mirror of
https://github.com/linux-do/new-api.git
synced 2025-11-07 22:53:41 +08:00
fix: try to fix tencent hunyuan #336
This commit is contained in:
@@ -24,7 +24,7 @@ func UpdateTaskBulk() {
|
|||||||
//imageModel := "midjourney"
|
//imageModel := "midjourney"
|
||||||
for {
|
for {
|
||||||
time.Sleep(time.Duration(15) * time.Second)
|
time.Sleep(time.Duration(15) * time.Second)
|
||||||
common.SysLog("任务进度轮询开始")
|
common.SysLog(" 任务进度轮询开始")
|
||||||
ctx := context.TODO()
|
ctx := context.TODO()
|
||||||
allTasks := model.GetAllUnFinishSyncTasks(500)
|
allTasks := model.GetAllUnFinishSyncTasks(500)
|
||||||
platformTask := make(map[constant.TaskPlatform][]*model.Task)
|
platformTask := make(map[constant.TaskPlatform][]*model.Task)
|
||||||
|
|||||||
@@ -6,18 +6,26 @@ import (
|
|||||||
"github.com/gin-gonic/gin"
|
"github.com/gin-gonic/gin"
|
||||||
"io"
|
"io"
|
||||||
"net/http"
|
"net/http"
|
||||||
|
"one-api/common"
|
||||||
"one-api/dto"
|
"one-api/dto"
|
||||||
"one-api/relay/channel"
|
"one-api/relay/channel"
|
||||||
relaycommon "one-api/relay/common"
|
relaycommon "one-api/relay/common"
|
||||||
"one-api/service"
|
"one-api/service"
|
||||||
|
"strconv"
|
||||||
"strings"
|
"strings"
|
||||||
)
|
)
|
||||||
|
|
||||||
type Adaptor struct {
|
type Adaptor struct {
|
||||||
Sign string
|
Sign string
|
||||||
|
Action string
|
||||||
|
Version string
|
||||||
|
Timestamp int64
|
||||||
}
|
}
|
||||||
|
|
||||||
func (a *Adaptor) Init(info *relaycommon.RelayInfo, request dto.GeneralOpenAIRequest) {
|
func (a *Adaptor) Init(info *relaycommon.RelayInfo, request dto.GeneralOpenAIRequest) {
|
||||||
|
a.Action = "ChatCompletions"
|
||||||
|
a.Version = "2023-09-01"
|
||||||
|
a.Timestamp = common.GetTimestamp()
|
||||||
}
|
}
|
||||||
|
|
||||||
func (a *Adaptor) GetRequestURL(info *relaycommon.RelayInfo) (string, error) {
|
func (a *Adaptor) GetRequestURL(info *relaycommon.RelayInfo) (string, error) {
|
||||||
@@ -27,7 +35,9 @@ func (a *Adaptor) GetRequestURL(info *relaycommon.RelayInfo) (string, error) {
|
|||||||
func (a *Adaptor) SetupRequestHeader(c *gin.Context, req *http.Request, info *relaycommon.RelayInfo) error {
|
func (a *Adaptor) SetupRequestHeader(c *gin.Context, req *http.Request, info *relaycommon.RelayInfo) error {
|
||||||
channel.SetupApiRequestHeader(info, c, req)
|
channel.SetupApiRequestHeader(info, c, req)
|
||||||
req.Header.Set("Authorization", a.Sign)
|
req.Header.Set("Authorization", a.Sign)
|
||||||
req.Header.Set("X-TC-Action", info.UpstreamModelName)
|
req.Header.Set("X-TC-Action", a.Action)
|
||||||
|
req.Header.Set("X-TC-Version", a.Version)
|
||||||
|
req.Header.Set("X-TC-Timestamp", strconv.FormatInt(a.Timestamp, 10))
|
||||||
return nil
|
return nil
|
||||||
}
|
}
|
||||||
|
|
||||||
@@ -37,15 +47,13 @@ func (a *Adaptor) ConvertRequest(c *gin.Context, relayMode int, request *dto.Gen
|
|||||||
}
|
}
|
||||||
apiKey := c.Request.Header.Get("Authorization")
|
apiKey := c.Request.Header.Get("Authorization")
|
||||||
apiKey = strings.TrimPrefix(apiKey, "Bearer ")
|
apiKey = strings.TrimPrefix(apiKey, "Bearer ")
|
||||||
