Compare commits

...
7 Commits
Author SHA1 Message Date
Wesley Liddick 8f2cbb53d9 Merge pull request #8177 from feeeei/fix/gemini-negative-input-tokens
fix(gemini): prompt 被拦截只报缓存命中时 input_tokens 不再为负
2026-10-19 20:45:15 +08:00
Wesley Liddick 0fa02816af Merge pull request #8168 from feeeei/fix/gemini-stream-client-disconnect-billing
fix(gemini): 流式中途客户端断开时按已收集用量计费
2026-10-19 20:45:02 +08:00
Wesley Liddick 378984d87f Merge pull request #8223 from Wei-Shaw/fix/async-image-download-policy
fix(images): 异步生图转存的图片下载与同步回填策略对齐
2026-10-19 18:33:04 +08:00
shaw ab71e12ec4 fix(images): 异步生图转存的图片下载与同步回填策略对齐
异步生图任务把上游返回的图片 url 下载后转存对象存储,此前沿用默认 http.Client,
与同步 Images 端点 url 回填(c227863d)的下载策略不一致。

- 下载目的地为 localhost、回环、私网、链路本地、CGNAT 等非公网地址时不下载,
  重定向的每一跳同样校验;直连时在 socket 层按实际连接 IP 判断
- 继续支持环境变量代理:经代理时先校验目标主机的解析结果
- 响应声明类型与字节嗅探都不是图片时任务失败,不再按 png 落盘
- 下载客户端进程内共享,复用连接池;超时与大小上限不变
2026-10-19 17:51:18 +08:00
feeeei 9c41b32695 fix(gemini): prompt 被拦截只报缓存命中时 input_tokens 不再为负
- Vertex 拦截 prompt(promptFeedback.blockReason,如 PROHIBITED_CONTENT)时 usageMetadata
  可能只带 cachedContentTokenCount、缺 promptTokenCount,promptTokenCount − cachedContentTokenCount
  得到负的 input_tokens 与负费用
- extractGeminiUsage 与 antigravity 非流式 / 流式(累计用量、message_start)四处换算改为差值下限 0,
  缓存命中仍照上游所报计入 cache_read_input_tokens
- 补充表测试、原生流式拦截用例与 antigravity 两条路径用例
2026-10-17 19:09:23 +08:00
feeeei 2b1f129385 fix(gemini): 流式中途客户端断开时按已收集用量计费
- 原生 /v1beta、Claude 兼容 /v1/messages、Chat 兼容三条流式路径:读取中断由客户端
  断开引起(context.Canceled 或请求 context 已取消),且已观测到 token 用量或已回吐
  图片时,以已收集结果收尾并返回成功,handler 照常提交 usage;ForwardResult 标记
  ClientDisconnect
- 读到结束分片后断开按完整用量计费;中途断开按断开前最后一份 usageMetadata 计费,
  断开前无任何用量时仍不计费
- 图片数:收到 finishReason 前断开只计已送达的图片(原生按已观测 inlineData,Claude/
  Chat 兼容流不转发图片计 0);收到 finishReason 后断开与正常完成同口径
- Chat 兼容流写失败断开的既有返回同样带上断开标记与结束状态
- 原生路径读错误时不处理未以换行结尾的残行,避免截断的 usageMetadata 覆盖已收集用量
- 上游 context 不解绑:Gemini 兼容链路没有流空闲超时,解绑后上游挂起会让读取无限期阻塞
- 客户端未断开时的上游读错误、非流式客户端断开、Antigravity 非流式转换的读错误
  行为不变
2026-10-17 01:00:51 +08:00
github-actions[bot] b0a02a3304 chore: sync VERSION to 0.2.17 [skip ci] 2026-10-16 04:42:02 +00:00
12 changed files with 815 additions and 48 deletions
+1 -1
View File
@@ -1 +1 @@
0.2.16
0.2.17
@@ -277,11 +277,12 @@ func (p *NonStreamingProcessor) buildResponse(geminiResp *GeminiResponse, respon
}
// 注意:Gemini 的 promptTokenCount 包含 cachedContentTokenCount,
// 但 Claude 的 input_tokens 不包含 cache_read_input_tokens,需要减去
// 但 Claude 的 input_tokens 不包含 cache_read_input_tokens,需要减去。
// prompt 被拦截时上游可能只报 cachedContentTokenCount,差值下限为 0。
usage := ClaudeUsage{}
if geminiResp.UsageMetadata != nil {
cached := geminiResp.UsageMetadata.CachedContentTokenCount
usage.InputTokens = geminiResp.UsageMetadata.PromptTokenCount - cached
usage.InputTokens = max(geminiResp.UsageMetadata.PromptTokenCount-cached, 0)
usage.OutputTokens = geminiResp.UsageMetadata.CandidatesTokenCount + geminiResp.UsageMetadata.ThoughtsTokenCount
usage.CacheReadInputTokens = cached
usage.ImageOutputTokens = geminiResp.UsageMetadata.ImageOutputTokens()
@@ -109,10 +109,11 @@ func (p *StreamingProcessor) ProcessLine(line string) []byte {
// 更新 usage
// 注意:Gemini 的 promptTokenCount 包含 cachedContentTokenCount,
// 但 Claude 的 input_tokens 不包含 cache_read_input_tokens,需要减去
// 但 Claude 的 input_tokens 不包含 cache_read_input_tokens,需要减去。
// prompt 被拦截时上游可能只报 cachedContentTokenCount,差值下限为 0。
if geminiResp.UsageMetadata != nil {
cached := geminiResp.UsageMetadata.CachedContentTokenCount
p.inputTokens = geminiResp.UsageMetadata.PromptTokenCount - cached
p.inputTokens = max(geminiResp.UsageMetadata.PromptTokenCount-cached, 0)
p.outputTokens = geminiResp.UsageMetadata.CandidatesTokenCount + geminiResp.UsageMetadata.ThoughtsTokenCount
p.cacheReadTokens = cached
