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
@@ -3,7 +3,6 @@ package relay
|
||||
import (
|
||||
"bytes"
|
||||
"encoding/json"
|
||||
"errors"
|
||||
"fmt"
|
||||
"io"
|
||||
"net/http"
|
||||
@@ -12,7 +11,6 @@ import (
|
||||
relaycommon "one-api/relay/common"
|
||||
"one-api/relay/helper"
|
||||
"one-api/service"
|
||||
"one-api/setting"
|
||||
"one-api/setting/model_setting"
|
||||
"one-api/types"
|
||||
"strings"
|
||||
@@ -20,82 +18,24 @@ import (
|
||||
"github.com/gin-gonic/gin"
|
||||
)
|
||||
|
||||
func getAndValidateResponsesRequest(c *gin.Context) (*dto.OpenAIResponsesRequest, error) {
|
||||
request := &dto.OpenAIResponsesRequest{}
|
||||
err := common.UnmarshalBodyReusable(c, request)
|
||||
if err != nil {
|
||||
return nil, err
|
||||
}
|
||||
if request.Model == "" {
|
||||
return nil, errors.New("model is required")
|
||||
}
|
||||
if len(request.Input) == 0 {
|
||||
return nil, errors.New("input is required")
|
||||
}
|
||||
return request, nil
|
||||
func ResponsesHelper(c *gin.Context, info *relaycommon.RelayInfo) (newAPIError *types.NewAPIError) {
|
||||
info.InitChannelMeta(c)
|
||||
|
||||
}
|
||||
|
||||
func checkInputSensitive(textRequest *dto.OpenAIResponsesRequest, info *relaycommon.RelayInfo) ([]string, error) {
|
||||
sensitiveWords, err := service.CheckSensitiveInput(textRequest.Input)
|
||||
return sensitiveWords, err
|
||||
}
|
||||
|
||||
func getInputTokens(req *dto.OpenAIResponsesRequest, info *relaycommon.RelayInfo) int {
|
||||
inputTokens := service.CountTokenInput(req.Input, req.Model)
|
||||
info.PromptTokens = inputTokens
|
||||
return inputTokens
|
||||
}
|
||||
|
||||
func ResponsesHelper(c *gin.Context) (newAPIError *types.NewAPIError) {
|
||||
req, err := getAndValidateResponsesRequest(c)
|
||||
if err != nil {
|
||||
common.LogError(c, fmt.Sprintf("getAndValidateResponsesRequest error: %s", err.Error()))
|
||||
return types.NewError(err, types.ErrorCodeInvalidRequest, types.ErrOptionWithSkipRetry())
|
||||
request, ok := info.Request.(*dto.OpenAIResponsesRequest)
|
||||
if !ok {
|
||||
common.FatalLog(fmt.Sprintf("invalid request type, expected dto.OpenAIResponsesRequest, got %T", info.Request))
|
||||
}
|
||||
|
||||
relayInfo := relaycommon.GenRelayInfoResponses(c, req)
|
||||
|
||||
if setting.ShouldCheckPromptSensitive() {
|
||||
sensitiveWords, err := checkInputSensitive(req, relayInfo)
|
||||
if err != nil {
|
||||
common.LogWarn(c, fmt.Sprintf("user sensitive words detected: %s", strings.Join(sensitiveWords, ", ")))
|
||||
return types.NewError(err, types.ErrorCodeSensitiveWordsDetected, types.ErrOptionWithSkipRetry())
|
||||
}
|
||||
}
|
||||
|
||||
err = helper.ModelMappedHelper(c, relayInfo, req)
|
||||
err := helper.ModelMappedHelper(c, info, request)
|
||||
if err != nil {
|
||||
return types.NewError(err, types.ErrorCodeChannelModelMappedError, types.ErrOptionWithSkipRetry())
|
||||
}
|
||||
|
||||
if value, exists := c.Get("prompt_tokens"); exists {
|
||||
promptTokens := value.(int)
|
||||
relayInfo.SetPromptTokens(promptTokens)
