feat: log shows request conversion - #2703
Merged
Calcium-Ion merged 4 commits intoJan 21, 2026
Merged
Conversation
Contributor
WalkthroughThis PR adds request-conversion tracking to RelayInfo, records conversions from converted requests across multiple handlers, surfaces a human-readable conversion chain in logs, strips admin-only debug fields when serializing logs, and exposes conversion info to the admin UI with i18n entries. Changes
Sequence DiagramsequenceDiagram
participant Client
participant Handler
participant RelayCommon as RelayInfo/Common
participant Logger
participant UI
Client->>Handler: Send request
Handler->>Handler: Convert request to target format
Handler->>RelayCommon: AppendRequestConversionFromRequest(info, convertedRequest)
RelayCommon->>RelayCommon: GuessRelayFormatFromRequest / AppendRequestConversion
Handler->>Logger: Include RelayInfo with RequestConversionChain
Logger->>Logger: appendRequestConversionChain -> other["request_conversion"]
Logger->>UI: Emit log data (admin-only request_conversion)
UI->>UI: Format conversion chain for display
UI-->>Client: Render conversion row (admin visible)
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~20 minutes Possibly related PRs
Suggested reviewers
Poem
🚥 Pre-merge checks | ✅ 2 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (2 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing touches
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
Actionable comments posted: 2
🤖 Fix all issues with AI agents
In `@relay/common/relay_info.go`:
- Around line 454-500: The Task and MjProxy branches in GenRelayInfo call
genBaseRelayInfo(c, nil) but do not set RelayInfo.RelayFormat, so
InitRequestConversionChain() has no starting format; update the
types.RelayFormatTask and types.RelayFormatMjProxy cases in GenRelayInfo to
create the base RelayInfo via genBaseRelayInfo and then set info.RelayFormat =
types.RelayFormatTask (or types.RelayFormatMjProxy) respectively before
returning so InitRequestConversionChain can seed the conversion chain correctly.
In `@service/log_info_generate.go`:
- Around line 73-103: appendRequestConversionChain currently maps
relayInfo.RequestConversionChain entries to hard-coded English labels ("OpenAI
Compatible", "Claude Messages", etc.); change it to emit locale-agnostic
canonical identifiers instead (e.g., the enum/string values from
types.RelayFormat*). In function appendRequestConversionChain, replace the
switch that appends human-readable strings with logic that appends the canonical
identifier (for example string(f) or f.String() if available) into the chain
slice, and keep setting other["request_conversion"] = chain so the frontend can
handle localization.
Comment on lines
454
to
+500
| func GenRelayInfo(c *gin.Context, relayFormat types.RelayFormat, request dto.Request, ws *websocket.Conn) (*RelayInfo, error) { | ||
| var info *RelayInfo | ||
| var err error | ||
| switch relayFormat { | ||
| case types.RelayFormatOpenAI: | ||
| return GenRelayInfoOpenAI(c, request), nil | ||
| info = GenRelayInfoOpenAI(c, request) | ||
| case types.RelayFormatOpenAIAudio: | ||
| return GenRelayInfoOpenAIAudio(c, request), nil | ||
| info = GenRelayInfoOpenAIAudio(c, request) | ||
| case types.RelayFormatOpenAIImage: | ||
| return GenRelayInfoImage(c, request), nil | ||
| info = GenRelayInfoImage(c, request) | ||
| case types.RelayFormatOpenAIRealtime: | ||
| return GenRelayInfoWs(c, ws), nil | ||
| info = GenRelayInfoWs(c, ws) | ||
| case types.RelayFormatClaude: | ||
| return GenRelayInfoClaude(c, request), nil | ||
| info = GenRelayInfoClaude(c, request) | ||
| case types.RelayFormatRerank: | ||
| if request, ok := request.(*dto.RerankRequest); ok { | ||
| return GenRelayInfoRerank(c, request), nil | ||
| info = GenRelayInfoRerank(c, request) | ||
| break | ||
| } | ||
| return nil, errors.New("request is not a RerankRequest") | ||
| err = errors.New("request is not a RerankRequest") | ||
| case types.RelayFormatGemini: | ||
| return GenRelayInfoGemini(c, request), nil | ||
| info = GenRelayInfoGemini(c, request) | ||
| case types.RelayFormatEmbedding: | ||
| return GenRelayInfoEmbedding(c, request), nil | ||
| info = GenRelayInfoEmbedding(c, request) | ||
| case types.RelayFormatOpenAIResponses: | ||
| if request, ok := request.(*dto.OpenAIResponsesRequest); ok { | ||
| return GenRelayInfoResponses(c, request), nil | ||
| info = GenRelayInfoResponses(c, request) | ||
| break | ||
| } | ||
| return nil, errors.New("request is not a OpenAIResponsesRequest") | ||
| err = errors.New("request is not a OpenAIResponsesRequest") | ||
| case types.RelayFormatTask: | ||
| return genBaseRelayInfo(c, nil), nil | ||
| info = genBaseRelayInfo(c, nil) | ||
| case types.RelayFormatMjProxy: | ||
| return genBaseRelayInfo(c, nil), nil | ||
| info = genBaseRelayInfo(c, nil) | ||
| default: | ||
| return nil, errors.New("invalid relay format") | ||
| err = errors.New("invalid relay format") | ||
| } | ||
|
|
||
| if err != nil { | ||
| return nil, err | ||
| } | ||
| if info == nil { | ||
| return nil, errors.New("failed to build relay info") | ||
| } | ||
|
|
||
| info.InitRequestConversionChain() | ||
| return info, nil |
Contributor
There was a problem hiding this comment.