appId, secretId, secretKey, err := parseTencentConfig(apiKey)
|
_, secretId, secretKey, err := parseTencentConfig(apiKey)
|
||||||
if err != nil {
|
if err != nil {
|
||||||
return nil, err
|
return nil, err
|
||||||
}
|
}
|
||||||
tencentRequest := requestOpenAI2Tencent(*request)
|
tencentRequest := requestOpenAI2Tencent(*request)
|
||||||
tencentRequest.AppId = appId
|
|
||||||
tencentRequest.SecretId = secretId
|
|
||||||
// we have to calculate the sign here
|
// we have to calculate the sign here
|
||||||
a.Sign = getTencentSign(*tencentRequest, secretKey)
|
a.Sign = getTencentSign(*tencentRequest, a, secretId, secretKey)
|
||||||
return tencentRequest, nil
|
return tencentRequest, nil
|
||||||
}
|
}
|
||||||
|
|
||||||
|
|||||||
@@ -1,62 +1,71 @@
|
|||||||
package tencent
|
package tencent
|
||||||
|
|
||||||
import "one-api/dto"
|
|
||||||
|
|
||||||
type TencentMessage struct {
|
type TencentMessage struct {
|
||||||
Role string `json:"role"`
|
Role string `json:"Role"`
|
||||||
Content string `json:"content"`
|
Content string `json:"Content"`
|
||||||
}
|
}
|
||||||
|
|
||||||
type TencentChatRequest struct {
|
type TencentChatRequest struct {
|
||||||
AppId int64 `json:"app_id"` // 腾讯云账号的 APPID
|
// 模型名称,可选值包括 hunyuan-lite、hunyuan-standard、hunyuan-standard-256K、hunyuan-pro。
|
||||||
SecretId string `json:"secret_id"` // 官网 SecretId
|
// 各模型介绍请阅读 [产品概述](https://cloud.tencent.com/document/product/1729/104753) 中的说明。
|
||||||
// Timestamp当前 UNIX 时间戳,单位为秒,可记录发起 API 请求的时间。
|
//
|
||||||
// 例如1529223702,如果与当前时间相差过大,会引起签名过期错误
|
// 注意:
|
||||||
Timestamp int64 `json:"timestamp"`
|
// 不同的模型计费不同,请根据 [购买指南](https://cloud.tencent.com/document/product/1729/97731) 按需调用。
|
||||||
// Expired 签名的有效期,是一个符合 UNIX Epoch 时间戳规范的数值,
|
Model *string `json:"Model"`
|
||||||
// 单位为秒;Expired 必须大于 Timestamp 且 Expired-Timestamp 小于90天
|
// 聊天上下文信息。
|
||||||
Expired int64 `json:"expired"`
|
// 说明:
|
||||||
QueryID string `json:"query_id"` //请求 Id,用于问题排查
|
// 1. 长度最多为 40,按对话时间从旧到新在数组中排列。
|
||||||
// Temperature 较高的数值会使输出更加随机,而较低的数值会使其更加集中和确定
|
// 2. Message.Role 可选值:system、user、assistant。
|
||||||
// 默认 1.0,取值区间为[0.0,2.0],非必要不建议使用,不合理的取值会影响效果
|
// 其中,system 角色可选,如存在则必须位于列表的最开始。user 和 assistant 需交替出现(一问一答),以 user 提问开始和结束,且 Content 不能为空。Role 的顺序示例:[system(可选) user assistant user assistant user ...]。
|
||||||
// 建议该参数和 top_p 只设置1个,不要同时更改 top_p
|
// 3. Messages 中 Content 总长度不能超过模型输入长度上限(可参考 [产品概述](https://cloud.tencent.com/document/product/1729/104753) 文档),超过则会截断最前面的内容,只保留尾部内容。
|
||||||
Temperature float64 `json:"temperature"`
|
Messages []*TencentMessage `json:"Messages"`
|
||||||
// TopP 影响输出文本的多样性,取值越大,生成文本的多样性越强
|
// 流式调用开关。
|
||||||
// 默认1.0,取值区间为[0.0, 1.0],非必要不建议使用, 不合理的取值会影响效果
|
// 说明:
|
||||||
// 建议该参数和 temperature 只设置1个,不要同时更改
|
// 1. 未传值时默认为非流式调用(false)。
|
||||||
TopP float64 `json:"top_p"`
|
// 2. 流式调用时以 SSE 协议增量返回结果(返回值取 Choices[n].Delta 中的值,需要拼接增量数据才能获得完整结果)。
|
||||||
// Stream 0:同步,1:流式 (默认,协议:SSE)
|
// 3. 非流式调用时:
|
||||||
// 同步请求超时:60s,如果内容较长建议使用流式
|
// 调用方式与普通 HTTP 请求无异。
|
||||||
Stream int `json:"stream"`
|
// 接口响应耗时较长,**如需更低时延建议设置为 true**。
|
||||||
// Messages 会话内容, 长度最多为40, 按对话时间从旧到新在数组中排列
|