p.imageOutputTokens = geminiResp.UsageMetadata.ImageOutputTokens()
@@ -190,7 +191,7 @@ func (p *StreamingProcessor) emitMessageStart(v1Resp *V1InternalResponse) []byte
usage := ClaudeUsage{}
if v1Resp.Response.UsageMetadata != nil {
cached := v1Resp.Response.UsageMetadata.CachedContentTokenCount
usage.InputTokens = v1Resp.Response.UsageMetadata.PromptTokenCount - cached
usage.InputTokens = max(v1Resp.Response.UsageMetadata.PromptTokenCount-cached, 0)
usage.OutputTokens = v1Resp.Response.UsageMetadata.CandidatesTokenCount + v1Resp.Response.UsageMetadata.ThoughtsTokenCount
usage.CacheReadInputTokens = cached
usage.ImageOutputTokens = v1Resp.Response.UsageMetadata.ImageOutputTokens()
@@ -38,3 +38,29 @@ func TestGeminiUsageMapping_NoCacheCreationTokens(t *testing.T) {
require.Zero(t, usage.CacheCreationInputTokens)
})
}
// usageMetadata 缺 promptTokenCount、只有 cachedContentTokenCount 时,input_tokens 下限为 0,
// 缓存命中仍计入 cache_read_input_tokens。
func TestGeminiUsageMapping_MissingPromptTokenCountKeepsInputNonNegative(t *testing.T) {
const geminiBody = `{"candidates":[{"content":{"parts":[{"text":"hi"}],"role":"model"},"finishReason":"STOP"}],` +
`"usageMetadata":{"cachedContentTokenCount":30}}`
t.Run("non-stream", func(t *testing.T) {
_, usage, err := TransformGeminiToClaude([]byte(geminiBody), "gemini-3.1-pro-preview")
require.NoError(t, err)
require.NotNil(t, usage)
require.Zero(t, usage.InputTokens)
require.Equal(t, 30, usage.CacheReadInputTokens)
})
t.Run("stream", func(t *testing.T) {
p := NewStreamingProcessor("gemini-3.1-pro-preview")
out := p.ProcessLine(`data: {"response":` + geminiBody + `}`)
require.True(t, strings.Contains(string(out), `"message_start"`))
require.NotContains(t, string(out), `"input_tokens":-`, "message_start 的 input_tokens 不应为负")
_, usage := p.Finish()
require.NotNil(t, usage)
require.Zero(t, usage.InputTokens)
require.Equal(t, 30, usage.CacheReadInputTokens)
})
}
@@ -250,6 +250,8 @@ func (s *GeminiMessagesCompatService) forwardClaudeBodyAsChatCompletions(
var usage *ClaudeUsage
var firstTokenMs *int
clientDisconnect := false
streamFinished := false
if clientStream {
streamRes, err := s.handleChatCompletionsStreamingResponseFromGemini(c, resp, startTime, originalModel, account.Type == AccountTypeOAuth, includeUsage)
if err != nil {
@@ -257,6 +259,8 @@ func (s *GeminiMessagesCompatService) forwardClaudeBodyAsChatCompletions(
}
usage = streamRes.usage
firstTokenMs = streamRes.firstTokenMs
clientDisconnect = streamRes.clientDisconnect
streamFinished = streamRes.finished
} else if useUpstreamStream {
collected, usageObj, err := collectGeminiSSE(resp.Body, account.Type == AccountTypeOAuth)
if err != nil {
@@ -287,6 +291,10 @@ func (s *GeminiMessagesCompatService) forwardClaudeBodyAsChatCompletions(
if isImageGenerationModel(originalModel) {
imageCount = 1
}
// 客户端在收到结束分片前断开时只计已送达的图片:Chat 流不向客户端转发图片。
if clientDisconnect && !streamFinished {
imageCount = 0
}
return &ForwardResult{
RequestID: requestID,
@@ -301,7 +309,7 @@ func (s *GeminiMessagesCompatService) forwardClaudeBodyAsChatCompletions(
ImageCount: imageCount,
ImageSize: imageSize,
ImageInputSize: imageInputSize,
ClientDisconnect: false,
ClientDisconnect: clientDisconnect,
}, nil
}
@@ -532,6 +540,11 @@ func (s *GeminiMessagesCompatService) handleChatCompletionsStreamingResponseFrom
var usage ClaudeUsage
var firstTokenMs *int
firstChunk := true
finishReason := ""
clientDisconnect := false
streamResult := func() *geminiStreamResult {
return &geminiStreamResult{usage: &usage, firstTokenMs: firstTokenMs, clientDisconnect: clientDisconnect, finished: finishReason != ""}
}
writeChatChunk := func(chunk apicompat.ChatCompletionsChunk) bool {
sse, err := apicompat.ChatChunkToSSE(chunk)