Set RelayFormat for task/mj_proxy to initialize the conversion chain.
InitRequestConversionChain seeds from RelayInfo.RelayFormat, but the Task/MjProxy cases only call genBaseRelayInfo, leaving the chain empty. Assign the format in those cases so logs include the starting format.
🔧 Suggested fix
case types.RelayFormatTask:
info = genBaseRelayInfo(c, nil)
+ info.RelayFormat = types.RelayFormatTask
case types.RelayFormatMjProxy:
info = genBaseRelayInfo(c, nil)
+ info.RelayFormat = types.RelayFormatMjProxy🤖 Prompt for AI Agents
In `@relay/common/relay_info.go` around lines 454 - 500, The Task and MjProxy
branches in GenRelayInfo call genBaseRelayInfo(c, nil) but do not set
RelayInfo.RelayFormat, so InitRequestConversionChain() has no starting format;
update the types.RelayFormatTask and types.RelayFormatMjProxy cases in
GenRelayInfo to create the base RelayInfo via genBaseRelayInfo and then set
info.RelayFormat = types.RelayFormatTask (or types.RelayFormatMjProxy)
respectively before returning so InitRequestConversionChain can seed the
conversion chain correctly.
Comment on lines
+73
to
+103
| appendRequestConversionChain(relayInfo, other) | ||
| return other | ||
| } | ||
|
|
||
| func appendRequestConversionChain(relayInfo *relaycommon.RelayInfo, other map[string]interface{}) { | ||
| if relayInfo == nil || other == nil { | ||
| return | ||
| } | ||
| if len(relayInfo.RequestConversionChain) == 0 { | ||
| return | ||
| } | ||
| chain := make([]string, 0, len(relayInfo.RequestConversionChain)) | ||
| for _, f := range relayInfo.RequestConversionChain { | ||
| switch f { | ||
| case types.RelayFormatOpenAI: | ||
| chain = append(chain, "OpenAI Compatible") | ||
| case types.RelayFormatClaude: | ||
| chain = append(chain, "Claude Messages") | ||
| case types.RelayFormatGemini: | ||
| chain = append(chain, "Google Gemini") | ||
| case types.RelayFormatOpenAIResponses: | ||
| chain = append(chain, "OpenAI Responses") | ||
| default: | ||
| chain = append(chain, string(f)) | ||
| } | ||
| } | ||
| if len(chain) == 0 { | ||
| return | ||
| } | ||
| other["request_conversion"] = chain | ||
| } |
Contributor
There was a problem hiding this comment.
Consider emitting locale-agnostic format identifiers.
appendRequestConversionChain hard-codes English labels (e.g., “OpenAI Compatible”), which will show untranslated in non-English UIs. Prefer storing canonical relay-format values and localizing in the frontend.
🤖 Prompt for AI Agents
In `@service/log_info_generate.go` around lines 73 - 103,
appendRequestConversionChain currently maps relayInfo.RequestConversionChain
entries to hard-coded English labels ("OpenAI Compatible", "Claude Messages",
etc.); change it to emit locale-agnostic canonical identifiers instead (e.g.,
the enum/string values from types.RelayFormat*). In function
appendRequestConversionChain, replace the switch that appends human-readable
strings with logic that appends the canonical identifier (for example string(f)
or f.String() if available) into the chain slice, and keep setting
other["request_conversion"] = chain so the frontend can handle localization.
Contributor
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Fix all issues with AI agents
In `@web/src/hooks/usage-logs/useUsageLogsData.jsx`:
- Around line 309-317: The function requestConversionDisplayValue currently
hides single-step conversions by returning "原生格式" when chain.length <= 1; update
its logic so that only empty chains are treated as native by changing the
condition to check chain.length === 0 (preserve the Array.isArray filtering
behavior and the join for non-empty chains) so arrays like ["OpenAI Compatible"]
are displayed rather than labeled native.
Comment on lines
+309
to
+317
| const requestConversionDisplayValue = (conversionChain) => { | ||
| const chain = Array.isArray(conversionChain) | ||
| ? conversionChain.filter(Boolean) | ||
| : []; | ||
| if (chain.length <= 1) { | ||
| return t('原生格式'); | ||
| } | ||
| return `${chain.join(' -> ')}`; | ||
| }; |
Contributor
There was a problem hiding this comment.
