Files
new-api/relay/responses_handler.go
T
Calcium-Ion 0ed497f066 feat(relay): hosted-tool conversion fidelity, reasoning normalization, and billing usage integrity (#7137)
* feat(relaykit): preserve hosted tools across conversions

- add protocol-neutral hosted-tool DTOs, conversion metadata, and loss policies
- bridge citations, grounding metadata, and hosted-tool stream lifecycles
- document the public conversion behavior and channel policy controls

* refactor(relaykit): normalize reasoning and thinking intent

- centralize provider-neutral reasoning intent, effort, and budget mappings
- parse model suffixes at the host entry boundary while preserving provider-owned tails
- keep adaptive Claude thinking and explicit zero-token compatibility consistent

* fix(billing): preserve authoritative usage across relay hops

- carry native BillingUsage sidecars through direct and streamed protocol bridges
- merge partial and terminal usage monotonically with safe fallback settlement
- retain cache metadata, penultimate usage, and per-call Gemini tool surcharges

* feat(relay): bridge Responses with Claude and Gemini protocols

- add direct request, response, and stream converters across supported relay formats
- expose Claude count_tokens and Chat-to-Responses compatibility endpoints
- carry conversion diagnostics through the host while retaining the curated public goldens

* fix(relay): wire relaykit conversions into host channels

- connect handlers, adaptors, and channel settings to the standalone conversion layer
- keep model mapping, pricing identity, retries, and provider-specific suffix behavior aligned
- ignore local audit artifacts and retain focused public regression coverage
2026-09-01 21:53:35 +08:00