@@ -539,6 +552,7 @@ func (s *GeminiMessagesCompatService) handleChatCompletionsStreamingResponseFrom
return false
}
if _, err := io.WriteString(c.Writer, sse); err != nil {
clientDisconnect = true
return true
}
return false
@@ -571,10 +585,9 @@ func (s *GeminiMessagesCompatService) handleChatCompletionsStreamingResponseFrom
Usage: apicompat.AnthropicUsage{},
},
}) {
return &geminiStreamResult{usage: &usage, firstTokenMs: firstTokenMs}, nil
return streamResult(), nil
}
finishReason := ""
sawToolUse := false
nextBlockIndex := 0
openBlockIndex := -1
@@ -643,12 +656,12 @@ func (s *GeminiMessagesCompatService) handleChatCompletionsStreamingResponseFrom
}
if openToolIndex >= 0 {
if closeOpenTool() {
return &geminiStreamResult{usage: &usage, firstTokenMs: firstTokenMs}, nil
return streamResult(), nil
}
}
if openBlockType != "thinking" {
if closeOpenBlock() {
return &geminiStreamResult{usage: &usage, firstTokenMs: firstTokenMs}, nil
return streamResult(), nil
}
idx := nextBlockIndex
nextBlockIndex++
@@ -659,21 +672,21 @@ func (s *GeminiMessagesCompatService) handleChatCompletionsStreamingResponseFrom
Index: &idx,
ContentBlock: &apicompat.AnthropicContentBlock{Type: "thinking"},
}) {
return &geminiStreamResult{usage: &usage, firstTokenMs: firstTokenMs}, nil
return streamResult(), nil
}
}
if emitAnthropicEvent(&apicompat.AnthropicStreamEvent{
Type: "content_block_delta",
Delta: &apicompat.AnthropicDelta{Type: "thinking_delta", Thinking: delta},
}) {
return &geminiStreamResult{usage: &usage, firstTokenMs: firstTokenMs}, nil
return streamResult(), nil
}
continue
}
if text, ok := part["text"].(string); ok && text != "" {
if openToolIndex >= 0 {
if closeOpenTool() {
return &geminiStreamResult{usage: &usage, firstTokenMs: firstTokenMs}, nil
return streamResult(), nil
}
}
delta, newSeen := computeGeminiTextDelta(seenText, text)
@@ -683,7 +696,7 @@ func (s *GeminiMessagesCompatService) handleChatCompletionsStreamingResponseFrom
}
if openBlockType != "text" {
if closeOpenBlock() {
return &geminiStreamResult{usage: &usage, firstTokenMs: firstTokenMs}, nil
return streamResult(), nil
}
idx := nextBlockIndex
nextBlockIndex++
@@ -697,7 +710,7 @@ func (s *GeminiMessagesCompatService) handleChatCompletionsStreamingResponseFrom
Text: "",
},
}) {
return &geminiStreamResult{usage: &usage, firstTokenMs: firstTokenMs}, nil
return streamResult(), nil
}
}
if emitAnthropicEvent(&apicompat.AnthropicStreamEvent{
@@ -707,7 +720,7 @@ func (s *GeminiMessagesCompatService) handleChatCompletionsStreamingResponseFrom
Text: delta,
},
}) {
return &geminiStreamResult{usage: &usage, firstTokenMs: firstTokenMs}, nil
return streamResult(), nil
}
continue
}
@@ -718,11 +731,11 @@ func (s *GeminiMessagesCompatService) handleChatCompletionsStreamingResponseFrom
name = "tool"
}
if closeOpenBlock() {
return &geminiStreamResult{usage: &usage, firstTokenMs: firstTokenMs}, nil
return streamResult(), nil
}
if openToolIndex >= 0 && openToolName != name {
if closeOpenTool() {
return &geminiStreamResult{usage: &usage, firstTokenMs: firstTokenMs}, nil
return streamResult(), nil
}
}
if openToolIndex < 0 {
@@ -741,7 +754,7 @@ func (s *GeminiMessagesCompatService) handleChatCompletionsStreamingResponseFrom
Input: json.RawMessage(`{}`),
},
}) {
return &geminiStreamResult{usage: &usage, firstTokenMs: firstTokenMs}, nil
return streamResult(), nil
}
}
@@ -767,7 +780,7 @@ func (s *GeminiMessagesCompatService) handleChatCompletionsStreamingResponseFrom
PartialJSON: delta,
},
}) {
return &geminiStreamResult{usage: &usage, firstTokenMs: firstTokenMs}, nil
return streamResult(), nil
}
}
}
@@ -781,15 +794,20 @@ func (s *GeminiMessagesCompatService) handleChatCompletionsStreamingResponseFrom
break
}
if err != nil {
