feat: chat with file function is ready

api/handler/chatimpl/chat_handler.go

@@ -323,20 +323,46 @@ func (h *ChatHandler) sendMessage(ctx context.Context, session *types.ChatSessio
 		reqMgs = append(reqMgs, m)
 	}
 
+	fullPrompt := prompt
+	text := prompt
+	// extract files in prompt
+	files := utils.ExtractFileURLs(prompt)
+	logger.Debugf("detected FILES: %+v", files)
+	if len(files) > 0 {
+		contents := make([]string, 0)
+		var file model.File
+		for _, v := range files {
+			h.DB.Where("url = ?", v).First(&file)
+			content, err := utils.ReadFileContent(v)
+			if err == nil {
+				contents = append(contents, fmt.Sprintf("%s 文件内容：%s", file.Name, content))
+			}
+			text = strings.Replace(text, v, "", 1)
+		}
+		if len(contents) > 0 {
+			fullPrompt = fmt.Sprintf("请根据提供的文件内容信息回答问题(其中Excel 已转成 HTML)：\n\n %s\n\n 问题：%s", strings.Join(contents, "\n"), text)
+		}
+
+		tokens, _ := utils.CalcTokens(fullPrompt, req.Model)
+		if tokens > session.Model.MaxContext {
+			return fmt.Errorf("文件的长度超出模型允许的最大上下文长度，请减少文件内容数量或文件大小。")
+		}
+	}
+	logger.Debug("最终Prompt：", fullPrompt)
+
 	if session.Model.Platform == types.QWen.Value {
 		req.Input = make(map[string]interface{})
 		reqMgs = append(reqMgs, types.Message{
 			Role:    "user",
-			Content: prompt,
+			Content: fullPrompt,
 		})
 		req.Input["messages"] = reqMgs
 	} else if session.Model.Platform == types.OpenAI.Value || session.Model.Platform == types.Azure.Value { // extract image for gpt-vision model
-		imgURLs := utils.ExtractImgURL(prompt)
+		imgURLs := utils.ExtractImgURLs(prompt)
 		logger.Debugf("detected IMG: %+v", imgURLs)
 		var content interface{}
 		if len(imgURLs) > 0 {
 			data := make([]interface{}, 0)
-			text := prompt
 			for _, v := range imgURLs {
 				text = strings.Replace(text, v, "", 1)
 				data = append(data, gin.H{
@@ -352,7 +378,7 @@ func (h *ChatHandler) sendMessage(ctx context.Context, session *types.ChatSessio
 			})
 			content = data
 		} else {
-			content = prompt
+			content = fullPrompt
 		}
 		req.Messages = append(reqMgs, map[string]interface{}{
 			"role":    "user",
@@ -361,7 +387,7 @@ func (h *ChatHandler) sendMessage(ctx context.Context, session *types.ChatSessio
 	} else {
 		req.Messages = append(reqMgs, map[string]interface{}{
 			"role":    "user",
-			"content": prompt,
+			"content": fullPrompt,
 		})
 	}
 
@@ -454,7 +480,7 @@ func (h *ChatHandler) StopGenerate(c *gin.Context) {
 func (h *ChatHandler) doRequest(ctx context.Context, req types.ApiRequest, session *types.ChatSession, apiKey *model.ApiKey) (*http.Response, error) {
 	// if the chat model bind a KEY, use it directly
 	if session.Model.KeyId > 0 {
-		h.DB.Debug().Where("id", session.Model.KeyId).Where("enabled", true).Find(apiKey)
+		h.DB.Where("id", session.Model.KeyId).Find(apiKey)
 	}
 	// use the last unused key
 	if apiKey.Id == 0 {

api/handler/chatimpl/xunfei_handler.go

@@ -79,7 +79,7 @@ func (h *ChatHandler) sendXunFeiMessage(
 	var res *gorm.DB
 	// use the bind key
 	if session.Model.KeyId > 0 {
-		res = h.DB.Where("id", session.Model.KeyId).Where("enabled", true).Find(&apiKey)
+		res = h.DB.Where("id", session.Model.KeyId).Find(&apiKey)
 	}
 	// use the last unused key
 	if apiKey.Id == 0 {