175 lines
6.2 KiB
Go

package relay
import (
"fmt"
"io"
"net/http"
"strings"
"github.com/QuantumNous/new-api/common"
"github.com/QuantumNous/new-api/logger"
relaycommon "github.com/QuantumNous/new-api/relay/common"
relayconstant "github.com/QuantumNous/new-api/relay/constant"
"github.com/QuantumNous/new-api/relay/helper"
"github.com/QuantumNous/new-api/relaykit/dto"
"github.com/QuantumNous/new-api/relaykit/types"
"github.com/QuantumNous/new-api/service"
"github.com/QuantumNous/new-api/setting/model_setting"
"github.com/gin-gonic/gin"
)
func ResponsesHelper(c *gin.Context, info *relaycommon.RelayInfo) (newAPIError *types.NewAPIError) {
info.InitChannelMeta(c)
if info.RelayMode == relayconstant.RelayModeResponsesCompact &&
!common.SupportsResponsesCompact(info.ChannelType, info.ApiType) {
return types.NewErrorWithStatusCode(
fmt.Errorf("unsupported endpoint %q for api type %d", "/v1/responses/compact", info.ApiType),
types.ErrorCodeInvalidRequest,
http.StatusBadRequest,
types.ErrOptionWithSkipRetry(),
)
}
var responsesReq *dto.OpenAIResponsesRequest
switch req := info.Request.(type) {
case *dto.OpenAIResponsesRequest:
responsesReq = req
case *dto.OpenAIResponsesCompactionRequest:
// Only fields documented for POST /v1/responses/compact are forwarded:
// model, input, instructions, previous_response_id, prompt_cache_key,
// prompt_cache_options, prompt_cache_retention, service_tier.
// Undocumented Codex-parity fields (tools, reasoning, text) are parsed
// for client compatibility but intentionally not sent upstream.
responsesReq = &dto.OpenAIResponsesRequest{
Model: req.Model,
Input: req.Input,
Instructions: req.Instructions,
PreviousResponseID: req.PreviousResponseID,
ParallelToolCalls: req.ParallelToolCalls,
ServiceTier: req.ServiceTier,
PromptCacheKey: req.PromptCacheKey,
PromptCacheOptions: req.PromptCacheOptions,
PromptCacheRetention: req.PromptCacheRetention,
}
default:
return types.NewErrorWithStatusCode(
fmt.Errorf("invalid request type, expected dto.OpenAIResponsesRequest or dto.OpenAIResponsesCompactionRequest, got %T", info.Request),
types.ErrorCodeInvalidRequest,
http.StatusBadRequest,
types.ErrOptionWithSkipRetry(),
)
}
request, err := common.DeepCopy(responsesReq)
if err != nil {
return types.NewError(fmt.Errorf("failed to copy request to GeneralOpenAIRequest: %w", err), types.ErrorCodeInvalidRequest, types.ErrOptionWithSkipRetry())
}
err = helper.ModelMappedHelper(c, info, request)
if err != nil {
return types.NewError(err, types.ErrorCodeChannelModelMappedError, types.ErrOptionWithSkipRetry())
}
if err = helper.ApplyReasoningModelSuffix(info, request); err != nil {
return newConvertRequestFailedError(c, info, err)
}
adaptor := GetAdaptor(info.ApiType)
if adaptor == nil {
return types.NewError(fmt.Errorf("invalid api type: %d", info.ApiType), types.ErrorCodeInvalidApiType, types.ErrOptionWithSkipRetry())
}
adaptor.Init(info)
var requestBody io.Reader
if model_setting.GetGlobalSettings().PassThroughRequestEnabled || info.ChannelSetting.PassThroughBodyEnabled {
storage, err := common.GetBodyStorage(c)
if err != nil {
return types.NewError(err, types.ErrorCodeReadRequestBodyFailed, types.ErrOptionWithSkipRetry())
}
requestBody = common.NewReplayableBodyReader(storage)
} else {
convertedRequest, err := adaptor.ConvertOpenAIResponsesRequest(c, info, *request)
if err != nil {
return newConvertRequestFailedError(c, info, err)
}
relaycommon.AppendRequestConversionFromRequest(info, convertedRequest)
jsonData, err := common.Marshal(convertedRequest)
if err != nil {
return types.NewError(err, types.ErrorCodeConvertRequestFailed, types.ErrOptionWithSkipRetry())
}
// remove disabled fields for OpenAI Responses API
jsonData, err = relaycommon.RemoveDisabledFields(jsonData, info.ChannelOtherSettings, info.ChannelSetting.PassThroughBodyEnabled)
if err != nil {
return types.NewError(err, types.ErrorCodeConvertRequestFailed, types.ErrOptionWithSkipRetry())
}
// apply param override
if len(info.ParamOverride) > 0 {
jsonData, err = relaycommon.ApplyParamOverrideWithRelayInfo(jsonData, info)
if err != nil {
return newAPIErrorFromParamOverride(err)
}
}
logger.LogDebug(c, "requestBody: %s", jsonData)
body, closer, err := relaycommon.NewOutboundJSONBody(jsonData)
if err != nil {
return types.NewError(err, types.ErrorCodeConvertRequestFailed, types.ErrOptionWithSkipRetry())
}
defer closer.Close()
jsonData = nil
requestBody = body
}
var httpResp *http.Response
resp, err := adaptor.DoRequest(c, info, requestBody)
if err != nil {
return types.NewOpenAIError(err, types.ErrorCodeDoRequestFailed, http.StatusInternalServerError)
}
statusCodeMappingStr := c.GetString("status_code_mapping")
if resp != nil {
httpResp = resp.(*http.Response)
if httpResp.StatusCode != http.StatusOK {
newAPIError = service.RelayErrorHandler(c.Request.Context(), httpResp, false)
// reset status code 重置状态码
service.ResetStatusCode(newAPIError, statusCodeMappingStr)
return newAPIError
}
}
usage, newAPIError := adaptor.DoResponse(c, httpResp, info)
if newAPIError != nil {
// reset status code 重置状态码
service.ResetStatusCode(newAPIError, statusCodeMappingStr)
return newAPIError
}
usageDto := usage.(*dto.Usage)
if info.RelayMode == relayconstant.RelayModeResponsesCompact {
originModelName := info.OriginModelName
originPriceData := info.PriceData
_, err := helper.ModelPriceHelper(c, info, info.GetEstimatePromptTokens(), &types.TokenCountMeta{})
if err != nil {
info.OriginModelName = originModelName
info.PriceData = originPriceData
return types.NewError(err, types.ErrorCodeModelPriceError, types.ErrOptionWithSkipRetry(), types.ErrOptionWithStatusCode(http.StatusBadRequest))
}
service.PostTextConsumeQuota(c, info, usageDto, nil)
info.OriginModelName = originModelName
info.PriceData = originPriceData
return nil
}
if strings.HasPrefix(info.OriginModelName, "gpt-4o-audio") {
service.PostAudioConsumeQuota(c, info, usageDto, "")
} else {
service.PostTextConsumeQuota(c, info, usageDto, nil)
}
return nil
}