Files
sub2api/backend/internal/handler/gateway_handler_error_fallback_test.go
T
feeeei f4f8ff04db fix(gateway): 客户端断开统一标记 499,不再被记成 502/200 上游错误
客户端在响应提交前断开时,只有排队/换号阶段和 OpenAI 的兜底写入会标 499。
Claude 非流式与 Gemini 的 Chat Completions / Messages / Responses 兼容入口
在等待上游时断开,由 GatewayHandler.ensureForwardErrorResponse 补写 502;
Gemini 原生入口什么都不写,落成 200 并被 ops 记成一条 provider 的
"Recovered upstream error";Antigravity 回写 502。这些行都按上游失败计入
SLA 与上游错误数。

- GatewayHandler.ensureForwardErrorResponse 补客户端断开检查,与 OpenAI 版一致
- GatewayHandler 的 Messages / ChatCompletions / Responses 与 Gemini 原生入口:
  首次选号失败时先判断客户端断开;ChatCompletions / Responses 循环开头改用
  failoverClientGone,不再直接 return 落成 200
- Gemini 原生入口:forward 失败时补 499 标记;并发槽错误改走
  concurrencyErrorResponse 统一映射(断开 499、排队满/超时 429、其余 503)
- Antigravity 上游响应前断开回写 499;googleapi 与 Antigravity 的 Google 错误
  格式把 499 映射为 CANCELLED
- OpenAI alpha search 非 failover 错误分支补断开检查
- Claude / Gemini / OpenAI 传输层错误:请求 context 已取消时不记 ops 上游事件
- ops:IgnoreContextCanceled 开启时跳过没有上游错误上下文的 499;带上游错误的
  499(上游失败后客户端没等到换号结果就离开)仍按上游失败落库

计费不变:Gemini 流式中途断开的用量处理不在本次范围。新增 handler/service
单测覆盖各入口的断开场景与 ops 过滤规则。
2026-09-26 17:13:10 +08:00

223 lines
8.3 KiB
Go
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
package handler
import (
"context"
"encoding/json"
"errors"
"net/http"
"net/http/httptest"
"strings"
"testing"
"github.com/Wei-Shaw/sub2api/internal/service"
"github.com/gin-gonic/gin"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
func TestGatewayEnsureForwardErrorResponse_WritesFallbackWhenNotWritten(t *testing.T) {
gin.SetMode(gin.TestMode)
w := httptest.NewRecorder()
c, _ := gin.CreateTestContext(w)
c.Request = httptest.NewRequest(http.MethodGet, "/", nil)
h := &GatewayHandler{}
wrote := h.ensureForwardErrorResponse(c, false)
require.True(t, wrote)
require.Equal(t, http.StatusBadGateway, w.Code)
var parsed map[string]any
err := json.Unmarshal(w.Body.Bytes(), &parsed)
require.NoError(t, err)
assert.Equal(t, "error", parsed["type"])
errorObj, ok := parsed["error"].(map[string]any)
require.True(t, ok)
assert.Equal(t, "upstream_error", errorObj["type"])
assert.Equal(t, "Upstream request failed", errorObj["message"])
}
// Writer 已写后 ensureForwardErrorResponse 必须把错误以 SSE 形式追加,
// 而不是 silent EOF。非 /responses 路径走 legacy data:{"type":"error"} 分支。
func TestGatewayEnsureForwardErrorResponse_AppendsSSEAfterWritten(t *testing.T) {
gin.SetMode(gin.TestMode)
w := httptest.NewRecorder()
c, _ := gin.CreateTestContext(w)
c.Request = httptest.NewRequest(http.MethodGet, "/", nil)
c.String(http.StatusTeapot, "already written")
h := &GatewayHandler{}
wrote := h.ensureForwardErrorResponse(c, false)
require.True(t, wrote)
require.Equal(t, http.StatusTeapot, w.Code)
assert.Contains(t, w.Body.String(), "already written")
assert.Contains(t, w.Body.String(), `data: {"type":"error"`)
}
func TestGatewayEnsureForwardErrorResponse_SkipsCommittedSSEError(t *testing.T) {
gin.SetMode(gin.TestMode)
w := httptest.NewRecorder()