api/handler/markmap_handler.go

@@ -215,7 +215,7 @@ func (h *MarkMapHandler) doRequest(req types.ApiRequest, chatModel model.ChatMod
 	// if the chat model bind a KEY, use it directly
 	var res *gorm.DB
 	if chatModel.KeyId > 0 {
-		res = h.DB.Where("id", chatModel.KeyId).Where("enabled", true).Find(apiKey)
+		res = h.DB.Where("id", chatModel.KeyId).Find(apiKey)
 	}
 	// use the last unused key
 	if apiKey.Id == 0 {

api/test/test.go

@@ -7,7 +7,7 @@ import (
 
 func main() {
 	file := "http://nk.img.r9it.com/chatgpt-plus/1719389335351828.xlsx"
-	content, err := utils.ReadPdf(file)
+	content, err := utils.ReadFileContent(file)
 	if err != nil {
 		panic(err)
 	}

api/utils/file.go

@@ -13,7 +13,8 @@ import (
 	"github.com/google/go-tika/tika"
 )
 
-func ReadPdf(filePath string) (string, error) {
+func ReadFileContent(filePath string) (string, error) {
+	// for remote file, download it first
 	if strings.HasPrefix(filePath, "http") {
 		file, err := downloadFile(filePath)
 		if err != nil {
@@ -31,22 +32,34 @@ func ReadPdf(filePath string) (string, error) {
 	defer file.Close()
 
 	// 使用 Tika 提取 PDF 文件的文本内容
-	html, err := client.Parse(context.TODO(), file)
+	content, err := client.Parse(context.TODO(), file)
 	if err != nil {
 		return "", fmt.Errorf("error with parse file: %v", err)
 	}
 
-	fmt.Println(html)
+	ext := filepath.Ext(filePath)
-
+	switch ext {
-	return cleanBlankLine(html), nil
+	case ".doc", ".docx", ".pdf", ".pptx", "ppt":
+		return cleanBlankLine(cleanHtml(content, false)), nil
+	case ".xls", ".xlsx":
+		return cleanBlankLine(cleanHtml(content, true)), nil
+	default:
+		return cleanBlankLine(content), nil
+	}
 
 }
 
 // 清理文本内容
-func cleanHtml(html string) string {
+func cleanHtml(html string, keepTable bool) string {
 	// 清理 HTML 标签
-	p := bluemonday.StrictPolicy()
+	var policy *bluemonday.Policy
-	return p.Sanitize(html)
+	if keepTable {
+		policy = bluemonday.NewPolicy()
+		policy.AllowElements("table", "thead", "tbody", "tfoot", "tr", "td", "th")
+	} else {
+		policy = bluemonday.StrictPolicy()
+	}
+	return policy.Sanitize(html)
 }
 
 func cleanBlankLine(content string) string {
@@ -57,6 +70,12 @@ func cleanBlankLine(content string) string {
 		if len(line) < 2 {
 			continue
 		}
+		// discard image
+		if strings.HasSuffix(line, ".png") ||
+			strings.HasSuffix(line, ".jpg") ||
+			strings.HasSuffix(line, ".jpeg") {
+			continue
+		}
 		texts = append(texts, line)
 	}
 

api/utils/upload.go

@@ -88,7 +88,7 @@ func GetImgExt(filename string) string {
 	return ext
 }
 
-func ExtractImgURL(text string) []string {
+func ExtractImgURLs(text string) []string {
 	re := regexp.MustCompile(`(http[s]?:\/\/.*?\.(?:png|jpg|jpeg|gif))`)
 	matches := re.FindAllStringSubmatch(text, 10)
 	urls := make([]string, 0)
@@ -99,3 +99,15 @@ func ExtractImgURL(text string) []string {
 	}
 	return urls
 }
+
+func ExtractFileURLs(text string) []string {
+	re := regexp.MustCompile(`(http[s]?:\/\/.*?\.(?:docx?|pdf|pptx?|xlsx?|txt))`)
+	matches := re.FindAllStringSubmatch(text, 10)
+	urls := make([]string, 0)
+	if len(matches) > 0 {
+		for _, m := range matches {
+			urls = append(urls, m[1])
+		}
+	}
+	return urls
+}