// 只返回一次最终结果(返回值取 Choices[n].Message 中的值)。
|
||||||
// 输入 content 总数最大支持 3000 token。
|
//
|
||||||
Messages []TencentMessage `json:"messages"`
|
// 注意:
|
||||||
Model string `json:"model"` // 模型名称
|
// 通过 SDK 调用时,流式和非流式调用需用**不同的方式**获取返回值,具体参考 SDK 中的注释或示例(在各语言 SDK 代码仓库的 examples/hunyuan/v20230901/ 目录中)。
|
||||||
|
Stream *bool `json:"Stream"`
|
||||||
|
// 说明:
|
||||||
|
// 1. 影响输出文本的多样性,取值越大,生成文本的多样性越强。
|
||||||
|
// 2. 取值区间为 [0.0, 1.0],未传值时使用各模型推荐值。
|
||||||
|
// 3. 非必要不建议使用,不合理的取值会影响效果。
|
||||||
|
TopP *float64 `json:"TopP"`
|
||||||
|
// 说明:
|
||||||
|
// 1. 较高的数值会使输出更加随机,而较低的数值会使其更加集中和确定。
|
||||||
|
// 2. 取值区间为 [0.0, 2.0],未传值时使用各模型推荐值。
|
||||||
|
// 3. 非必要不建议使用,不合理的取值会影响效果。
|
||||||
|
Temperature *float64 `json:"Temperature"`
|
||||||
}
|
}
|
||||||
|
|
||||||
type TencentError struct {
|
type TencentError struct {
|
||||||
Code int `json:"code"`
|
Code int `json:"Code"`
|
||||||
Message string `json:"message"`
|
Message string `json:"Message"`
|
||||||
}
|
}
|
||||||
|
|
||||||
type TencentUsage struct {
|
type TencentUsage struct {
|
||||||
InputTokens int `json:"input_tokens"`
|
PromptTokens int `json:"PromptTokens"`
|
||||||
OutputTokens int `json:"output_tokens"`
|
CompletionTokens int `json:"CompletionTokens"`
|
||||||
TotalTokens int `json:"total_tokens"`
|
TotalTokens int `json:"TotalTokens"`
|
||||||
}
|
}
|
||||||
|
|
||||||
type TencentResponseChoices struct {
|
type TencentResponseChoices struct {
|
||||||
FinishReason string `json:"finish_reason,omitempty"` // 流式结束标志位,为 stop 则表示尾包
|
FinishReason string `json:"FinishReason,omitempty"` // 流式结束标志位,为 stop 则表示尾包
|
||||||
Messages TencentMessage `json:"messages,omitempty"` // 内容,同步模式返回内容,流模式为 null 输出 content 内容总数最多支持 1024token。
|
Messages TencentMessage `json:"Message,omitempty"` // 内容,同步模式返回内容,流模式为 null 输出 content 内容总数最多支持 1024token。
|
||||||
Delta TencentMessage `json:"delta,omitempty"` // 内容,流模式返回内容,同步模式为 null 输出 content 内容总数最多支持 1024token。
|
Delta TencentMessage `json:"Delta,omitempty"` // 内容,流模式返回内容,同步模式为 null 输出 content 内容总数最多支持 1024token。
|
||||||
}
|
}
|
||||||
|
|
||||||
type TencentChatResponse struct {
|
type TencentChatResponse struct {
|
||||||
Choices []TencentResponseChoices `json:"choices,omitempty"` // 结果
|
Choices []TencentResponseChoices `json:"Choices,omitempty"` // 结果
|
||||||
Created string `json:"created,omitempty"` // unix 时间戳的字符串
|
Created int64 `json:"Created,omitempty"` // unix 时间戳的字符串
|
||||||
Id string `json:"id,omitempty"` // 会话 id
|
Id string `json:"Id,omitempty"` // 会话 id
|
||||||
Usage dto.Usage `json:"usage,omitempty"` // token 数量
|
Usage TencentUsage `json:"Usage,omitempty"` // token 数量
|
||||||
Error TencentError `json:"error,omitempty"` // 错误信息 注意:此字段可能返回 null,表示取不到有效值
|
Error TencentError `json:"Error,omitempty"` // 错误信息 注意:此字段可能返回 null,表示取不到有效值
|
||||||
Note string `json:"note,omitempty"` // 注释
|
Note string `json:"Note,omitempty"` // 注释
|
||||||
ReqID string `json:"req_id,omitempty"` // 唯一请求 Id,每次请求都会返回。用于反馈接口入参
|
ReqID string `json:"Req_id,omitempty"` // 唯一请求 Id,每次请求都会返回。用于反馈接口入参
|
||||||
}
|
}
|
||||||
|
|||||||
@@ -3,8 +3,8 @@ package tencent
|
|||||||
import (
|
import (