if geminiStreamAbortedWithBillableOutput(c, err, &usage) {
logger.LegacyPrintf("service.gemini_messages_compat", "[GeminiAPI] Client disconnected during chat completions stream, billing collected usage: %v", err)
clientDisconnect = true
return streamResult(), nil
}
return nil, fmt.Errorf("stream read error: %w", err)
}
}
if closeOpenBlock() {
return &geminiStreamResult{usage: &usage, firstTokenMs: firstTokenMs}, nil
return streamResult(), nil
}
if closeOpenTool() {
return &geminiStreamResult{usage: &usage, firstTokenMs: firstTokenMs}, nil
return streamResult(), nil
}
stopReason := mapGeminiFinishReasonToClaudeStopReason(finishReason)
@@ -810,30 +828,30 @@ func (s *GeminiMessagesCompatService) handleChatCompletionsStreamingResponseFrom
CacheReadInputTokens: usage.CacheReadInputTokens,
},
}) {
return &geminiStreamResult{usage: &usage, firstTokenMs: firstTokenMs}, nil
return streamResult(), nil
}
if emitAnthropicEvent(&apicompat.AnthropicStreamEvent{Type: "message_stop"}) {
return &geminiStreamResult{usage: &usage, firstTokenMs: firstTokenMs}, nil
return streamResult(), nil
}
for _, resEvt := range apicompat.FinalizeAnthropicResponsesStream(anthState) {
chunks := apicompat.ResponsesEventToChatChunks(&resEvt, ccState)
for _, chunk := range chunks {
if disconnected := writeChatChunk(chunk); disconnected {
return &geminiStreamResult{usage: &usage, firstTokenMs: firstTokenMs}, nil
return streamResult(), nil
}
}
}
for _, chunk := range apicompat.FinalizeResponsesChatStream(ccState) {
if disconnected := writeChatChunk(chunk); disconnected {
return &geminiStreamResult{usage: &usage, firstTokenMs: firstTokenMs}, nil
return streamResult(), nil
}
}
_, _ = io.WriteString(c.Writer, "data: [DONE]\n\n")
flusher.Flush()
return &geminiStreamResult{usage: &usage, firstTokenMs: firstTokenMs}, nil
return streamResult(), nil
}
func (s *GeminiMessagesCompatService) writeGeminiChatCompletionsMappedError(
@@ -1015,6 +1015,8 @@ func (s *GeminiMessagesCompatService) Forward(ctx context.Context, c *gin.Contex
var usage *ClaudeUsage
var firstTokenMs *int
clientDisconnect := false
streamFinished := false
if req.Stream {
streamRes, err := s.handleStreamingResponse(c, resp, startTime, originalModel)
if err != nil {
@@ -1022,6 +1024,8 @@ func (s *GeminiMessagesCompatService) Forward(ctx context.Context, c *gin.Contex
}
usage = streamRes.usage
firstTokenMs = streamRes.firstTokenMs
clientDisconnect = streamRes.clientDisconnect
streamFinished = streamRes.finished
} else {
if useUpstreamStream {
collected, usageObj, err := collectGeminiSSE(resp.Body, true)
@@ -1049,6 +1053,10 @@ func (s *GeminiMessagesCompatService) Forward(ctx context.Context, c *gin.Contex
imageInputSize := s.extractImageInputSize(body)
imageSize := normalizeOpenAIImageSizeTier(imageInputSize)
imageCount := resolveGeminiImageCount(c, originalModel, mappedModel)
// 客户端在收到结束分片前断开时只计已送达的图片:Claude 兼容流不向客户端转发图片。
if clientDisconnect && !streamFinished {
imageCount = 0
}
return &ForwardResult{
RequestID: requestID,
@@ -1062,6 +1070,7 @@ func (s *GeminiMessagesCompatService) Forward(ctx context.Context, c *gin.Contex
Duration: time.Since(startTime),
FirstTokenMs: firstTokenMs,
ReasoningEffort: extractGeminiReasoningEffortFromBody(geminiReq),
ClientDisconnect: clientDisconnect,
ImageCount: imageCount,
ImageSize: imageSize,
ImageInputSize: imageInputSize,
@@ -1530,6 +1539,8 @@ func (s *GeminiMessagesCompatService) ForwardNative(ctx context.Context, c *gin.
var usage *ClaudeUsage
var firstTokenMs *int
clientDisconnect := false
streamFinished := false
if stream {
streamRes, err := s.handleNativeStreamingResponse(c, resp, startTime, isOAuth, account, requestID)
@@ -1538,6 +1549,8 @@ func (s *GeminiMessagesCompatService) ForwardNative(ctx context.Context, c *gin.