web/babel.config.js

@@ -1,5 +1,5 @@
 module.exports = {
   presets: [
-    '@vue/cli-plugin-babel/preset'
+    '@vue.css/cli-plugin-babel/preset'
   ]
 }

web/src/assets/css/markdown/vue.css

@@ -0,0 +1,237 @@
+.chat-line {
+    ol, ul {
+        margin: 0.8em 0;
+        list-style: normal;
+    }
+    a {
+        color: #42b983;
+        font-weight: 600;
+        padding: 0 2px;
+        text-decoration: none;
+    }
+
+    h1,
+    h2,
+    h3,
+    h4,
+    h5,
+    h6 {
+        position: relative;
+        margin-top: 1rem;
+        margin-bottom: 1rem;
+        font-weight: bold;
+        line-height: 1.4;
+        cursor: text;
+    }
+
+    h1:hover a.anchor,
+    h2:hover a.anchor,
+    h3:hover a.anchor,
+    h4:hover a.anchor,
+    h5:hover a.anchor,
+    h6:hover a.anchor {
+        text-decoration: none;
+    }
+
+    h1 tt,
+    h1 code {
+        font-size: inherit !important;
+    }
+
+    h2 tt,
+    h2 code {
+        font-size: inherit !important;
+    }
+
+    h3 tt,
+    h3 code {
+        font-size: inherit !important;
+    }
+
+    h4 tt,
+    h4 code {
+        font-size: inherit !important;
+    }
+
+    h5 tt,
+    h5 code {
+        font-size: inherit !important;
+    }
+
+    h6 tt,
+    h6 code {
+        font-size: inherit !important;
+    }
+
+    h2 a,
+    h3 a {
+        color: #34495e;
+    }
+
+    h1 {
+        padding-bottom: .4rem;
+        font-size: 2.2rem;
+        line-height: 1.3;
+    }
+
+    h2 {
+        font-size: 1.75rem;
+        line-height: 1.225;
+        margin: 35px 0 15px;
+        padding-bottom: 0.5em;
+        border-bottom: 1px solid #ddd;
+    }
+
+    h3 {
+        font-size: 1.4rem;
+        line-height: 1.43;
+        margin: 20px 0 7px;
+    }
+
+    h4 {
+        font-size: 1.2rem;
+    }
+
+    h5 {
+        font-size: 1rem;
+    }
+
+    h6 {
+        font-size: 1rem;
+        color: #777;
+    }
+
+    p,
+    blockquote,
+    ul,
+    ol,
+    dl,
+    table {
+        margin: 0.8em 0;
+    }
+
+    li > ol,
+    li > ul {
+        margin: 0 0;
+    }
+
+    hr {
+        height: 2px;
+        padding: 0;
+        margin: 16px 0;
+        background-color: #e7e7e7;
+        border: 0 none;
+        overflow: hidden;
+        box-sizing: content-box;
+    }
+
+    body > h2:first-child {
+        margin-top: 0;
+        padding-top: 0;
+    }
+
+    body > h1:first-child {
+        margin-top: 0;
+        padding-top: 0;
+    }
+
+    body > h1:first-child + h2 {
+        margin-top: 0;
+        padding-top: 0;
+    }
+
+    body > h3:first-child,
+    body > h4:first-child,
+    body > h5:first-child,
+    body > h6:first-child {
+        margin-top: 0;
+        padding-top: 0;
+    }
+
+    a:first-child h1,
+    a:first-child h2,
+    a:first-child h3,
+    a:first-child h4,
+    a:first-child h5,
+    a:first-child h6 {
+        margin-top: 0;
+        padding-top: 0;
+    }
+
+    h1 p,
+    h2 p,
+    h3 p,
+    h4 p,
+    h5 p,
+    h6 p {
+        margin-top: 0;
+    }
+
+    li p.first {
+        display: inline-block;
+    }
+
+    ul,
+    ol {
+        padding-left: 30px;
+    }
+
+    ul:first-child,
+    ol:first-child {
+        margin-top: 0;
+    }
+
+    ul:last-child,
+    ol:last-child {
+        margin-bottom: 0;
+    }
+
+    blockquote {
+        border-left: 4px solid #42b983;
+        padding: 10px 15px;
+        color: #777;
+        background-color: rgba(66, 185, 131, .1);
+    }
+
+    table {
+        padding: 0;
+        word-break: initial;
+    }
+
+    table tr {
+        border-top: 1px solid #dfe2e5;
+        margin: 0;
+        padding: 0;
+    }
+
+    table tr:nth-child(2n),
+    thead {
+        background-color: #fafafa;
+    }
+
+    table tr th {
+        font-weight: bold;
+        border: 1px solid #dfe2e5;
+        border-bottom: 0;
+        text-align: left;
+        margin: 0;
+        padding: 6px 13px;
+    }
+
+    table tr td {
+        border: 1px solid #dfe2e5;
+        text-align: left;
+        margin: 0;
+        padding: 6px 13px;
+    }
+
+    table tr th:first-child,
+    table tr td:first-child {
+        margin-top: 0;
+    }
+
+    table tr th:last-child,
+    table tr td:last-child {
+        margin-bottom: 0;
+    }
+}