|
||||||
"bufio"
|
"bufio"
|
||||||
"crypto/hmac"
|
"crypto/hmac"
|
||||||
"crypto/sha1"
|
"crypto/sha256"
|
||||||
"encoding/base64"
|
"encoding/hex"
|
||||||
"encoding/json"
|
"encoding/json"
|
||||||
"errors"
|
"errors"
|
||||||
"fmt"
|
"fmt"
|
||||||
@@ -15,46 +15,28 @@ import (
|
|||||||
"one-api/dto"
|
"one-api/dto"
|
||||||
relaycommon "one-api/relay/common"
|
relaycommon "one-api/relay/common"
|
||||||
"one-api/service"
|
"one-api/service"
|
||||||
"sort"
|
|
||||||
"strconv"
|
"strconv"
|
||||||
"strings"
|
"strings"
|
||||||
|
"time"
|
||||||
)
|
)
|
||||||
|
|
||||||
// https://cloud.tencent.com/document/product/1729/97732
|
// https://cloud.tencent.com/document/product/1729/97732
|
||||||
|
|
||||||
func requestOpenAI2Tencent(request dto.GeneralOpenAIRequest) *TencentChatRequest {
|
func requestOpenAI2Tencent(request dto.GeneralOpenAIRequest) *TencentChatRequest {
|
||||||
messages := make([]TencentMessage, 0, len(request.Messages))
|
messages := make([]*TencentMessage, 0, len(request.Messages))
|
||||||
for i := 0; i < len(request.Messages); i++ {
|
for i := 0; i < len(request.Messages); i++ {
|
||||||
message := request.Messages[i]
|
message := request.Messages[i]
|
||||||
if message.Role == "system" {
|
messages = append(messages, &TencentMessage{
|
||||||
messages = append(messages, TencentMessage{
|
|
||||||
Role: "user",
|
|
||||||
Content: message.StringContent(),
|
|
||||||
})
|
|
||||||
messages = append(messages, TencentMessage{
|
|
||||||
Role: "assistant",
|
|
||||||
Content: "Okay",
|
|
||||||
})
|
|
||||||
continue
|
|
||||||
}
|
|
||||||
messages = append(messages, TencentMessage{
|
|
||||||
Content: message.StringContent(),
|
Content: message.StringContent(),
|
||||||
Role: message.Role,
|
Role: message.Role,
|
||||||
})
|
})
|
||||||
}
|
}
|
||||||
stream := 0
|
|
||||||
if request.Stream {
|
|
||||||
stream = 1
|
|
||||||
}
|
|
||||||
return &TencentChatRequest{
|
return &TencentChatRequest{
|
||||||
Timestamp: common.GetTimestamp(),
|
Temperature: &request.Temperature,
|
||||||
Expired: common.GetTimestamp() + 24*60*60,
|
TopP: &request.TopP,
|
||||||
QueryID: common.GetUUID(),
|
Stream: &request.Stream,
|
||||||
Temperature: request.Temperature,
|
|
||||||
TopP: request.TopP,
|
|
||||||
Stream: stream,
|
|
||||||
Messages: messages,
|
Messages: messages,
|
||||||
Model: request.Model,
|
Model: &request.Model,
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
@@ -62,7 +44,11 @@ func responseTencent2OpenAI(response *TencentChatResponse) *dto.OpenAITextRespon
|
|||||||
fullTextResponse := dto.OpenAITextResponse{
|
fullTextResponse := dto.OpenAITextResponse{
|
||||||
Object: "chat.completion",
|
Object: "chat.completion",
|
||||||
Created: common.GetTimestamp(),
|
Created: common.GetTimestamp(),
|
||||||
Usage: response.Usage,
|
Usage: dto.Usage{
|
||||||
|
PromptTokens: response.Usage.PromptTokens,
|
||||||
|
CompletionTokens: response.Usage.CompletionTokens,
|
||||||
|
TotalTokens: response.Usage.TotalTokens,
|
||||||
|
},
|
||||||
}
|
}
|
||||||
if len(response.Choices) > 0 {
|
if len(response.Choices) > 0 {
|
||||||