|
||||
} else {
|
||||
promptTokens := getInputTokens(req, relayInfo)
|
||||
c.Set("prompt_tokens", promptTokens)
|
||||
}
|
||||
|
||||
priceData, err := helper.ModelPriceHelper(c, relayInfo, relayInfo.PromptTokens, int(req.MaxOutputTokens))
|
||||
if err != nil {
|
||||
return types.NewError(err, types.ErrorCodeModelPriceError, types.ErrOptionWithSkipRetry())
|
||||
}
|
||||
// pre consume quota
|
||||
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)
|
||||
var requestBody io.Reader
|
||||
if model_setting.GetGlobalSettings().PassThroughRequestEnabled {
|
||||
body, err := common.GetRequestBody(c)
|
||||
@@ -104,7 +44,7 @@ func ResponsesHelper(c *gin.Context) (newAPIError *types.NewAPIError) {
|
||||
}
|
||||
requestBody = bytes.NewBuffer(body)
|
||||
} else {
|
||||
convertedRequest, err := adaptor.ConvertOpenAIResponsesRequest(c, relayInfo, *req)
|
||||
convertedRequest, err := adaptor.ConvertOpenAIResponsesRequest(c, info, *request)
|
||||
if err != nil {
|
||||
return types.NewError(err, types.ErrorCodeConvertRequestFailed, types.ErrOptionWithSkipRetry())
|
||||
}
|
||||
@@ -113,13 +53,13 @@ func ResponsesHelper(c *gin.Context) (newAPIError *types.NewAPIError) {
|
||||
return types.NewError(err, types.ErrorCodeConvertRequestFailed, types.ErrOptionWithSkipRetry())
|
||||
}
|
||||
// apply param override
|
||||
if len(relayInfo.ParamOverride) > 0 {
|
||||
if len(info.ParamOverride) > 0 {
|
||||
reqMap := make(map[string]interface{})
|
||||
err = json.Unmarshal(jsonData, &reqMap)
|
||||
if err != nil {
|
||||
return types.NewError(err, types.ErrorCodeChannelParamOverrideInvalid, types.ErrOptionWithSkipRetry())
|
||||
}
|
||||
for key, value := range relayInfo.ParamOverride {
|
||||
for key, value := range info.ParamOverride {
|
||||
reqMap[key] = value
|
||||
}
|
||||
jsonData, err = json.Marshal(reqMap)
|
||||
@@ -135,7 +75,7 @@ func ResponsesHelper(c *gin.Context) (newAPIError *types.NewAPIError) {
|
||||
}
|
||||
|
||||
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)
|
||||
}
|
||||
@@ -153,17 +93,17 @@ func ResponsesHelper(c *gin.Context) (newAPIError *types.NewAPIError) {
|
||||
}
|
||||
}
|
||||
|
||||
usage, newAPIError := adaptor.DoResponse(c, httpResp, relayInfo)
|
||||
usage, newAPIError := adaptor.DoResponse(c, httpResp, info)
|
||||
if newAPIError != nil {
|
||||
// reset status code 重置状态码
|
||||
service.ResetStatusCode(newAPIError, statusCodeMappingStr)
|
||||
return newAPIError
|
||||
}
|
||||
|
||||
if strings.HasPrefix(relayInfo.OriginModelName, "gpt-4o-audio") {
|
||||
service.PostAudioConsumeQuota(c, relayInfo, usage.(*dto.Usage), preConsumedQuota, userQuota, priceData, "")
|
||||
if strings.HasPrefix(info.OriginModelName, "gpt-4o-audio") {
|
||||
service.PostAudioConsumeQuota(c, info, usage.(*dto.Usage), "")
|
||||
} else {
|
||||
postConsumeQuota(c, relayInfo, usage.(*dto.Usage), preConsumedQuota, userQuota, priceData, "")
|
||||
postConsumeQuota(c, info, usage.(*dto.Usage), "")
|
||||
}
|
||||
return nil
|
||||
}
|
||||
|
||||
Reference in New Issue
Block a user