refactor: Introduce pre-consume quota and unify relay handlers
This commit introduces a major architectural refactoring to improve quota management, centralize logging, and streamline the relay handling logic. Key changes: - **Pre-consume Quota:** Implements a new mechanism to check and reserve user quota *before* making the request to the upstream provider. This ensures more accurate quota deduction and prevents users from exceeding their limits due to concurrent requests. - **Unified Relay Handlers:** Refactors the relay logic to use generic handlers (e.g., `ChatHandler`, `ImageHandler`) instead of provider-specific implementations. This significantly reduces code duplication and simplifies adding new channels. - **Centralized Logger:** A new dedicated `logger` package is introduced, and all system logging calls are migrated to use it, moving this responsibility out of the `common` package. - **Code Reorganization:** DTOs are generalized (e.g., `dalle.go` -> `openai_image.go`) and utility code is moved to more appropriate packages (e.g., `common/http.go` -> `service/http.go`) for better code structure.
This commit is contained in:
CaIon
@@ -2,7 +2,6 @@ package relay
|
||||
|
||||
import (
|
||||
"bytes"
|
||||
"errors"
|
||||
"fmt"
|
||||
"io"
|
||||
"net/http"
|
||||
@@ -18,68 +17,26 @@ import (
|
||||
"github.com/gin-gonic/gin"
|
||||
)
|
||||
|
||||
func getAndValidateClaudeRequest(c *gin.Context) (textRequest *dto.ClaudeRequest, err error) {
|
||||
textRequest = &dto.ClaudeRequest{}
|
||||
err = c.ShouldBindJSON(textRequest)
|
||||
if err != nil {
|
||||
return nil, err
|
||||
}
|
||||
if textRequest.Messages == nil || len(textRequest.Messages) == 0 {
|
||||
return nil, errors.New("field messages is required")
|
||||
}
|
||||
if textRequest.Model == "" {
|
||||
return nil, errors.New("field model is required")
|
||||
}
|
||||
return textRequest, nil
|
||||
}
|
||||
func ClaudeHelper(c *gin.Context, info *relaycommon.RelayInfo) (newAPIError *types.NewAPIError) {
|
||||
|
||||
func ClaudeHelper(c *gin.Context) (newAPIError *types.NewAPIError) {
|
||||
info.InitChannelMeta(c)
|
||||
|
||||
relayInfo := relaycommon.GenRelayInfoClaude(c)
|
||||
textRequest, ok := info.Request.(*dto.ClaudeRequest)
|
||||
|
||||
// get & validate textRequest 获取并验证文本请求
|
||||
textRequest, err := getAndValidateClaudeRequest(c)
|
||||
if err != nil {
|
||||
return types.NewError(err, types.ErrorCodeInvalidRequest, types.ErrOptionWithSkipRetry())
|
||||
if !ok {
|
||||
common.FatalLog(fmt.Sprintf("invalid request type, expected dto.ClaudeRequest, got %T", info.Request))
|
||||
}
|
||||
|
||||
if textRequest.Stream {
|
||||
relayInfo.IsStream = true
|
||||
}
|
||||
|
||||
err = helper.ModelMappedHelper(c, relayInfo, textRequest)
|
||||
err := helper.ModelMappedHelper(c, info, textRequest)
|
||||
if err != nil {
|
||||
return types.NewError(err, types.ErrorCodeChannelModelMappedError, types.ErrOptionWithSkipRetry())
|
||||
}
|
||||
|
||||
promptTokens, err := getClaudePromptTokens(textRequest, relayInfo)
|
||||
// count messages token error 计算promptTokens错误
|
||||
if err != nil {
|
||||
return types.NewError(err, types.ErrorCodeCountTokenFailed, types.ErrOptionWithSkipRetry())
|
||||
}
|
||||
|
||||
priceData, err := helper.ModelPriceHelper(c, relayInfo, promptTokens, int(textRequest.MaxTokens))
|
||||
if err != nil {
|
||||
return types.NewError(err, types.ErrorCodeModelPriceError, types.ErrOptionWithSkipRetry())
|
||||
}
|
||||
|
||||
// pre-consume quota 预消耗配额
|
||||
preConsumedQuota, userQuota, newAPIError := preConsumeQuota(c, priceData.ShouldPreConsumedQuota, relayInfo)
|
||||
|
||||
if newAPIError != nil {
|
||||
return newAPIError
|
||||
}
|
||||
defer func() {
|
||||
if newAPIError != nil {
|
||||
returnPreConsumedQuota(c, relayInfo, userQuota, preConsumedQuota)
|
||||
}
|
||||
}()
|
||||
|
||||
adaptor := GetAdaptor(relayInfo.ApiType)
|
||||
adaptor := GetAdaptor(info.ApiType)
|
||||
if adaptor == nil {
|
||||
return types.NewError(fmt.Errorf("invalid api type: %d", relayInfo.ApiType), types.ErrorCodeInvalidApiType, types.ErrOptionWithSkipRetry())
|
||||