content, _ := json.Marshal(response.Choices[0].Messages.Content)
|
content, _ := json.Marshal(response.Choices[0].Messages.Content)
|
||||||
@@ -99,64 +85,46 @@ func streamResponseTencent2OpenAI(TencentResponse *TencentChatResponse) *dto.Cha
|
|||||||
func tencentStreamHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIErrorWithStatusCode, string) {
|
func tencentStreamHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIErrorWithStatusCode, string) {
|
||||||
var responseText string
|
var responseText string
|
||||||
scanner := bufio.NewScanner(resp.Body)
|
scanner := bufio.NewScanner(resp.Body)
|
||||||
scanner.Split(func(data []byte, atEOF bool) (advance int, token []byte, err error) {
|
scanner.Split(bufio.ScanLines)
|
||||||
if atEOF && len(data) == 0 {
|
|
||||||
return 0, nil, nil
|
|
||||||
}
|
|
||||||
if i := strings.Index(string(data), "\n"); i >= 0 {
|
|
||||||
return i + 1, data[0:i], nil
|
|
||||||
}
|
|
||||||
if atEOF {
|
|
||||||
return len(data), data, nil
|
|
||||||
}
|
|
||||||
return 0, nil, nil
|
|
||||||
})
|
|
||||||
dataChan := make(chan string)
|
|
||||||
stopChan := make(chan bool)
|
|
||||||
go func() {
|
|
||||||
for scanner.Scan() {
|
|
||||||
data := scanner.Text()
|
|
||||||
if len(data) < 5 { // ignore blank line or wrong format
|
|
||||||
continue
|
|
||||||
}
|
|
||||||
if data[:5] != "data:" {
|
|
||||||
continue
|
|
||||||
}
|
|
||||||
data = data[5:]
|
|
||||||
dataChan <- data
|
|
||||||
}
|
|
||||||
stopChan <- true
|
|
||||||
}()
|
|
||||||
service.SetEventStreamHeaders(c)
|
service.SetEventStreamHeaders(c)
|
||||||
c.Stream(func(w io.Writer) bool {
|
|
||||||
select {
|
for scanner.Scan() {
|
||||||
case data := <-dataChan:
|
data := scanner.Text()
|
||||||
var TencentResponse TencentChatResponse
|
if len(data) < 5 || !strings.HasPrefix(data, "data:") {
|
||||||
err := json.Unmarshal([]byte(data), &TencentResponse)
|
continue
|
||||||
if err != nil {
|
|
||||||
common.SysError("error unmarshalling stream response: " + err.Error())
|
|
||||||
return true
|
|
||||||
}
|
|
||||||
response := streamResponseTencent2OpenAI(&TencentResponse)
|
|
||||||
if len(response.Choices) != 0 {
|
|
||||||
responseText += response.Choices[0].Delta.GetContentString()
|
|
||||||
}
|
|
||||||
jsonResponse, err := json.Marshal(response)
|
|
||||||
if err != nil {
|
|
||||||
common.SysError("error marshalling stream response: " + err.Error())
|
|
||||||
return true
|
|
||||||
}
|
|
||||||
c.Render(-1, common.CustomEvent{Data: "data: " + string(jsonResponse)})
|
|
||||||
return true
|
|
||||||
case <-stopChan:
|
|
||||||
c.Render(-1, common.CustomEvent{Data: "data: [DONE]"})
|
|
||||||
return false
|
|
||||||
}
|
}
|
||||||
})
|
data = strings.TrimPrefix(data, "data:")
|
||||||
|
|
||||||
|
var tencentResponse TencentChatResponse
|
||||||
|
err := json.Unmarshal([]byte(data), &tencentResponse)
|
||||||
|
if err != nil {
|
||||||
|
common.SysError("error unmarshalling stream response: " + err.Error())
|
||||||
|
continue
|
||||||
|
}
|
||||||
|
|
||||||
|
response := streamResponseTencent2OpenAI(&tencentResponse)
|
||||||
|
if len(response.Choices) != 0 {
|
||||||
