mirror of
https://github.com/yangjian102621/geekai.git
synced 2025-09-17 16:56:38 +08:00
316 lines
8.9 KiB
Go
316 lines
8.9 KiB
Go
package chatimpl
|
||
|
||
// * +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
|
||
// * Copyright 2023 The Geek-AI Authors. All rights reserved.
|
||
// * Use of this source code is governed by a Apache-2.0 license
|
||
// * that can be found in the LICENSE file.
|
||
// * @Author yangjian102621@163.com
|
||
// * +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
|
||
|
||
import (
|
||
"bufio"
|
||
"context"
|
||
"encoding/json"
|
||
"errors"
|
||
"fmt"
|
||
"html/template"
|
||
"io"
|
||
"net/http"
|
||
"strings"
|
||
"time"
|
||
"unicode/utf8"
|
||
|
||
"github.com/levigross/grequests"
|
||
|
||
"geekai/core/types"
|
||
"geekai/store/model"
|
||
"geekai/store/vo"
|
||
"geekai/utils"
|
||
)
|
||
|
||
type ollamaResp struct {
|
||
Id string `json:"id"`
|
||
Model string `json:"model"`
|
||
|
||
CreatedAt string `json:"created_at"`
|
||
Response string `json:"response"`
|
||
Done bool `json:"done"`
|
||
DoneReason string `json:"done_reason"`
|
||
Context []int `json:"context"`
|
||
|
||
TotalDuration int64 `json:"total_duration"` // 生成响应所花费的总时间
|
||
LoadDuration int64 `json:"load_duration"` // 以纳秒为单位加载模型所花费的时间
|
||
PromptEvalCount int `json:"prompt_eval_count"` // 提示文本中的标记（tokens）数量
|
||
PromptEvalDuration int64 `json:"prompt_eval_duration"` // 以纳秒为单位评估提示文本所花费的时间
|
||
EvalCount int64 `json:"eval_count"` // 生成响应中的标记数量
|
||
EvalDuration int64 `json:"eval_duration"` // 以纳秒为单位生成响应所花费的时间
|
||
}
|
||
|
||
// 通义千问消息发送实现
|
||
func (h *ChatHandler) sendOllamaMessage(
|
||
chatCtx []types.Message,
|
||
req types.ApiRequest,
|
||
userVo vo.User,
|
||
ctx context.Context,
|
||
session *types.ChatSession,
|
||
role model.ChatRole,
|
||
prompt string,
|
||
ws *types.WsClient) error {
|
||
|
||
promptCreatedAt := time.Now() // 记录提问时间
|
||
start := time.Now()
|
||
|
||
//var apiKey = model.ApiKey{}
|
||
//response, err := h.doRequest(ctx, req, session, &apiKey)
|
||
response, err := h.sendOllamaRequest(session, prompt)
|
||
defer response.Body.Close()
|
||
|
||
logger.Info("HTTP请求完成，耗时：", time.Now().Sub(start))
|
||
|
||
if err != nil {
|
||
h.processError(err, prompt, ws)
|
||
}
|
||
|
||
contentType := response.Header.Get("Content-Type")
|
||
if strings.Contains(contentType, "application/x-ndjson") {
|
||
|
||
h.processOllamaStreamResponse(chatCtx, req, userVo, response, ws, prompt, session, role, promptCreatedAt)
|
||
|
||
} else {
|
||
if err = h.processOllamaJsonResponse(response, ws); err != nil {
|
||
return err
|
||
}
|
||
}
|
||
|
||
return nil
|
||
}
|
||
|
||
func (h *ChatHandler) sendOllamaRequest(session *types.ChatSession, prompt string) (*http.Response, error) {
|
||
apiKey, err := h.queryApiKey(session)
|
||
if err != nil {
|
||
return nil, err
|
||
}
|
||
|
||
// todo add context to request body
|
||
postData := map[string]interface{}{
|
||
"model": session.Model.Value,
|
||
"stream": true,
|
||
"prompt": prompt,
|
||
}
|
||
headers := map[string]string{
|
||
"Content-Type": "application/json",
|
||
"Authorization": "Bearer " + apiKey.Value,
|
||
}
|
||
|
||
ro := &grequests.RequestOptions{
|
||