c, _ := gin.CreateTestContext(w)
c.Request = httptest.NewRequest(http.MethodPost, EndpointResponses, nil)
c.Header("Content-Type", "text/event-stream")
_, _ = c.Writer.WriteString("event: error\ndata: {\"type\":\"error\"}\n\n")
service.MarkResponseCommitted(c)
h := &GatewayHandler{}
wrote := h.ensureForwardErrorResponse(c, true)
require.False(t, wrote)
require.Equal(t, 1, strings.Count(w.Body.String(), "event: error"))
}
// 客户端已断开且响应未提交:不补写 502,标记 499 供访问日志与 ops 归类。
func TestGatewayEnsureForwardErrorResponse_SkipsCanceledClient(t *testing.T) {
gin.SetMode(gin.TestMode)
w := httptest.NewRecorder()
c, _ := gin.CreateTestContext(w)
ctx, cancel := context.WithCancel(context.Background())
c.Request = httptest.NewRequest(http.MethodPost, EndpointMessages, nil).WithContext(ctx)
cancel()
h := &GatewayHandler{}
wrote := h.ensureForwardErrorResponse(c, false)
require.False(t, wrote)
require.Equal(t, statusClientClosedRequest, c.Writer.Status())
require.Empty(t, w.Body.String())
}
// 客户端已断开但流已开始:状态码已固化为 200,不再向断开的连接追加错误帧。
func TestGatewayEnsureForwardErrorResponse_CanceledClientAfterStreamStartedAppendsNothing(t *testing.T) {
gin.SetMode(gin.TestMode)
w := httptest.NewRecorder()
c, _ := gin.CreateTestContext(w)
ctx, cancel := context.WithCancel(context.Background())
c.Request = httptest.NewRequest(http.MethodPost, EndpointMessages, nil).WithContext(ctx)
c.Header("Content-Type", "text/event-stream")
_, _ = c.Writer.WriteString(":\n\n")
cancel()
h := &GatewayHandler{}
wrote := h.ensureForwardErrorResponse(c, true)
require.False(t, wrote)
require.Equal(t, http.StatusOK, c.Writer.Status())
require.Equal(t, ":\n\n", w.Body.String())
require.Empty(t, service.GetOpsStreamErrors(c))
}
// case B 回归:Anthropic-backed /responses,Writer 已被写过时
// ensureForwardErrorResponse 仍要发 response.failed。
func TestGatewayEnsureForwardErrorResponse_ResponsesRouteAfterWrittenEmitsResponseFailed(t *testing.T) {
gin.SetMode(gin.TestMode)
w := httptest.NewRecorder()
c, _ := gin.CreateTestContext(w)
c.Request = httptest.NewRequest(http.MethodPost, EndpointResponses, nil)
_, _ = c.Writer.WriteString(":\n\n")
h := &GatewayHandler{}
wrote := h.ensureForwardErrorResponse(c, false)
require.True(t, wrote)
body := w.Body.String()
assert.Contains(t, body, ":\n\n")
assert.Contains(t, body, "event: response.failed\n")
assert.Contains(t, body, `"type":"response.failed"`)
}
func TestGatewayForwardErrorAlreadyCommunicated(t *testing.T) {
gin.SetMode(gin.TestMode)
t.Run("json error already written", func(t *testing.T) {
w := httptest.NewRecorder()
c, _ := gin.CreateTestContext(w)
c.Request = httptest.NewRequest(http.MethodPost, EndpointMessages, nil)
before := c.Writer.Size()
c.JSON(http.StatusBadGateway, gin.H{
"type": "error",
"error": gin.H{
"type": "upstream_error",
"message": "Your Claude Code version (2.1.39) is below the minimum required version (2.1.81). Please update: npm update -g @anthropic-ai/claude-code",
},
})