|
responseText += response.Choices[0].Delta.GetContentString()
|
||||||
|
}
|
||||||
|
|
||||||
|
err = service.ObjectData(c, response)
|
||||||
|
if err != nil {
|
||||||
|
common.SysError(err.Error())
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
|
if err := scanner.Err(); err != nil {
|
||||||
|
common.SysError("error reading stream: " + err.Error())
|
||||||
|
}
|
||||||
|
|
||||||
|
service.Done(c)
|
||||||
|
|
||||||
err := resp.Body.Close()
|
err := resp.Body.Close()
|
||||||
if err != nil {
|
if err != nil {
|
||||||
return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), ""
|
return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), ""
|
||||||
}
|
}
|
||||||
|
|
||||||
return nil, responseText
|
return nil, responseText
|
||||||
}
|
}
|
||||||
|
|
||||||
@@ -206,29 +174,62 @@ func parseTencentConfig(config string) (appId int64, secretId string, secretKey
|
|||||||
return
|
return
|
||||||
}
|
}
|
||||||
|
|
||||||
func getTencentSign(req TencentChatRequest, secretKey string) string {
|
func sha256hex(s string) string {
|
||||||
params := make([]string, 0)
|
b := sha256.Sum256([]byte(s))
|
||||||
params = append(params, "app_id="+strconv.FormatInt(req.AppId, 10))
|
return hex.EncodeToString(b[:])
|
||||||
params = append(params, "secret_id="+req.SecretId)
|
}
|
||||||
params = append(params, "timestamp="+strconv.FormatInt(req.Timestamp, 10))
|
|
||||||
params = append(params, "query_id="+req.QueryID)
|
func hmacSha256(s, key string) string {
|
||||||
params = append(params, "temperature="+strconv.FormatFloat(req.Temperature, 'f', -1, 64))
|
hashed := hmac.New(sha256.New, []byte(key))
|
||||||
params = append(params, "top_p="+strconv.FormatFloat(req.TopP, 'f', -1, 64))
|
hashed.Write([]byte(s))
|
||||||
params = append(params, "stream="+strconv.Itoa(req.Stream))
|
return string(hashed.Sum(nil))
|
||||||
params = append(params, "expired="+strconv.FormatInt(req.Expired, 10))
|
}
|
||||||
|
|
||||||
var messageStr string
|
func getTencentSign(req TencentChatRequest, adaptor *Adaptor, secId, secKey string) string {
|
||||||
for _, msg := range req.Messages {
|
// build canonical request string
|
||||||
messageStr += fmt.Sprintf(`{"role":"%s","content":"%s"},`, msg.Role, msg.Content)
|
host := "hunyuan.tencentcloudapi.com"
|
||||||
}
|
httpRequestMethod := "POST"
|
||||||
messageStr = strings.TrimSuffix(messageStr, ",")
|
canonicalURI := "/"
|
||||||
params = append(params, "messages=["+messageStr+"]")
|
canonicalQueryString := ""
|
||||||
|
canonicalHeaders := fmt.Sprintf("content-type:%s\nhost:%s\nx-tc-action:%s\n",
|
||||||
sort.Sort(sort.StringSlice(params))
|
"application/json", host, strings.ToLower(adaptor.Action))
|
||||||
url := "hunyuan.cloud.tencent.com/hyllm/v1/chat/completions?" + strings.Join(params, "&")
|
signedHeaders := "content-type;host;x-tc-action"
|
||||||
mac := hmac.New(sha1.New, []byte(secretKey))
|
payload, _ := json.Marshal(req)
|
||||||
signURL := url
|
hashedRequestPayload := sha256hex(string(payload))
|
||||||
mac.Write([]byte(signURL))
|
canonicalRequest := fmt.Sprintf("%s\n%s\n%s\n%s\n%s\n%s",
|
||||||