JSON: postData,
|
||
Headers: headers,
|
||
}
|
||
resp, err := grequests.Post(apiKey.ApiURL, ro)
|
||
if err != nil {
|
||
return nil, err
|
||
}
|
||
|
||
if !resp.Ok {
|
||
return nil, resp.Error
|
||
}
|
||
|
||
return resp.RawResponse, nil
|
||
}
|
||
|
||
func (h *ChatHandler) queryApiKey(session *types.ChatSession) (*model.ApiKey, error) {
|
||
apiKey := &model.ApiKey{}
|
||
|
||
// 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)
|
||
}
|
||
// use the last unused key
|
||
if apiKey.Id == 0 {
|
||
h.DB.Debug().Where("platform", session.Model.Platform).Where("type", "chat").Where("enabled", true).Order("last_used_at ASC").First(apiKey)
|
||
}
|
||
if apiKey.Id == 0 {
|
||
return nil, errors.New("no available key, please import key")
|
||
}
|
||
|
||
h.DB.Model(apiKey).UpdateColumn("last_used_at", time.Now().Unix())
|
||
|
||
return apiKey, nil
|
||
}
|
||
|
||
func (h *ChatHandler) processOllamaStreamResponse(
|
||
chatCtx []types.Message, req types.ApiRequest, userVo vo.User,
|
||
response *http.Response, ws *types.WsClient, prompt string,
|
||
session *types.ChatSession, role model.ChatRole, promptCreatedAt time.Time) {
|
||
|
||
// 记录回复时间
|
||
replyCreatedAt := time.Now()
|
||
// 循环读取 Chunk 消息
|
||
var message = types.Message{}
|
||
scanner := bufio.NewScanner(response.Body)
|
||
|
||
var content string
|
||
var replyTokens int
|
||
|
||
for scanner.Scan() {
|
||
var resp ollamaResp
|
||
line := scanner.Text()
|
||
|
||
err := utils.JsonDecode(line, &resp)
|
||
if err != nil {
|
||
logger.Error("error with parse data line: ", content)
|
||
utils.ReplyMessage(ws, fmt.Sprintf("**解析数据行失败：%s**", err))
|
||
break
|
||
}
|
||
|
||
if resp.Done == true && resp.DoneReason == "stop" {
|
||
utils.ReplyChunkMessage(ws, types.WsMessage{Type: types.WsEnd})
|
||
message.Content = utils.InterfaceToString(resp.Context)
|
||
replyTokens = resp.PromptEvalCount
|
||
|
||
// 消息发送成功后做记录工作
|
||
h.recordInfoAfterSendMessage(chatCtx, req, userVo, message, prompt, session, role, promptCreatedAt, replyTokens, replyCreatedAt)
|
||
|
||
break
|
||
} else if resp.Done == true && resp.DoneReason != "stop" {
|
||
utils.ReplyMessage(ws, fmt.Sprintf("**API 返回错误：%s**", resp.DoneReason))
|
||
break
|
||
}
|
||
|
||
if len(resp.Id) > 0 {
|
||
utils.ReplyChunkMessage(ws, types.WsMessage{Type: types.WsStart})
|
||
}
|
||
|
||
if len(resp.Response) > 0 {
|
||
utils.ReplyChunkMessage(ws, types.WsMessage{
|
||
Type: types.WsMiddle,
|
||
Content: utils.InterfaceToString(resp.Response),
|
||
})
|
||
}
|
||
|
||
}
|
||
|
||
if err := scanner.Err(); err != nil {
|
||
if strings.Contains(err.Error(), "context canceled") {
|
||
logger.Info("用户取消了请求：", prompt)
|
||
} else {
|
||
logger.Error("信息读取出错：", err)
|
||
}
|
||
}
|
||
}
|
||
|
||
func (h *ChatHandler) processOllamaJsonResponse(response *http.Response, ws *types.WsClient) error {
|
||
body, err := io.ReadAll(response.Body)
|
||
if err != nil {
|
||
return fmt.Errorf("error with reading response: %v", err)
|
||
}
|
||
|
||
var res struct {
|
||
Code int `json:"error_code"`
|
||
Msg string `json:"error_msg"`
|
||
}
|
||
err = json.Unmarshal(body, &res)
|
||
if err != nil {
|
||
return fmt.Errorf("error with decode response: %v", err)
