Merge branch 'main' into feat/voice-input

[Cherry Pick] Improve [Utils] Check Vision Model

README.md

@@ -25,7 +25,7 @@ One-Click to get a well-designed cross-platform ChatGPT web UI, with GPT3, GPT4
 [MacOS-image]: https://img.shields.io/badge/-MacOS-black?logo=apple
 [Linux-image]: https://img.shields.io/badge/-Linux-333?logo=ubuntu
 
-[![Deploy with Vercel](https://vercel.com/button)](https://vercel.com/new/clone?repository-url=https%3A%2F%2Fgithub.com%2FYidadaa%2FChatGPT-Next-Web&env=OPENAI_API_KEY&env=CODE&env=GOOGLE_API_KEY&project-name=chatgpt-next-web&repository-name=ChatGPT-Next-Web)
+[![Deploy with Vercel](https://vercel.com/button)](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/button.svg)](https://zeabur.com/templates/ZBUEFA)
 

app/components/chat.module.scss

@@ -58,7 +58,7 @@
     box-shadow: var(--card-shadow);
     transition: width ease 0.3s;
     align-items: center;
-    height: 16px;
+    height: 24px;
     width: var(--icon-width);
     overflow: hidden;
 
@@ -68,7 +68,6 @@
 
     .text {
       white-space: nowrap;
-      padding-left: 5px;
       opacity: 0;
       transform: translateX(-5px);
       transition: all ease 0.3s;
@@ -610,10 +609,6 @@
 .chat-input-send {
   background-color: var(--primary);
   color: white;
-
-  position: absolute;
-  right: 30px;
-  bottom: 32px;
 }
 
 @media only screen and (max-width: 600px) {

app/components/chat.tsx

@@ -97,7 +97,7 @@ import { ExportMessageModal } from "./exporter";
 import { getClientConfig } from "../config/client";
 import { useAllModels } from "../utils/hooks";
 import { MultimodalContent } from "../client/api";
-
+import SpeechRecorder from "./chat/speechRecorder";
 const Markdown = dynamic(async () => (await import("./markdown")).Markdown, {
   loading: () => <LoadingIcon />,
 });
@@ -219,6 +219,8 @@ function useSubmitHandler() {
   }, []);
 
   const shouldSubmit = (e: React.KeyboardEvent<HTMLTextAreaElement>) => {
+    // Fix Chinese input method "Enter" on Safari
+    if (e.keyCode == 229) return false;
     if (e.key !== "Enter") return false;
     if (e.key === "Enter" && (e.nativeEvent.isComposing || isComposing.current))
       return false;
@@ -345,7 +347,7 @@ function ChatAction(props: {
     full: 16,
     icon: 16,
   });
-
+  const [isActive, setIsActive] = useState(false);
   function updateWidth() {
     if (!iconRef.current || !textRef.current) return;
     const getWidth = (dom: HTMLDivElement) => dom.getBoundingClientRect().width;
@@ -359,25 +361,22 @@ function ChatAction(props: {
 
   return (
     <div
-      className={`${styles["chat-input-action"]} clickable`}
+      className={`${styles["chat-input-action"]} clickable group`}
       onClick={() => {
         props.onClick();
         setTimeout(updateWidth, 1);
       }}
-      onMouseEnter={updateWidth}
-      onTouchStart={updateWidth}
-      style={
-        {
-          "--icon-width": `${width.icon}px`,
-          "--full-width": `${width.full}px`,
-        } as React.CSSProperties
-      }
     >
-      <div ref={iconRef} className={styles["icon"]}>
-        {props.icon}
-      </div>
-      <div className={styles["text"]} ref={textRef}>
-        {props.text}
+      <div className="flex">
+        <div ref={iconRef} className={styles["icon"]}>
+          {props.icon}
+        </div>
+        <div
+          className={`${styles["text"]} transition-all duration-1000 w-0 group-hover:w-[60px]`}
+          ref={textRef}
+        >
+          {props.text}
+        </div>
       </div>
     </div>
   );
@@ -422,6 +421,7 @@ export function ChatActions(props: {
   showPromptModal: () => void;
   scrollToBottom: () => void;
   showPromptHints: () => void;
+  setUserInput: (text: string) => void;
   hitBottom: boolean;
   uploading: boolean;
 }) {
@@ -1460,6 +1460,7 @@ function _Chat() {
           scrollToBottom={scrollToBottom}
           hitBottom={hitBottom}
           uploading={uploading}
+          setUserInput={setUserInput}
           showPromptHints={() => {
             // Click again to close
             if (promptHints.length > 0) {
@@ -1520,13 +1521,19 @@ function _Chat() {
               })}
             </div>
           )}
-          <IconButton
-            icon={<SendWhiteIcon />}
-            text={Locale.Chat.Send}
-            className={styles["chat-input-send"]}
-            type="primary"
-            onClick={() => doSubmit(userInput)}
-          />
+          <div className="flex gap-2 absolute left-[30px] bottom-[32px]">
+            <SpeechRecorder textUpdater={setUserInput}></SpeechRecorder>
+          </div>
+
+          <div className="flex gap-2 absolute right-[30px] bottom-[32px]">
+            <IconButton
+              icon={<SendWhiteIcon />}
+              text={Locale.Chat.Send}
+              className={styles["chat-input-send"]}
+              type="primary"
+              onClick={() => doSubmit(userInput)}
+            />
+          </div>
         </label>
       </div>
 