sign := mac.Sum([]byte(nil))
|
httpRequestMethod,
|
||||||
return base64.StdEncoding.EncodeToString(sign)
|
canonicalURI,
|
||||||
|
canonicalQueryString,
|
||||||
|
canonicalHeaders,
|
||||||
|
signedHeaders,
|
||||||
|
hashedRequestPayload)
|
||||||
|
// build string to sign
|
||||||
|
algorithm := "TC3-HMAC-SHA256"
|
||||||
|
requestTimestamp := strconv.FormatInt(adaptor.Timestamp, 10)
|
||||||
|
timestamp, _ := strconv.ParseInt(requestTimestamp, 10, 64)
|
||||||
|
t := time.Unix(timestamp, 0).UTC()
|
||||||
|
// must be the format 2006-01-02, ref to package time for more info
|
||||||
|
date := t.Format("2006-01-02")
|
||||||
|
credentialScope := fmt.Sprintf("%s/%s/tc3_request", date, "hunyuan")
|
||||||
|
hashedCanonicalRequest := sha256hex(canonicalRequest)
|
||||||
|
string2sign := fmt.Sprintf("%s\n%s\n%s\n%s",
|
||||||
|
algorithm,
|
||||||
|
requestTimestamp,
|
||||||
|
credentialScope,
|
||||||
|
hashedCanonicalRequest)
|
||||||
|
|
||||||
|
// sign string
|
||||||
|
secretDate := hmacSha256(date, "TC3"+secKey)
|
||||||
|
secretService := hmacSha256("hunyuan", secretDate)
|
||||||
|
secretKey := hmacSha256("tc3_request", secretService)
|
||||||
|
signature := hex.EncodeToString([]byte(hmacSha256(string2sign, secretKey)))
|
||||||
|
|
||||||
|
// build authorization
|
||||||
|
authorization := fmt.Sprintf("%s Credential=%s/%s, SignedHeaders=%s, Signature=%s",
|
||||||
|
algorithm,
|
||||||
|
secId,
|
||||||
|
credentialScope,
|
||||||
|
signedHeaders,
|
||||||
|
signature)
|
||||||
|
return authorization
|
||||||
}
|
}
|
||||||
|
|||||||
@@ -1,6 +1,12 @@
|
|||||||
package service
|
package service
|
||||||
|
|
||||||
import "github.com/gin-gonic/gin"
|
import (
|
||||||
|
"encoding/json"
|
||||||
|
"fmt"
|
||||||
|
"github.com/gin-gonic/gin"
|
||||||
|
"one-api/common"
|
||||||
|
"strings"
|
||||||
|
)
|
||||||
|
|
||||||
func SetEventStreamHeaders(c *gin.Context) {
|
func SetEventStreamHeaders(c *gin.Context) {
|
||||||
c.Writer.Header().Set("Content-Type", "text/event-stream")
|
c.Writer.Header().Set("Content-Type", "text/event-stream")
|
||||||
@@ -9,3 +15,23 @@ func SetEventStreamHeaders(c *gin.Context) {
|
|||||||
c.Writer.Header().Set("Transfer-Encoding", "chunked")
|
c.Writer.Header().Set("Transfer-Encoding", "chunked")
|
||||||
c.Writer.Header().Set("X-Accel-Buffering", "no")
|
c.Writer.Header().Set("X-Accel-Buffering", "no")
|
||||||
}
|
}
|
||||||
|
|
||||||
|
func StringData(c *gin.Context, str string) {
|
||||||
|
str = strings.TrimPrefix(str, "data: ")
|
||||||
|
str = strings.TrimSuffix(str, "\r")
|
||||||
|
c.Render(-1, common.CustomEvent{Data: "data: " + str})
|
||||||
|
c.Writer.Flush()
|
||||||
|
}
|
||||||
|
|
||||||
|
func ObjectData(c *gin.Context, object interface{}) error {
|
||||||
|
jsonData, err := json.Marshal(object)
|
||||||
|
if err != nil {
|
||||||
|
return fmt.Errorf("error marshalling object: %w", err)
|
||||||
|
}
|
||||||
|
StringData(c, string(jsonData))
|
||||||
|
return nil
|
||||||
|
}
|
||||||
|
|
||||||
|
func Done(c *gin.Context) {
|
||||||
|
StringData(c, "[DONE]")
|
||||||
|
}
|
||||||
|
|||||||
Reference in New Issue
Block a user