web/src/components/ChatPrompt.vue

@@ -16,7 +16,9 @@
                 <el-image :src="GetFileIcon(file.ext)" fit="cover"  />
               </div>
               <div class="body">
-                <div class="title">{{file.name}}</div>
+                <div class="title">
+                  <el-link :href="file.url" target="_blank" style="--el-font-weight-primary:bold">{{file.name}}</el-link>
+                </div>
                 <div class="info">
                   <span>{{GetFileType(file.ext)}}</span>
                   <span>{{FormatFileSize(file.size)}}</span>
@@ -121,6 +123,7 @@ onMounted(() => {
 </script>
 
 <style lang="stylus">
+@import '@/assets/css/markdown/vue.css';
 .chat-line-prompt {
   background-color #ffffff;
   justify-content: center;
@@ -214,11 +217,6 @@ onMounted(() => {
           margin 10px 0
         }
 
-
-        a {
-          color #20a0ff
-        }
-
         p {
           line-height 1.5
         }

web/src/components/ChatReply.vue

@@ -97,6 +97,7 @@ const reGenerate = (prompt) => {
 </script>
 
 <style lang="stylus">
+@import '@/assets/css/markdown/vue.css';
 .common-layout {
   .chat-line-reply {
     justify-content: center;
@@ -132,18 +133,12 @@ const reGenerate = (prompt) => {
         .content {
           min-height 20px;
           word-break break-word;
-          padding: 6px 10px;
+          padding: 0 10px;
           color #374151;
           font-size: var(--content-font-size);
           border-radius: 5px;
           overflow auto;
 
-          a {
-            color #20a0ff
-          }
-
-          // control the image size in content
-
           img {
             max-width: 600px;
             border-radius: 10px;
@@ -170,10 +165,11 @@ const reGenerate = (prompt) => {
 
           .code-container {
             position relative
+            display flex
 
             .hljs {
               border-radius 10px
-              line-height 1.5
+              width 100%
             }
 