return types.NewError(fmt.Errorf("invalid api type: %d", info.ApiType), types.ErrorCodeInvalidApiType, types.ErrOptionWithSkipRetry())
|
||||
}
|
||||
adaptor.Init(relayInfo)
|
||||
adaptor.Init(info)
|
||||
|
||||
if textRequest.MaxTokens == 0 {
|
||||
textRequest.MaxTokens = uint(model_setting.GetClaudeSettings().GetDefaultMaxTokens(textRequest.Model))
|
||||
@@ -104,18 +61,18 @@ func ClaudeHelper(c *gin.Context) (newAPIError *types.NewAPIError) {
|
||||
textRequest.Temperature = common.GetPointer[float64](1.0)
|
||||
}
|
||||
textRequest.Model = strings.TrimSuffix(textRequest.Model, "-thinking")
|
||||
relayInfo.UpstreamModelName = textRequest.Model
|
||||
info.UpstreamModelName = textRequest.Model
|
||||
}
|
||||
|
||||
var requestBody io.Reader
|
||||
if model_setting.GetGlobalSettings().PassThroughRequestEnabled || relayInfo.ChannelSetting.PassThroughBodyEnabled {
|
||||
if model_setting.GetGlobalSettings().PassThroughRequestEnabled || info.ChannelSetting.PassThroughBodyEnabled {
|
||||
body, err := common.GetRequestBody(c)
|
||||
if err != nil {
|
||||
return types.NewErrorWithStatusCode(err, types.ErrorCodeReadRequestBodyFailed, http.StatusBadRequest, types.ErrOptionWithSkipRetry())
|
||||
}
|
||||
requestBody = bytes.NewBuffer(body)
|
||||
} else {
|
||||
convertedRequest, err := adaptor.ConvertClaudeRequest(c, relayInfo, textRequest)
|
||||
convertedRequest, err := adaptor.ConvertClaudeRequest(c, info, textRequest)
|
||||
if err != nil {
|
||||
return types.NewError(err, types.ErrorCodeConvertRequestFailed, types.ErrOptionWithSkipRetry())
|
||||
}
|
||||
@@ -125,10 +82,10 @@ func ClaudeHelper(c *gin.Context) (newAPIError *types.NewAPIError) {
|
||||
}
|
||||
|
||||
// apply param override
|
||||
if len(relayInfo.ParamOverride) > 0 {
|
||||
if len(info.ParamOverride) > 0 {
|
||||
reqMap := make(map[string]interface{})
|
||||
_ = common.Unmarshal(jsonData, &reqMap)
|
||||
for key, value := range relayInfo.ParamOverride {
|
||||
for key, value := range info.ParamOverride {
|
||||
reqMap[key] = value
|
||||
}
|
||||
jsonData, err = common.Marshal(reqMap)
|
||||
@@ -145,14 +102,14 @@ func ClaudeHelper(c *gin.Context) (newAPIError *types.NewAPIError) {
|
||||
|
||||
statusCodeMappingStr := c.GetString("status_code_mapping")
|
||||
var httpResp *http.Response
|
||||
resp, err := adaptor.DoRequest(c, relayInfo, requestBody)
|
||||
resp, err := adaptor.DoRequest(c, info, requestBody)
|
||||
if err != nil {
|
||||
return types.NewOpenAIError(err, types.ErrorCodeDoRequestFailed, http.StatusInternalServerError)
|
||||
}
|
||||
|
||||
if resp != nil {
|
||||
httpResp = resp.(*http.Response)
|
||||
relayInfo.IsStream = relayInfo.IsStream || strings.HasPrefix(httpResp.Header.Get("Content-Type"), "text/event-stream")
|
||||
info.IsStream = info.IsStream || strings.HasPrefix(httpResp.Header.Get("Content-Type"), "text/event-stream")
|
||||
if httpResp.StatusCode != http.StatusOK {
|
||||
newAPIError = service.RelayErrorHandler(httpResp, false)
|
||||
// reset status code 重置状态码
|
||||
@@ -161,24 +118,14 @@ func ClaudeHelper(c *gin.Context) (newAPIError *types.NewAPIError) {
|
||||
}
|
||||
}
|
||||
|
||||
usage, newAPIError := adaptor.DoResponse(c, httpResp, relayInfo)
|
||||
usage, newAPIError := adaptor.DoResponse(c, httpResp, info)
|
||||
//log.Printf("usage: %v", usage)
|
||||
if newAPIError != nil {
|
||||
// reset status code 重置状态码
|
||||
service.ResetStatusCode(newAPIError, statusCodeMappingStr)
|
||||
return newAPIError
|
||||
}
|
||||
service.PostClaudeConsumeQuota(c, relayInfo, usage.(*dto.Usage), preConsumedQuota, userQuota, priceData, "")
|
||||
|
||||
service.PostClaudeConsumeQuota(c, info, usage.(*dto.Usage))
|
||||
return nil
|
||||
}
|
||||
|
||||
func getClaudePromptTokens(textRequest *dto.ClaudeRequest, info *relaycommon.RelayInfo) (int, error) {
|
||||
var promptTokens int
|
||||
var err error
|
||||
switch info.RelayMode {
|
||||
default:
|
||||
promptTokens, err = service.CountTokenClaudeRequest(*textRequest, info.UpstreamModelName)
|
||||
}
|
||||
info.PromptTokens = promptTokens
|
||||
return promptTokens, err
|
||||
}
|
||||
|
||||
Reference in New Issue
Block a user