app/components/chat/speechRecorder.tsx

@@ -0,0 +1,64 @@
+import React, { useState, useEffect } from "react";
+import VoiceIcon from "@/app/icons/voice.svg";
+import { getLang, formatLang } from "@/app/locales";
+type SpeechRecognitionType =
+  | typeof window.SpeechRecognition
+  | typeof window.webkitSpeechRecognition;
+
+export default function SpeechRecorder({
+  textUpdater,
+  onStop,
+}: {
+  textUpdater: (text: string) => void;
+  onStop?: () => void;
+}) {
+  const [speechRecognition, setSpeechRecognition] =
+    useState<SpeechRecognitionType | null>(null);
+  const [isRecording, setIsRecording] = useState(false);
+  useEffect(() => {
+    if ("SpeechRecognition" in window) {
+      setSpeechRecognition(new (window as any).SpeechRecognition());
+    } else if ("webkitSpeechRecognition" in window) {
+      setSpeechRecognition(new (window as any).webkitSpeechRecognition());
+    }
+  }, []);
+  return (
+    <>
+      {speechRecognition && (
+        <div>
+          <button
+            onClick={() => {
+              if (!isRecording && speechRecognition) {
+                speechRecognition.continuous = true;
+                speechRecognition.lang = formatLang(getLang());
+                console.log(speechRecognition.lang);
+                speechRecognition.interimResults = true;
+                speechRecognition.start();
+                speechRecognition.onresult = function (event: any) {
+                  console.log(event);
+                  var transcript = event.results[0][0].transcript;
+                  console.log(transcript);
+                  textUpdater(transcript);
+                };
+                setIsRecording(true);
+              } else {
+                speechRecognition.stop();
+                setIsRecording(false);
+              }
+            }}
+          >
+            {isRecording ? (
+              <button className="p-2 rounded-full bg-blue-500 hover:bg-blue-600 ring-4 ring-blue-200 transition animate-pulse">
+                <VoiceIcon fill={"white"} />
+              </button>
+            ) : (
+              <button className="p-2 rounded-full bg-zinc-100 hover:bg-zinc-200 transition">
+                <VoiceIcon fill={"#8282A5"} />
+              </button>
+            )}
+          </button>
+        </div>
+      )}
+    </>
+  );
+}

app/icons/voice.svg

@@ -0,0 +1,11 @@
+<svg width="20" height="20" viewBox="0 0 20 20" fill="none" xmlns="http://www.w3.org/2000/svg">
+    <path fill-rule="evenodd" clip-rule="evenodd"
+        d="M5.9375 5.3125C5.9375 3.06884 7.75634 1.25 10 1.25C12.2437 1.25 14.0625 3.06884 14.0625 5.3125V8.75C14.0625 10.9937 12.2437 12.8125 10 12.8125C7.75634 12.8125 5.9375 10.9937 5.9375 8.75V5.3125Z"
+        fill="auto" />
+    <path fill-rule="evenodd" clip-rule="evenodd"
+        d="M3.35938 7.8125C3.79085 7.8125 4.14062 8.16228 4.14062 8.59375V8.75C4.14062 11.986 6.76396 14.6094 10 14.6094C13.236 14.6094 15.8594 11.986 15.8594 8.75V8.59375C15.8594 8.16228 16.2092 7.8125 16.6406 7.8125C17.0721 7.8125 17.4219 8.16228 17.4219 8.59375V8.75C17.4219 12.849 14.099 16.1719 10 16.1719C5.90101 16.1719 2.57812 12.849 2.57812 8.75V8.59375C2.57812 8.16228 2.9279 7.8125 3.35938 7.8125Z"
+        fill="auto" />
+    <path
+        d="M9.21875 15.4688C9.21875 15.0373 9.56853 14.6875 10 14.6875C10.4315 14.6875 10.7812 15.0373 10.7812 15.4688V17.9688C10.7812 18.4002 10.4315 18.75 10 18.75C9.56853 18.75 9.21875 18.4002 9.21875 17.9688V15.4688Z"
+        fill="auto" />
+</svg>

app/locales/index.ts

@@ -70,6 +70,28 @@ export const ALL_LANG_OPTIONS: Record<Lang, string> = {
   sk: "Slovensky",
 };
 