🧩 Analysis chain
🏁 Script executed:
# First, verify the file exists and check the specific lines
fd "useUsageLogsData.jsx" --type fRepository: QuantumNous/new-api
Length of output: 108
🏁 Script executed:
# Examine the file around lines 309-317
if [ -f "web/src/hooks/usage-logs/useUsageLogsData.jsx" ]; then
echo "=== Lines 300-330 ===" && sed -n '300,330p' "web/src/hooks/usage-logs/useUsageLogsData.jsx"
echo ""
echo "=== Searching for request_conversion usage ===" && grep -n "request_conversion\|conversionChain\|requestConversionDisplayValue" "web/src/hooks/usage-logs/useUsageLogsData.jsx" | head -20
else
echo "File not found"
fiRepository: QuantumNous/new-api
Length of output: 1308
🏁 Script executed:
# Search for where other.request_conversion comes from (likely API response)
rg "request_conversion|requestConversionDisplayValue" web/src/hooks/usage-logs/ -A 2 -B 2Repository: QuantumNous/new-api
Length of output: 893
🏁 Script executed:
# Look for type definitions or API contracts related to request_conversion
fd -e "ts" -e "tsx" -e "d.ts" | xargs rg "request_conversion" 2>/dev/null | head -20Repository: QuantumNous/new-api
Length of output: 304
🏁 Script executed:
# Find and examine service/log_info_generate.go
fd "log_info_generate.go" --type fRepository: QuantumNous/new-api
Length of output: 91
🏁 Script executed:
# Look at the context around request_conversion assignment in the backend
if [ -f "service/log_info_generate.go" ]; then
# Find lines with request_conversion and show context
rg "request_conversion|chain" "service/log_info_generate.go" -B 5 -A 5 | head -60
else
echo "File not found"
fiRepository: QuantumNous/new-api
Length of output: 1094
🏁 Script executed:
# Also check model/log.go to understand when request_conversion is deleted
if [ -f "model/log.go" ]; then
rg "request_conversion" "model/log.go" -B 3 -A 3
else
echo "File not found"
fiRepository: QuantumNous/new-api
Length of output: 284
🏁 Script executed:
# Search for type definitions or comments about the conversion chain structure
rg "type.*[Cc]onversion|[Cc]onversion.*struct" --type go | head -20Repository: QuantumNous/new-api
Length of output: 45
🏁 Script executed:
# Verify if there are any tests or comments that clarify the expected behavior
rg "RequestConversionChain|request_conversion" --type go -B 2 -A 2 | grep -E "comment|//|len\(chain\)|len\(|if.*==" | head -30Repository: QuantumNous/new-api
Length of output: 880
🏁 Script executed:
# Check if there are any tests related to request_conversion formatting
fd -e "test.go" -e "_test.go" | xargs rg "request_conversion\|requestConversion" 2>/dev/nullRepository: QuantumNous/new-api
Length of output: 47
🏁 Script executed:
# Double-check the exact behavior: when is request_conversion NOT set vs when it IS set
sed -n '1,100p' "service/log_info_generate.go" | grep -A 50 "RequestConversionChain"Repository: QuantumNous/new-api
Length of output: 872
🏁 Script executed:
# Confirm the actual behavior when request_conversion is undefined/missing
sed -n '490,510p' "web/src/hooks/usage-logs/useUsageLogsData.jsx"Repository: QuantumNous/new-api
Length of output: 661
Display single-step conversions instead of hiding them as native format.
The condition chain.length <= 1 incorrectly treats single-element conversion arrays as native format. The backend only sets request_conversion when there are one or more conversions (e.g., ["OpenAI Compatible"]), so admins will see "原生格式" even when a conversion occurred. Change the condition to chain.length === 0 to properly distinguish native requests from converted ones.
Suggested fix
- const requestConversionDisplayValue = (conversionChain) => {
- const chain = Array.isArray(conversionChain)
- ? conversionChain.filter(Boolean)
- : [];
- if (chain.length <= 1) {
- return t('原生格式');
- }
- return `${chain.join(' -> ')}`;
- };
+ const requestConversionDisplayValue = (conversionChain) => {
+ const chain = Array.isArray(conversionChain)
+ ? conversionChain.filter(Boolean)
+ : conversionChain
+ ? [conversionChain]
+ : [];
+ if (chain.length === 0) {
+ return t('原生格式');
+ }
+ return chain.join(' -> ');
+ };🤖 Prompt for AI Agents
In `@web/src/hooks/usage-logs/useUsageLogsData.jsx` around lines 309 - 317, The
function requestConversionDisplayValue currently hides single-step conversions
by returning "原生格式" when chain.length <= 1; update its logic so that only empty
chains are treated as native by changing the condition to check chain.length ===
0 (preserve the Array.isArray filtering behavior and the join for non-empty
chains) so arrays like ["OpenAI Compatible"] are displayed rather than labeled
native.
ennnnny
pushed a commit
to ennnnny/new-api
that referenced
this pull request
Mar 17, 2026
…ion-info feat: log shows request conversion
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary by CodeRabbit
New Features
Refactor
Documentation
✏️ Tip: You can customize this high-level summary in your review settings.