}
usage = streamRes.usage
firstTokenMs = streamRes.firstTokenMs
clientDisconnect = streamRes.clientDisconnect
streamFinished = streamRes.finished
} else {
if useUpstreamStream {
var best geminiResponseSignal
@@ -1572,6 +1585,10 @@ func (s *GeminiMessagesCompatService) ForwardNative(ctx context.Context, c *gin.
imageInputSize := s.extractImageInputSize(body)
imageSize := normalizeOpenAIImageSizeTier(imageInputSize)
imageCount := resolveGeminiImageCount(c, originalModel, mappedModel)
// 客户端在收到结束分片前断开时只计已送达的图片。
if clientDisconnect && !streamFinished {
imageCount = observedGeminiImageOutputs(c)
}
return &ForwardResult{
RequestID: requestID,
@@ -1585,6 +1602,7 @@ func (s *GeminiMessagesCompatService) ForwardNative(ctx context.Context, c *gin.
ReasoningEffort: extractGeminiReasoningEffortFromBody(body),
Duration: time.Since(startTime),
FirstTokenMs: firstTokenMs,
ClientDisconnect: clientDisconnect,
ImageCount: imageCount,
ImageSize: imageSize,
ImageInputSize: imageInputSize,
@@ -2068,8 +2086,10 @@ func mapGeminiStatusToClaudeErrorType(status string) string {
}
type geminiStreamResult struct {
usage *ClaudeUsage
firstTokenMs *int
usage *ClaudeUsage
firstTokenMs *int
clientDisconnect bool // 客户端在流式传输中途断开,结果为断开前已收集的部分
finished bool // 断开前已收到上游 finishReason
}
func (s *GeminiMessagesCompatService) handleNonStreamingResponse(c *gin.Context, resp *http.Response, originalModel string) (*ClaudeUsage, error) {
@@ -2151,6 +2171,15 @@ func (s *GeminiMessagesCompatService) handleStreamingResponse(c *gin.Context, re
for {
line, err := reader.ReadString('\n')
if err != nil && !errors.Is(err, io.EOF) {
if geminiStreamAbortedWithBillableOutput(c, err, &usage) {
logger.LegacyPrintf("service.gemini_messages_compat", "[GeminiAPI] Client disconnected during messages stream, billing collected usage: %v", err)
return &geminiStreamResult{
usage: &usage,
firstTokenMs: firstTokenMs,
clientDisconnect: true,
finished: finishReason != "",
}, nil
}
return nil, fmt.Errorf("stream read error: %w", err)
}
@@ -2660,8 +2689,10 @@ func mergeCollectedTextParts(response map[string]any, textParts []string) map[st
}
type geminiNativeStreamResult struct {
usage *ClaudeUsage
firstTokenMs *int
usage *ClaudeUsage
firstTokenMs *int
clientDisconnect bool // 客户端在流式传输中途断开,结果为断开前已收集的部分
finished bool // 断开前已收到上游 finishReason
}
func isGeminiInsufficientScope(headers http.Header, body []byte) bool {
@@ -2838,10 +2869,13 @@ func (s *GeminiMessagesCompatService) handleNativeStreamingResponse(c *gin.Conte
var best geminiResponseSignal
sawDataEvent := false
fallback := &geminiSSEFallbackBody{}
clientDisconnect := false
finished := false
for {
line, err := reader.ReadString('\n')
if len(line) > 0 {
// 读错误时 line 是未以换行结尾的残行,不透传也不参与用量统计;EOF 时的末行完整有效。
if len(line) > 0 && (err == nil || errors.Is(err, io.EOF)) {
trimmed := strings.TrimRight(line, "\r\n")
if strings.HasPrefix(trimmed, "data:") {
payload := strings.TrimSpace(strings.TrimPrefix(trimmed, "data:"))
@@ -2874,6 +2908,10 @@ func (s *GeminiMessagesCompatService) handleNativeStreamingResponse(c *gin.Conte
observer.ObserveGemini(rawBytes)
observeGeminiImageOutputs(c, rawBytes)
if gjson.GetBytes(rawBytes, "candidates.0.finishReason").String() != "" {
finished = true
}
if firstTokenMs == nil {
ms := int(time.Since(startTime).Milliseconds())
firstTokenMs = &ms
@@ -2901,13 +2939,18 @@ func (s *GeminiMessagesCompatService) handleNativeStreamingResponse(c *gin.Conte
break
}
if err != nil {
return nil, err
if !geminiStreamAbortedWithBillableOutput(c, err, usage) {
return nil, err
}
clientDisconnect = true
logger.LegacyPrintf("service.gemini_messages_compat", "[GeminiAPI] Client disconnected during native stream (account=%d), billing collected usage: %v", account.ID, err)
break
}
}
s.finalizeGeminiSSESignal(c, account, true, upstreamRequestID, best, sawDataEvent, fallback)
