fix(openai): harden Chat-to-Responses compatibility (#5772)

Add a shared Responses-to-Chat stream state machine and use it from the OpenAI relay path. Preserve assistant text alongside tool calls, bind tool argument deltas by output_index, map incomplete finish reasons, support reasoning/custom tool events, and buffer upstream SSE for non-stream Chat clients. Add deterministic service tests and relay SSE tests for the conversion path. Related to #5745.
2026-09-07 01:56:53 +00:00 · 2026-06-27 19:16:56 +08:00
parent 626dadb556
commit 3a506f50f0
9 changed files with 1516 additions and 470 deletions
@@ -0,0 +1,128 @@
+package openai
+
+import (
+	"io"
+	"net/http"
+	"net/http/httptest"
+	"strings"
+	"testing"
+
+	"github.com/QuantumNous/new-api/common"
+	"github.com/QuantumNous/new-api/constant"
+	relaycommon "github.com/QuantumNous/new-api/relay/common"
+	"github.com/QuantumNous/new-api/types"
+	"github.com/gin-gonic/gin"
+	"github.com/stretchr/testify/require"
+)
+
+func newResponsesChatTestContext(t *testing.T, body string, isStream bool) (*gin.Context, *httptest.ResponseRecorder, *http.Response, *relaycommon.RelayInfo) {
+	t.Helper()
+
+	recorder := httptest.NewRecorder()
+	c, _ := gin.CreateTestContext(recorder)
+	c.Request = httptest.NewRequest(http.MethodPost, "/v1/chat/completions", nil)
+	c.Set(common.RequestIdKey, "responses-test")
+
+	resp := &http.Response{
+		StatusCode: http.StatusOK,
+		Body:       io.NopCloser(strings.NewReader(body)),
+		Header:     http.Header{"Content-Type": []string{"text/event-stream"}},
+	}
+	info := &relaycommon.RelayInfo{
+		ChannelMeta:        &relaycommon.ChannelMeta{UpstreamModelName: "gpt-test"},
+		IsStream:           isStream,
+		RelayFormat:        types.RelayFormatOpenAI,
+		ShouldIncludeUsage: true,
+		DisablePing:        true,
+	}
+	return c, recorder, resp, info
+}
+
+func TestOaiResponsesToChatStreamHandlerConvertsSSEOrderAndUsage(t *testing.T) {
+	oldMode := gin.Mode()
+	gin.SetMode(gin.TestMode)
+	t.Cleanup(func() { gin.SetMode(oldMode) })
+
+	oldTimeout := constant.StreamingTimeout
+	constant.StreamingTimeout = 30
+	t.Cleanup(func() { constant.StreamingTimeout = oldTimeout })
+
+	body := strings.Join([]string{
+		`data: {"type":"response.created","response":{"id":"resp_1","model":"gpt-test","created_at":1710000000}}`,
+		`data: {"type":"response.output_text.delta","delta":"hello"}`,
+		`data: {"type":"response.output_item.added","output_index":1,"item":{"type":"function_call","id":"fc_1","call_id":"call_1","name":"lookup"}}`,
+		`data: {"type":"response.function_call_arguments.delta","output_index":1,"delta":"{\"q\":\"x\"}"}`,
+		`data: {"type":"response.completed","response":{"status":"completed","usage":{"input_tokens":2,"output_tokens":3,"total_tokens":5}}}`,
+		`data: [DONE]`,
+		``,
+	}, "\n")
+
+	c, recorder, resp, info := newResponsesChatTestContext(t, body, true)
+
+	usage, err := OaiResponsesToChatStreamHandler(c, info, resp)
+	require.Nil(t, err)
+	require.NotNil(t, usage)
+	require.Equal(t, 2, usage.PromptTokens)
+	require.Equal(t, 3, usage.CompletionTokens)
+	require.Equal(t, 5, usage.TotalTokens)
+
+	got := recorder.Body.String()
+	require.Equal(t, "text/event-stream", recorder.Header().Get("Content-Type"))
+	require.Contains(t, got, `"role":"assistant"`)
+	require.Contains(t, got, `"content":"hello"`)
+	require.Contains(t, got, `"name":"lookup"`)
+	require.Contains(t, got, `"arguments":"{\"q\":\"x\"}"`)
+	require.Contains(t, got, `"finish_reason":"tool_calls"`)
+	require.Contains(t, got, `"usage":{"prompt_tokens":2,"completion_tokens":3,"total_tokens":5`)
+	require.Contains(t, got, `data: [DONE]`)
+	requireOrderedSubstrings(t, got,
+		`"role":"assistant"`,
+		`"content":"hello"`,
+		`"name":"lookup"`,
+		`"arguments":"{\"q\":\"x\"}"`,
+		`"finish_reason":"tool_calls"`,
+		`"usage":{"prompt_tokens":2,"completion_tokens":3,"total_tokens":5`,
+		`data: [DONE]`,
+	)
+}
+
+func TestOaiResponsesToChatBufferedStreamHandlerReturnsJSONFromSSE(t *testing.T) {
+	oldMode := gin.Mode()
+	gin.SetMode(gin.TestMode)
+	t.Cleanup(func() { gin.SetMode(oldMode) })
+
+	body := strings.Join([]string{
+		`data: {"type":"response.output_text.delta","delta":"buffered text"}`,
+		`data: {"type":"response.output_item.added","output_index":0,"item":{"type":"function_call","id":"fc_1","call_id":"call_1","name":"lookup"}}`,
+		`data: {"type":"response.function_call_arguments.delta","output_index":0,"delta":"{\"q\":\"x\"}"}`,
+		`data: {"type":"response.done","response":{"model":"gpt-test","status":"completed","usage":{"input_tokens":1,"output_tokens":2,"total_tokens":3}}}`,
+		`data: [DONE]`,
+		``,
+	}, "\n")
+
+	c, recorder, resp, info := newResponsesChatTestContext(t, body, false)
+
+	usage, err := OaiResponsesToChatBufferedStreamHandler(c, info, resp)
+	require.Nil(t, err)
+	require.NotNil(t, usage)
+	require.Equal(t, 3, usage.TotalTokens)
+
+	got := recorder.Body.String()
+	require.NotContains(t, got, `data:`)
+	require.Contains(t, got, `"object":"chat.completion"`)
+	require.Contains(t, got, `"content":"buffered text"`)
+	require.Contains(t, got, `"name":"lookup"`)
+	require.Contains(t, got, `"arguments":"{\"q\":\"x\"}"`)
+	require.Contains(t, got, `"finish_reason":"tool_calls"`)
+}
+
+func requireOrderedSubstrings(t *testing.T, s string, parts ...string) {
+	t.Helper()
+
+	offset := 0
+	for _, part := range parts {
+		idx := strings.Index(s[offset:], part)
+		require.NotEqualf(t, -1, idx, "missing %q after byte offset %d", part, offset)
+		offset += idx + len(part)
+	}
+}