onopen(res) { - clearTimeout(requestTimeoutId); - const contentType = res.headers.get("content-type"); - console.log( - "[OpenAI] request response content type: ", - contentType, - ); - - if (contentType?.startsWith("text/plain")) { - responseText = await res.clone().text(); - return finish(); - } - - if ( - !res.ok || - !res.headers - .get("content-type") - ?.startsWith(EventStreamContentType) || - res.status !== 200 - ) { - const responseTexts = [responseText]; - let extraInfo = await res.clone().text(); - try { - const resJson = await res.clone().json(); - extraInfo = prettyObject(resJson); - } catch {} - - if (res.status === 401) { - responseTexts.push(Locale.Error.Unauthorized); - } - - if (extraInfo) { - responseTexts.push(extraInfo); - } - - responseText = responseTexts.join("\n\n"); - - return finish(); - } - }, - onmessage(msg) { - if (msg.data === "[DONE]" || finished) { - return finish(); - } - const text = msg.data; - try { - const json = JSON.parse(text); - const choices = json.choices as Array<{ - delta: { content: string }; - }>; - const delta = choices[0]?.delta?.content; - const textmoderation = json?.prompt_filter_results; - - if (delta) { - remainText += delta; - } - - if ( - textmoderation && - textmoderation.length > 0 && - ServiceProvider.Azure - ) { - const contentFilterResults = - textmoderation[0]?.content_filter_results; - console.log( - `[${ServiceProvider.Azure}] [Text Moderation] flagged categories result:`, - contentFilterResults, - ); - } - } catch (e) { - console.error("[Request] parse error", text, msg); - } - }, - onclose() { - finish(); - }, - onerror(e) { - options.onError?.(e); - throw e; - }, - openWhenHidden: true, - }); - } else { const res = await fetch(chatPath, chatPayload); clearTimeout(requestTimeoutId); const resJson = await res.json(); - const message = this.extractMessage(resJson); + const message = await this.extractMessage(resJson); options.onFinish(message); } } catch (e) { @@ -404,20 +410,26 @@ export class ChatGPTApi implements LLMApi { }); const resJson = (await res.json()) as OpenAIListModelResponse; - const chatModels = resJson.data?.filter((m) => m.id.startsWith("gpt-")); + const chatModels = resJson.data?.filter( + (m) => m.id.startsWith("gpt-") || m.id.startsWith("chatgpt-"), + ); console.log("[Models]", chatModels); if (!chatModels) { return []; } + //由于目前 OpenAI 的 disableListModels 默认为 true，所以当前实际不会运行到这场 + let seq = 1000; //同 Constant.ts 中的排序保持一致 return chatModels.map((m) => ({ name: m.id, available: true, + sorted: seq++, provider: { id: "openai", providerName: "OpenAI", providerType: "openai", + sorted: 1, }, })); } diff --git a/app/client/platforms/tencent.ts b/app/client/platforms/tencent.ts index e9e49d3f0..579008a9b 100644 --- a/app/client/platforms/tencent.ts +++ b/app/client/platforms/tencent.ts @@ -91,8 +91,9 @@ export class HunyuanApi implements LLMApi { async chat(options: ChatOptions) { const visionModel = isVisionModel(options.config.model); - const messages = options.messages.map((v) => ({ - role: v.role, + const messages = options.messages.map((v, index) => ({ + // "Messages 中 system 角色必须位于列表的最开始" + role: index !== 0 && v.role === "system" ? "user" : v.role, content: visionModel ? v.content : getMessageTextContent(v), })); diff --git a/app/command.ts b/app/command.ts index e515e5f0b..bea4e06f3 100644 --- a/app/command.ts +++ b/app/command.ts @@ -41,13 +41,16 @@ interface ChatCommands { del?: Command; } -export const ChatCommandPrefix = ":"; +// Compatible with Chinese colon character "：" +export const ChatCommandPrefix = /^[:：]/; export function useChatCommand(commands: ChatCommands = {}) { function extract(userInput: string) { - return ( - userInput.startsWith(ChatCommandPrefix) ? userInput.slice(1) : userInput - ) as keyof ChatCommands; + const match = userInput.match(ChatCommandPrefix); + if (match) { + return userInput.slice(1) as keyof ChatCommands; + } + return userInput as keyof ChatCommands; } function search(userInput: string) { @@ -57,7 +60,7 @@ export function useChatCommand(commands: ChatCommands = {}) { .filter((c) => c.startsWith(input)) .map((c) => ({ title: desc[c as keyof ChatCommands], - content: ChatCommandPrefix + c, + content: ":" + c, })); } diff --git a/app/components/artifacts.tsx b/app/components/artifacts.tsx index 326891e73..d725ee659 100644 --- a/app/components/artifacts.tsx +++ b/app/components/artifacts.tsx @@ -1,4 +1,11 @@ -import { useEffect, useState, useRef, useMemo } from "react"; +import { + useEffect, + useState, + useRef, + useMemo, + forwardRef, + useImperativeHandle, +} from "react"; import { useParams } from "react-router"; import { useWindowSize } from "@/app/utils"; import { IconButton } from "./button"; @@ -8,6 +15,7 @@ import CopyIcon from "../icons/copy.svg"; import DownloadIcon from "../icons/download.svg"; import GithubIcon from "../icons/github.svg"; import LoadingButtonIcon from "../icons/loading.svg"; +import ReloadButtonIcon from "../icons/reload.svg"; import Locale from "../locales"; import { Modal, showToast } from "./ui-lib"; import { copyToClipboard, downloadAs } from "../utils"; @@ -15,73 +23,89 @@ import { Path, ApiPath, REPO_URL } from "@/app/constant"; import { Loading } from "./home"; import styles from "./artifacts.module.scss"; -export function HTMLPreview(props: { +type HTMLPreviewProps = { code: string; autoHeight?: boolean; height?: number | string; onLoad?: (title?: string) => void; -}) { - const ref = useRef(null); - const frameId = useRef(nanoid()); - const [iframeHeight, setIframeHeight] = useState(600); - const [title, setTitle] = useState(""); - /* - * https://stackoverflow.com/questions/19739001/what-is-the-difference-between-srcdoc-and-src-datatext-html-in-an - * 1. using srcdoc - * 2. using src with dataurl: - * easy to share - * length limit (Data URIs cannot be larger than 32,768 characters.) - */ +}; - useEffect(() => { - const handleMessage = (e: any) => { - const { id, height, title } = e.data; - setTitle(title); - if (id == frameId.current) { - setIframeHeight(height); +export type HTMLPreviewHander = { + reload: () => void; +}; + +export const HTMLPreview = forwardRef( + function HTMLPreview(props, ref) { + const iframeRef = useRef(null); + const [frameId, setFrameId] = useState(nanoid()); + const [iframeHeight, setIframeHeight] = useState(600); + const [title, setTitle] = useState(""); + /* + * https://stackoverflow.com/questions/19739001/what-is-the-difference-between-srcdoc-and-src-datatext-html-in-an + * 1. using srcdoc + * 2. using src with dataurl: + * easy to share + * length limit (Data URIs cannot be larger than 32,768 characters.) + */ + + useEffect(() => { + const handleMessage = (e: any) => { + const { id, height, title } = e.data; + setTitle(title); + if (id == frameId) { + setIframeHeight(height); + } + }; + window.addEventListener("message", handleMessage); + return () => { + window.removeEventListener("message", handleMessage); + }; + }, [frameId]); + + useImperativeHandle(ref, () => ({ + reload: () => { + setFrameId(nanoid()); + }, + })); + + const height = useMemo(() => { + if (!props.autoHeight) return props.height || 600; + if (typeof props.height === "string") { + return props.height; + } + const parentHeight = props.height || 600; + return iframeHeight + 40 > parentHeight + ? parentHeight + : iframeHeight + 40; + }, [props.autoHeight, props.height, iframeHeight]); + + const srcDoc = useMemo(() => { + const script = ``; + if (props.code.includes("")) { + props.code.replace("", "" + script); + } + return script + props.code; + }, [props.code, frameId]); + + const handleOnLoad = () => { + if (props?.onLoad) { + props.onLoad(title); } }; - window.addEventListener("message", handleMessage); - return () => { - window.removeEventListener("message", handleMessage); - }; - }, []); - const height = useMemo(() => { - if (!props.autoHeight) return props.height || 600; - if (typeof props.height === "string") { - return props.height; - } - const parentHeight = props.height || 600; - return iframeHeight + 40 > parentHeight ? parentHeight : iframeHeight + 40; - }, [props.autoHeight, props.height, iframeHeight]); - - const srcDoc = useMemo(() => { - const script = ``; - if (props.code.includes("")) { - props.code.replace("", "" + script); - } - return props.code + script; - }, [props.code]); - - const handleOnLoad = () => { - if (props?.onLoad) { - props.onLoad(title); - } - }; - - return ( -