return &geminiNativeStreamResult{usage: usage, firstTokenMs: firstTokenMs}, nil
return &geminiNativeStreamResult{usage: usage, firstTokenMs: firstTokenMs, clientDisconnect: clientDisconnect, finished: finished}, nil
}
// unwrapGeminiResponse 解包 Gemini OAuth 响应中的 response 字段
@@ -3036,9 +3079,10 @@ func extractGeminiUsage(data []byte) *ClaudeUsage {
}
// 注意:Gemini 的 promptTokenCount 包含 cachedContentTokenCount,
// 但 Claude 的 input_tokens 不包含 cache_read_input_tokens,需要减去
// 但 Claude 的 input_tokens 不包含 cache_read_input_tokens,需要减去。
// prompt 被拦截时上游可能只报 cachedContentTokenCount,差值下限为 0。
return &ClaudeUsage{
InputTokens: prompt - cached,
InputTokens: max(prompt-cached, 0),
OutputTokens: cand + thoughts,
CacheReadInputTokens: cached,
ImageOutputTokens: imageTokens,
@@ -947,6 +947,16 @@ func TestExtractGeminiUsage(t *testing.T) {
CacheReadInputTokens: 0,
},
},
{
name: "prompt 被拦截只报 cachedContentTokenCount — input 下限为 0",
input: `{"promptFeedback":{"blockReason":"PROHIBITED_CONTENT"},"usageMetadata":{"cachedContentTokenCount":12980}}`,
wantNil: false,
wantUsage: &ClaudeUsage{
InputTokens: 0,
OutputTokens: 0,
CacheReadInputTokens: 12980,
},
},
{
name: "无 usageMetadata",
input: `{"candidates":[]}`,
@@ -180,6 +180,29 @@ func TestGeminiForwardNative_StreamPromptBlockedMarksContentPolicy(t *testing.T)
require.Empty(t, upstreamErrorEventsFromContext(t, c))
}
// Vertex 拦截 prompt 时 usageMetadata 可能只带 cachedContentTokenCount、缺 promptTokenCount:
// input 下限为 0,缓存命中照上游所报计入 cache_read。
func TestGeminiForwardNative_StreamPromptBlockedWithOnlyCachedTokensKeepsInputNonNegative(t *testing.T) {
gin.SetMode(gin.TestMode)
body := `data: {"promptFeedback":{"blockReason":"PROHIBITED_CONTENT"},"usageMetadata":{"cachedContentTokenCount":12980,"totalTokenCount":0,"trafficType":"ON_DEMAND"},"modelVersion":"gemini-3.5-flash-lite"}
`
svc := newGeminiSignalService("text/event-stream", body)
c, rec := newGeminiNativeTestContext(t)
result, err := svc.ForwardNative(context.Background(), c, geminiSignalTestAccount(),
"gemini-3.5-flash-lite", "streamGenerateContent", true, geminiSignalTestRequest())
require.NoError(t, err)
require.Equal(t, body, rec.Body.String())
require.Equal(t, 0, result.Usage.InputTokens)
require.Equal(t, 12980, result.Usage.CacheReadInputTokens)
require.Equal(t, 0, result.Usage.OutputTokens)
streamErrs := GetOpsStreamErrors(c)
require.Len(t, streamErrs, 1)
require.Equal(t, "PROHIBITED_CONTENT", streamErrs[0].Code)
}
func TestGeminiForwardNative_EmptyStreamMarksUpstreamFailure(t *testing.T) {
gin.SetMode(gin.TestMode)
svc := newGeminiSignalService("text/event-stream", "")
@@ -0,0 +1,25 @@
package service
import (
"context"
"errors"
"io"
"github.com/gin-gonic/gin"
)
// geminiStreamAbortedWithBillableOutput 判断流式读取中断是否由客户端断开引起,且断开前
// 已观测到可计费的输出(token 用量或已回吐的图片)。满足时调用方以已收集的结果收尾并照常计费。
// usageMetadata 为累计值,按断开前收到的最后一份计费:读到结束分片后断开即完整用量;
// 中途断开时中间分片可能只含输入计数或不含用量,此时按已有计数计费或不计费。
func geminiStreamAbortedWithBillableOutput(c *gin.Context, err error, usage *ClaudeUsage) bool {
if err == nil || errors.Is(err, io.EOF) {
return false
}
clientGone := errors.Is(err, context.Canceled) ||
(c != nil && c.Request != nil && c.Request.Context().Err() != nil)
if !clientGone {
return false
}
return usage.hasObservedTokens() || observedGeminiImageOutputs(c) > 0
}
@@ -0,0 +1,244 @@
//go:build unit
package service
import (
"bytes"
"context"
"io"
"net/http"
"net/http/httptest"
"testing"
"github.com/gin-gonic/gin"
"github.com/stretchr/testify/require"
)
// geminiAbortingStreamBody 先吐出 data,读完后返回 abortErr;onAbort 用于模拟客户端断开时请求 context 被取消。
type geminiAbortingStreamBody struct {
data *bytes.Reader
abortErr error
onAbort func()
}
func (b *geminiAbortingStreamBody) Read(p []byte) (int, error) {
n, err := b.data.Read(p)
if err == io.EOF {
if b.onAbort != nil {
b.onAbort()
}
return n, b.abortErr
}
return n, err
}