reported := gatewayForwardErrorAlreadyCommunicated(c, before, errors.New("upstream error: 400 message=version too low"))
require.True(t, reported)
body := w.Body.String()
assert.NotContains(t, body, `data: {"type":"error"`)
})
t.Run("sse ping still needs fallback", func(t *testing.T) {
w := httptest.NewRecorder()
c, _ := gin.CreateTestContext(w)
c.Request = httptest.NewRequest(http.MethodPost, EndpointMessages, nil)
c.Header("Content-Type", "text/event-stream")
before := c.Writer.Size()
_, _ = c.Writer.WriteString(":\n\n")
reported := gatewayForwardErrorAlreadyCommunicated(c, before, errors.New("stream read error: unexpected EOF"))
require.False(t, reported)
})
t.Run("no write still needs fallback", func(t *testing.T) {
w := httptest.NewRecorder()
c, _ := gin.CreateTestContext(w)
c.Request = httptest.NewRequest(http.MethodPost, EndpointMessages, nil)
reported := gatewayForwardErrorAlreadyCommunicated(c, c.Writer.Size(), errors.New("upstream request failed"))
require.False(t, reported)
})
// apikey 场景核心回归:复刻 GatewayService.handleErrorResponse 的 case 400 ——
// 原样透传上游 JSON body 后返回 err。此时错误已经完整告知客户端,
// handler 不得再追加 data:{"type":"error"} 帧,否则响应被污染成「JSON + 一行 data:」。
t.Run("upstream 400 json passthrough via c.Data", func(t *testing.T) {
w := httptest.NewRecorder()
c, _ := gin.CreateTestContext(w)
c.Request = httptest.NewRequest(http.MethodPost, EndpointMessages, nil)
before := c.Writer.Size()
upstreamBody := []byte(`{"type":"error","error":{"type":"upstream_error","message":"Your Claude Code version (2.1.39) is below the minimum required version (2.1.81). Please update: npm update -g @anthropic-ai/claude-code"}}`)
c.Data(http.StatusBadRequest, "application/json", upstreamBody)
reported := gatewayForwardErrorAlreadyCommunicated(c, before, errors.New("upstream error: 400 message=version too low"))
require.True(t, reported)
body := w.Body.String()
assert.NotContains(t, body, `data: {"type":"error"`)
// 客户端只应收到上游那一份错误,没有被追加第二份。
assert.Equal(t, 1, strings.Count(body, `"type":"error"`))
})
// 流式已开始(已 flush 真实 SSE 事件,不只是 ping)+ 上游中途 400:
// HTTP 200 已固化,仍需 handler 补协议级终止帧,故不算「已完整告知」。
t.Run("streaming 400 mid-stream still needs fallback", func(t *testing.T) {
w := httptest.NewRecorder()
c, _ := gin.CreateTestContext(w)
c.Request = httptest.NewRequest(http.MethodPost, EndpointMessages, nil)
c.Header("Content-Type", "text/event-stream")
before := c.Writer.Size()
_, _ = c.Writer.WriteString("event: message_start\ndata: {\"type\":\"message_start\"}\n\n")
reported := gatewayForwardErrorAlreadyCommunicated(c, before, errors.New("upstream error: 400 message=version too low"))
require.False(t, reported)
})
// 防御边界:err 为 nil 时永远不算「已告知」,避免在成功路径误吞兜底逻辑。
t.Run("nil error never reports communicated", func(t *testing.T) {
w := httptest.NewRecorder()
c, _ := gin.CreateTestContext(w)
c.Request = httptest.NewRequest(http.MethodPost, EndpointMessages, nil)
c.JSON(http.StatusOK, gin.H{"ok": true})
reported := gatewayForwardErrorAlreadyCommunicated(c, 0, nil)
require.False(t, reported)
})
}