+const LANG_CODE_MAPPING = {
+  cn: "zh-CN",
+  en: "en-US",
+  tw: "zh-TW",
+  pt: "pt-BR",
+  jp: "ja-JP",
+  ko: "ko-KR",
+  id: "id-ID",
+  fr: "fr-FR",
+  es: "es-ES",
+  it: "it-IT",
+  tr: "tr-TR",
+  de: "de-DE",
+  vi: "vi-VN",
+  ru: "ru-RU",
+  cs: "cs-CZ",
+  no: "nb-NO",
+  ar: "ar-SA",
+  bn: "bn-BD",
+  sk: "sk-SK",
+};
+
 const LANG_KEY = "lang";
 const DEFAULT_LANG = "en";
 
@@ -81,6 +103,13 @@ merge(fallbackLang, targetLang);
 
 export default fallbackLang as LocaleType;
 
+export const formatLang = (languageCode: string) => {
+  return (
+    LANG_CODE_MAPPING[languageCode as keyof typeof LANG_CODE_MAPPING] ||
+    languageCode
+  );
+};
+
 function getItem(key: string) {
   try {
     return localStorage.getItem(key);

app/store/sync.ts

@@ -118,7 +118,7 @@ export const useSyncStore = createPersistStore(
   }),
   {
     name: StoreKey.Sync,
-    version: 1.1,
+    version: 1.2,
 
     migrate(persistedState, version) {
       const newState = persistedState as typeof DEFAULT_SYNC_STATE;
@@ -127,6 +127,15 @@ export const useSyncStore = createPersistStore(
         newState.upstash.username = STORAGE_KEY;
       }
 
+      if (version < 1.2) {
+        if (
+          (persistedState as typeof DEFAULT_SYNC_STATE).proxyUrl ===
+          "/api/cors/"
+        ) {
+          newState.proxyUrl = "";
+        }
+      }
+
       return newState as any;
     },
   },

app/styles/globals.scss

@@ -1,3 +1,6 @@
+@tailwind base;
+@tailwind components;
+@tailwind utilities;
 @import "./animation.scss";
 @import "./window.scss";
 

app/utils.ts

@@ -292,9 +292,11 @@ export function getMessageImages(message: RequestMessage): string[] {
 }
 
 export function isVisionModel(model: string) {
-  return (
-    // model.startsWith("gpt-4-vision") ||
-    // model.startsWith("gemini-pro-vision") ||
-    model.includes("vision")
-  );
+  // Note: This is a better way using the TypeScript feature instead of `&&` or `||` (ts v5.5.0-dev.20240314 I've been using)
+  const visionKeywords = [
+    "vision",
+    "claude-3",
+  ];
+
+  return visionKeywords.some(keyword => model.includes(keyword));
 }

app/utils/cors.ts

@@ -4,6 +4,9 @@ import { ApiPath, DEFAULT_API_HOST } from "../constant";
 export function corsPath(path: string) {
   const baseUrl = getClientConfig()?.isApp ? `${DEFAULT_API_HOST}` : "";
 
+  if (baseUrl === "" && path === "") {
+    return "";
+  }
   if (!path.startsWith("/")) {
     path = "/" + path;
   }

package.json

@@ -50,6 +50,7 @@
     "@types/react-dom": "^18.2.7",
     "@types/react-katex": "^3.0.0",
     "@types/spark-md5": "^3.0.4",
+    "autoprefixer": "^10.4.18",
     "cross-env": "^7.0.3",
     "eslint": "^8.49.0",
     "eslint-config-next": "13.4.19",
@@ -57,12 +58,14 @@
     "eslint-plugin-prettier": "^4.2.1",
     "husky": "^8.0.0",
     "lint-staged": "^13.2.2",
+    "postcss": "^8.4.35",
     "prettier": "^3.0.2",
+    "tailwindcss": "^3.4.1",
     "typescript": "5.2.2",
     "webpack": "^5.88.1"
   },
   "resolutions": {
     "lint-staged/yaml": "^2.2.2"
-  },
+  }
   "packageManager": "yarn@1.22.19"
 }

postcss.config.js

@@ -0,0 +1,6 @@
+module.exports = {
+  plugins: {
+    tailwindcss: {},
+    autoprefixer: {},
+  },
+}

tailwind.config.js

@@ -0,0 +1,15 @@
+/** @type {import('tailwindcss').Config} */
+module.exports = {
+  content: [
+    "./app/**/*.{js,ts,jsx,tsx,mdx}",
+    "./pages/**/*.{js,ts,jsx,tsx,mdx}",
+    "./components/**/*.{js,ts,jsx,tsx,mdx}",
+
+    // Or if using `src` directory:
+    "./src/**/*.{js,ts,jsx,tsx,mdx}",
+  ],
+  theme: {
+    extend: {},
+  },
+  plugins: [],
+}

tsconfig.json

@@ -23,6 +23,12 @@
       "@/*": ["./*"]
     }
   },
-  "include": ["next-env.d.ts", "**/*.ts", "**/*.tsx", ".next/types/**/*.ts", "app/calcTextareaHeight.ts"],
+  "include": [
+    "next-env.d.ts",
+    "**/*.ts",
+    "**/*.tsx",
+    ".next/types/**/*.ts",
+    "app/calcTextareaHeight.ts"
+  ],
   "exclude": ["node_modules"]
 }