func (b *geminiAbortingStreamBody) Close() error { return nil }
const (
geminiAbortTextChunk = `data: {"candidates":[{"content":{"parts":[{"text":"hello"}],"role":"model"}}]}` + "\n\n"
geminiAbortFinalChunk = `data: {"candidates":[{"content":{"parts":[{"text":" world"}],"role":"model"},"finishReason":"STOP"}],"usageMetadata":{"promptTokenCount":10,"candidatesTokenCount":5,"totalTokenCount":15}}` + "\n\n"
geminiAbortImageChunk = `data: {"candidates":[{"content":{"parts":[{"inlineData":{"mimeType":"image/png","data":"aGVsbG8="}}],"role":"model"},"finishReason":"STOP"}],"usageMetadata":{"promptTokenCount":10,"candidatesTokenCount":1290,"totalTokenCount":1300}}` + "\n\n"
geminiAbortUsageChunk = `data: {"candidates":[{"content":{"parts":[{"text":"drawing"}],"role":"model"}}],"usageMetadata":{"promptTokenCount":10,"candidatesTokenCount":2,"totalTokenCount":12}}` + "\n\n"
)
// newGeminiAbortingStreamService 返回的上游流在吐完 sse 后中断;clientGone=true 时中断伴随请求 context 取消。
func newGeminiAbortingStreamService(t *testing.T, path string, reqBody []byte, sse string, clientGone bool) (*GeminiMessagesCompatService, *gin.Context, *httptest.ResponseRecorder) {
t.Helper()
gin.SetMode(gin.TestMode)
rec := httptest.NewRecorder()
c, _ := gin.CreateTestContext(rec)
reqCtx, cancel := context.WithCancel(context.Background())
t.Cleanup(cancel)
c.Request = httptest.NewRequest(http.MethodPost, path, bytes.NewReader(reqBody)).WithContext(reqCtx)
body := &geminiAbortingStreamBody{data: bytes.NewReader([]byte(sse)), abortErr: io.ErrUnexpectedEOF}
if clientGone {
body.abortErr = context.Canceled
body.onAbort = cancel
}
svc, httpStub := newGeminiVertexAPIKeyTestService()
httpStub.response = &http.Response{
StatusCode: http.StatusOK,
Header: http.Header{"Content-Type": []string{"text/event-stream"}},
Body: body,
}
return svc, c, rec
}
func TestGeminiForwardNative_ClientDisconnectAfterFinalChunkBillsCollectedUsage(t *testing.T) {
reqBody := []byte(`{"contents":[{"role":"user","parts":[{"text":"hi"}]}]}`)
svc, c, _ := newGeminiAbortingStreamService(t, "/v1beta/models/gemini-3-flash-preview:streamGenerateContent", reqBody, geminiAbortTextChunk+geminiAbortFinalChunk, true)
result, err := svc.ForwardNative(context.Background(), c, geminiVertexAPIKeyTestAccount(""), "gemini-3-flash-preview", "streamGenerateContent", true, reqBody)
require.NoError(t, err)
require.NotNil(t, result)
require.True(t, result.ClientDisconnect)
require.Equal(t, 10, result.Usage.InputTokens)
require.Equal(t, 5, result.Usage.OutputTokens)
require.Zero(t, result.ImageCount)
}
func TestGeminiForwardNative_ClientDisconnectWithoutUsageStaysUnbilled(t *testing.T) {
reqBody := []byte(`{"contents":[{"role":"user","parts":[{"text":"hi"}]}]}`)
svc, c, _ := newGeminiAbortingStreamService(t, "/v1beta/models/gemini-3-flash-preview:streamGenerateContent", reqBody, geminiAbortTextChunk, true)
result, err := svc.ForwardNative(context.Background(), c, geminiVertexAPIKeyTestAccount(""), "gemini-3-flash-preview", "streamGenerateContent", true, reqBody)
require.ErrorIs(t, err, context.Canceled)
require.Nil(t, result)
}
func TestGeminiForwardNative_UpstreamReadErrorWithLiveClientStaysError(t *testing.T) {
reqBody := []byte(`{"contents":[{"role":"user","parts":[{"text":"hi"}]}]}`)
svc, c, _ := newGeminiAbortingStreamService(t, "/v1beta/models/gemini-3-flash-preview:streamGenerateContent", reqBody, geminiAbortTextChunk+geminiAbortFinalChunk, false)
result, err := svc.ForwardNative(context.Background(), c, geminiVertexAPIKeyTestAccount(""), "gemini-3-flash-preview", "streamGenerateContent", true, reqBody)
require.ErrorIs(t, err, io.ErrUnexpectedEOF)
require.Nil(t, result)
}
func TestGeminiForwardNative_ClientDisconnectBillsOnlyDeliveredImages(t *testing.T) {
reqBody := []byte(`{"contents":[{"role":"user","parts":[{"text":"draw a cat"}]}]}`)
t.Run("image delivered before disconnect", func(t *testing.T) {
