chuan/CLIProxyAPI
1
0
Fork 0
mirror of https://github.com/router-for-me/CLIProxyAPI.git synced 2026-03-01 01:24:29 +08:00
Code Issues Packages Projects Releases Wiki Activity

fix(responses): include model and usage in translated streams

Browse Source 
Ensure response.created and response.completed chunks produced by the OpenAI/Gemini/Claude translators always include required fields (response.model and response.usage) so clients validating Responses SSE do not fail schema validation.
This commit is contained in:
canxin121
2026-02-23 13:22:06 +08:00
parent 49c8ec69d0
commit 5382764d8a
6 changed files with 196 additions and 36 deletions
Download Patch File Download Diff File Expand all files Collapse all files

30
internal/translator/gemini/openai/responses/gemini_openai-responses_response.go
View File

@@ -212,6 +212,7 @@ func ConvertGeminiResponseToOpenAIResponses(_ context.Context, modelName string,
created, _ = sjson.Set(created, "sequence_number", nextSeq())
created, _ = sjson.Set(created, "response.id", st.ResponseID)
created, _ = sjson.Set(created, "response.created_at", st.CreatedAt)
created, _ = sjson.Set(created, "response.model", modelName)
out = append(out, emitEvent("response.created", created))
inprog := `{"type":"response.in_progress","sequence_number":0,"response":{"id":"","object":"response","created_at":0,"status":"in_progress"}}`
@@ -529,31 +530,36 @@ func ConvertGeminiResponseToOpenAIResponses(_ context.Context, modelName string,
completed, _ = sjson.SetRaw(completed, "response.output", gjson.Get(outputsWrapper, "arr").Raw)
}
// usage mapping
input := int64(0)
cached := int64(0)
output := int64(0)
reasoning := int64(0)
total := int64(0)
if um := root.Get("usageMetadata"); um.Exists() {
// input tokens = prompt + thoughts
input := um.Get("promptTokenCount").Int() + um.Get("thoughtsTokenCount").Int()
completed, _ = sjson.Set(completed, "response.usage.input_tokens", input)
input = um.Get("promptTokenCount").Int() + um.Get("thoughtsTokenCount").Int()
// cached token details: align with OpenAI "cached_tokens" semantics.
completed, _ = sjson.Set(completed, "response.usage.input_tokens_details.cached_tokens", um.Get("cachedContentTokenCount").Int())
cached = um.Get("cachedContentTokenCount").Int()
// output tokens
if v := um.Get("candidatesTokenCount"); v.Exists() {
completed, _ = sjson.Set(completed, "response.usage.output_tokens", v.Int())
} else {
completed, _ = sjson.Set(completed, "response.usage.output_tokens", 0)
output = v.Int()
}
if v := um.Get("thoughtsTokenCount"); v.Exists() {
completed, _ = sjson.Set(completed, "response.usage.output_tokens_details.reasoning_tokens", v.Int())
} else {
completed, _ = sjson.Set(completed, "response.usage.output_tokens_details.reasoning_tokens", 0)
reasoning = v.Int()
}
if v := um.Get("totalTokenCount"); v.Exists() {
completed, _ = sjson.Set(completed, "response.usage.total_tokens", v.Int())
total = v.Int()
} else {
completed, _ = sjson.Set(completed, "response.usage.total_tokens", 0)
total = input + output
}
}
completed, _ = sjson.Set(completed, "response.usage.input_tokens", input)
completed, _ = sjson.Set(completed, "response.usage.input_tokens_details.cached_tokens", cached)
completed, _ = sjson.Set(completed, "response.usage.output_tokens", output)
completed, _ = sjson.Set(completed, "response.usage.output_tokens_details.reasoning_tokens", reasoning)
completed, _ = sjson.Set(completed, "response.usage.total_tokens", total)
out = append(out, emitEvent("response.completed", completed))
}

31
internal/translator/gemini/openai/responses/gemini_openai-responses_response_test.go
View File

@@ -53,6 +53,7 @@ func TestConvertGeminiResponseToOpenAIResponses_UnwrapAndAggregateText(t *testin
textDone string
messageText string
responseID string
createdModel string
instructions string
cachedTokens int64
@@ -68,6 +69,8 @@ func TestConvertGeminiResponseToOpenAIResponses_UnwrapAndAggregateText(t *testin
for i, chunk := range out {
ev, data := parseSSEEvent(t, chunk)
switch ev {
case "response.created":
createdModel = data.Get("response.model").String()
case "response.output_text.done":
gotTextDone = true
if posTextDone == -1 {
@@ -132,6 +135,9 @@ func TestConvertGeminiResponseToOpenAIResponses_UnwrapAndAggregateText(t *testin
if responseID != "resp_req_vrtx_1" {
t.Fatalf("unexpected response id: got %q", responseID)
}
if createdModel != "test-model" {
t.Fatalf("unexpected response.created model: got %q", createdModel)
}
if instructions != "test instructions" {
t.Fatalf("unexpected instructions echo: got %q", instructions)
}
@@ -153,6 +159,31 @@ func TestConvertGeminiResponseToOpenAIResponses_UnwrapAndAggregateText(t *testin
}
}
func TestConvertGeminiResponseToOpenAIResponses_CompletedAlwaysHasUsage(t *testing.T) {
in := `data: {"response":{"candidates":[{"content":{"role":"model","parts":[{"text":"hi"}]},"finishReason":"STOP"}],"modelVersion":"test-model","responseId":"req_no_usage"},"traceId":"t1"}`
var param any
out := ConvertGeminiResponseToOpenAIResponses(context.Background(), "test-model", nil, nil, []byte(in), &param)
gotCompleted := false
for _, chunk := range out {
ev, data := parseSSEEvent(t, chunk)
if ev != "response.completed" {
continue
}
gotCompleted = true
if !data.Get("response.usage.input_tokens").Exists() {
t.Fatalf("response.completed missing usage.input_tokens: %s", data.Raw)
}
if !data.Get("response.usage.output_tokens").Exists() {
t.Fatalf("response.completed missing usage.output_tokens: %s", data.Raw)
}
}
if !gotCompleted {
t.Fatalf("missing response.completed event")
}
}
func TestConvertGeminiResponseToOpenAIResponses_ReasoningEncryptedContent(t *testing.T) {
sig := "RXE0RENrZ0lDeEFDR0FJcVFOZDdjUzlleGFuRktRdFcvSzNyZ2MvWDNCcDQ4RmxSbGxOWUlOVU5kR1l1UHMrMGdkMVp0Vkg3ekdKU0g4YVljc2JjN3lNK0FrdGpTNUdqamI4T3Z0VVNETzdQd3pmcFhUOGl3U3hXUEJvTVFRQ09mWTFyMEtTWGZxUUlJakFqdmFGWk83RW1XRlBKckJVOVpkYzdDKw=="
in := []string{