             .copy-code-btn {
@@ -194,7 +190,7 @@ const reGenerate = (prompt) => {
           .lang-name {
             position absolute;
             right 10px
-            bottom 50px
+            bottom 20px
             padding 2px 6px 4px 6px
             background-color #444444
             border-radius 10px

web/src/components/FileSelect.vue

@@ -20,6 +20,7 @@
                   :auto-upload="true"
                   :show-file-list="false"
                   :http-request="afterRead"
+                  accept=".doc,.docx,.jpg,.png,.jpeg,.xls,.xlsx,.ppt,.pptx,.pdf"
               >
                 <el-icon class="avatar-uploader-icon">
                   <Plus/>

web/src/utils/libs.js

@@ -188,38 +188,7 @@ export function processContent(content) {
             }
         }
     }
-
+    return content
-    const lines = content.split("\n")
-    if (lines.length <= 1) {
-        return content
-    }
-    const texts = []
-    // 定义匹配数学公式的正则表达式
-    const formulaRegex = /^\s*[a-z|A-Z]+[^=]+\s*=\s*[^=]+$/;
-    let hasCode = false
-    for (let i = 0; i < lines.length; i++) {
-        // 处理引用块换行
-        if (lines[i].startsWith(">")) {
-            texts.push(lines[i])
-            texts.push("\n")
-            continue
-        }
-        // 如果包含代码块则跳过公式检测
-        if (lines[i].indexOf("```") !== -1) {
-            texts.push(lines[i])
-            hasCode = true
-            continue
-        }
-        // 识别并处理数学公式，需要排除那些已经被识别出来的公式
-        if (i > 0 && formulaRegex.test(lines[i]) && lines[i - 1].indexOf("$$") === -1 && !hasCode) {
-            texts.push("$$")
-            texts.push(lines[i])
-            texts.push("$$")
-            continue
-        }
-        texts.push(lines[i])
-    }
-    return texts.join("\n")
 }
 
 export function processPrompt(prompt) {

web/src/views/ChatPlus.vue

@@ -24,7 +24,7 @@
           <div class="content" :style="{height: leftBoxHeight+'px'}">
             <el-row v-for="chat in chatList" :key="chat.chat_id">
               <div :class="chat.chat_id === activeChat.chat_id?'chat-list-item active':'chat-list-item'"
-                   @click="changeChat(chat)">
+                   @click="loadChat(chat)">
                 <el-image :src="chat.icon" class="avatar"/>
                 <span class="chat-title-input" v-if="chat.edit">
                   <el-input v-model="tmpChatTitle" size="small" @keydown="titleKeydown($event, chat)"
@@ -424,12 +424,8 @@ const newChat = () => {
   connect(null, roleId.value)
 }
 
-// 切换会话
-const changeChat = (chat) => {
-  localStorage.setItem("chat_id", chat.chat_id)
-  loadChat(chat)
-}
 
+// 切换会话
 const loadChat = function (chat) {
   if (!isLogin.value) {
     store.setShowLoginDialog(true)
@@ -753,7 +749,7 @@ const sendMessage = function () {
   // 如果携带了文件，则串上文件地址
   let content = prompt.value
   if (files.value.length > 0) {
-    content = files.value.map(file => file.url).join(" ") + " " +  content
+    content += files.value.map(file => file.url).join(" ")
   }
   // 追加消息
   chatData.value.push({

web/src/views/admin/ChatModel.vue

@@ -162,7 +162,7 @@
         </el-form-item>
 
         <el-form-item label="绑定API-KEY：" prop="apikey">
-          <el-select v-model="item.key_id" placeholder="请选择 API KEY" clearable>
+          <el-select v-model="item.key_id" placeholder="请选择 API KEY" filterable clearable>
             <el-option v-for="v in apiKeys" :value="v.id" :label="v.name" :key="v.id">
               {{ v.name }}
               <el-text type="info" size="small">{{ substr(v.api_url, 50) }}</el-text>
@@ -229,7 +229,7 @@ const platforms = ref([])
 
 // 获取 API KEY
 const apiKeys = ref([])
-httpGet('/api/admin/apikey/list?status=true&type=chat').then(res => {
+httpGet('/api/admin/apikey/list?type=chat').then(res => {
   apiKeys.value = res.data
 }).catch(e => {
   ElMessage.error("获取 API KEY 失败：" + e.message)