|
||
}
|
||
utils.ReplyMessage(ws, "请求Ollama大模型 API 失败："+res.Msg)
|
||
return nil
|
||
}
|
||
|
||
func (h *ChatHandler) recordInfoAfterSendMessage(chatCtx []types.Message, req types.ApiRequest, userVo vo.User, message types.Message, prompt string, session *types.ChatSession, role model.ChatRole, promptCreatedAt time.Time, replyTokens int, replyCreatedAt time.Time) {
|
||
if message.Role == "" {
|
||
message.Role = "assistant"
|
||
}
|
||
|
||
useMsg := types.Message{Role: "user", Content: prompt}
|
||
|
||
// 更新上下文消息，如果是调用函数则不需要更新上下文
|
||
if h.App.SysConfig.EnableContext {
|
||
chatCtx = append(chatCtx, useMsg) // 提问消息
|
||
chatCtx = append(chatCtx, message) // 回复消息
|
||
h.App.ChatContexts.Put(session.ChatId, chatCtx)
|
||
}
|
||
|
||
// 追加聊天记录
|
||
// for prompt
|
||
promptToken, err := utils.CalcTokens(prompt, req.Model)
|
||
if err != nil {
|
||
logger.Error(err)
|
||
}
|
||
historyUserMsg := model.ChatMessage{
|
||
UserId: userVo.Id,
|
||
ChatId: session.ChatId,
|
||
RoleId: role.Id,
|
||
Type: types.PromptMsg,
|
||
Icon: userVo.Avatar,
|
||
Content: template.HTMLEscapeString(prompt),
|
||
Tokens: promptToken,
|
||
UseContext: true,
|
||
Model: req.Model,
|
||
}
|
||
historyUserMsg.CreatedAt = promptCreatedAt
|
||
historyUserMsg.UpdatedAt = promptCreatedAt
|
||
res := h.DB.Save(&historyUserMsg)
|
||
if res.Error != nil {
|
||
logger.Error("failed to save prompt history message: ", res.Error)
|
||
}
|
||
|
||
// for reply
|
||
// 计算本次对话消耗的总 token 数量
|
||
//totalTokens := replyTokens + getTotalTokens(req)
|
||
// todo rebuild the tokens
|
||
historyReplyMsg := model.ChatMessage{
|
||
UserId: userVo.Id,
|
||
ChatId: session.ChatId,
|
||
RoleId: role.Id,
|
||
Type: types.ReplyMsg,
|
||
Icon: role.Icon,
|
||
Content: message.Content,
|
||
Tokens: replyTokens,
|
||
UseContext: true,
|
||
Model: req.Model,
|
||
}
|
||
historyReplyMsg.CreatedAt = replyCreatedAt
|
||
historyReplyMsg.UpdatedAt = replyCreatedAt
|
||
res = h.DB.Create(&historyReplyMsg)
|
||
if res.Error != nil {
|
||
logger.Error("failed to save reply history message: ", res.Error)
|
||
}
|
||
|
||
// 更新用户算力
|
||
h.subUserPower(userVo, session, promptToken, replyTokens)
|
||
|
||
// 保存当前会话
|
||
var chatItem model.ChatItem
|
||
res = h.DB.Where("chat_id = ?", session.ChatId).First(&chatItem)
|
||
if res.Error != nil {
|
||
chatItem.ChatId = session.ChatId
|
||
chatItem.UserId = session.UserId
|
||
chatItem.RoleId = role.Id
|
||
chatItem.ModelId = session.Model.Id
|
||
if utf8.RuneCountInString(prompt) > 30 {
|
||
chatItem.Title = string([]rune(prompt)[:30]) + "..."
|
||
} else {
|
||
chatItem.Title = prompt
|
||
}
|
||
chatItem.Model = req.Model
|
||
h.DB.Create(&chatItem)
|
||
}
|
||
}
|
||
|
||
func (h *ChatHandler) processError(err error, prompt string, ws *types.WsClient) {
|
||
if strings.Contains(err.Error(), "context canceled") {
|
||
logger.Info("用户取消了请求：", prompt)
|
||
return
|
||
} else if strings.Contains(err.Error(), "no available key") {
|
||
utils.ReplyMessage(ws, "抱歉😔😔😔，系统已经没有可用的 API KEY，请联系管理员！")
|
||
return
|
||
} else {
|
||
logger.Error(err)
|
||
}
|
||
|
||
utils.ReplyMessage(ws, ErrorMsg)
|
||
utils.ReplyMessage(ws, ErrImg)
|
||
return
|
||
}
|