svc, c, _ := newGeminiAbortingStreamService(t, "/v1beta/models/gemini-3-pro-image:streamGenerateContent", reqBody, geminiAbortImageChunk, true)
result, err := svc.ForwardNative(context.Background(), c, geminiVertexAPIKeyTestAccount(""), "gemini-3-pro-image", "streamGenerateContent", true, reqBody)
require.NoError(t, err)
require.True(t, result.ClientDisconnect)
require.Equal(t, 1, result.ImageCount)
})
t.Run("disconnect before the image arrives", func(t *testing.T) {
svc, c, _ := newGeminiAbortingStreamService(t, "/v1beta/models/gemini-3-pro-image:streamGenerateContent", reqBody, geminiAbortUsageChunk, true)
result, err := svc.ForwardNative(context.Background(), c, geminiVertexAPIKeyTestAccount(""), "gemini-3-pro-image", "streamGenerateContent", true, reqBody)
require.NoError(t, err)
require.True(t, result.ClientDisconnect)
require.Zero(t, result.ImageCount)
require.Equal(t, 2, result.Usage.OutputTokens)
})
}
func TestGeminiCompatForward_ClientDisconnectAfterFinalChunkBillsCollectedUsage(t *testing.T) {
reqBody := []byte(`{"model":"gemini-3-flash-preview","max_tokens":16,"stream":true,"messages":[{"role":"user","content":"hi"}]}`)
t.Run("client gone", func(t *testing.T) {
svc, c, _ := newGeminiAbortingStreamService(t, "/v1/messages", reqBody, geminiAbortTextChunk+geminiAbortFinalChunk, true)
result, err := svc.Forward(context.Background(), c, geminiVertexAPIKeyTestAccount(""), reqBody)
require.NoError(t, err)
require.True(t, result.ClientDisconnect)
require.Equal(t, 10, result.Usage.InputTokens)
require.Equal(t, 5, result.Usage.OutputTokens)
})
t.Run("upstream read error with live client", func(t *testing.T) {
svc, c, _ := newGeminiAbortingStreamService(t, "/v1/messages", reqBody, geminiAbortTextChunk+geminiAbortFinalChunk, false)
result, err := svc.Forward(context.Background(), c, geminiVertexAPIKeyTestAccount(""), reqBody)
require.ErrorIs(t, err, io.ErrUnexpectedEOF)
require.Nil(t, result)
})
}
func TestGeminiForwardAsChatCompletions_ClientDisconnectAfterFinalChunkBillsCollectedUsage(t *testing.T) {
reqBody := []byte(`{"model":"gemini-3-flash-preview","stream":true,"messages":[{"role":"user","content":"hi"}]}`)
t.Run("client gone", func(t *testing.T) {
svc, c, _ := newGeminiAbortingStreamService(t, "/v1/chat/completions", reqBody, geminiAbortTextChunk+geminiAbortFinalChunk, true)
result, err := svc.ForwardAsChatCompletions(context.Background(), c, geminiVertexAPIKeyTestAccount(""), reqBody)
require.NoError(t, err)
require.True(t, result.ClientDisconnect)
require.Equal(t, 10, result.Usage.InputTokens)
require.Equal(t, 5, result.Usage.OutputTokens)
})
t.Run("client gone before usage", func(t *testing.T) {
svc, c, _ := newGeminiAbortingStreamService(t, "/v1/chat/completions", reqBody, geminiAbortTextChunk, true)
result, err := svc.ForwardAsChatCompletions(context.Background(), c, geminiVertexAPIKeyTestAccount(""), reqBody)
require.ErrorIs(t, err, context.Canceled)
require.Nil(t, result)
})
}
func TestGeminiForwardNative_ClientDisconnectAfterFinishKeepsImageFallback(t *testing.T) {
// 结束分片已到达但图片不以 inlineData 回吐时,断开与正常完成同口径按模型名兜底。
finishWithoutInline := `data: {"candidates":[{"content":{"parts":[{"fileData":{"mimeType":"image/png","fileUri":"gs://bucket/cat.png"}}],"role":"model"},"finishReason":"STOP"}],"usageMetadata":{"promptTokenCount":10,"candidatesTokenCount":1290,"totalTokenCount":1300}}` + "\n\n"
reqBody := []byte(`{"contents":[{"role":"user","parts":[{"text":"draw a cat"}]}]}`)
svc, c, _ := newGeminiAbortingStreamService(t, "/v1beta/models/gemini-3-pro-image:streamGenerateContent", reqBody, finishWithoutInline, true)
result, err := svc.ForwardNative(context.Background(), c, geminiVertexAPIKeyTestAccount(""), "gemini-3-pro-image", "streamGenerateContent", true, reqBody)
require.NoError(t, err)
require.True(t, result.ClientDisconnect)
require.Equal(t, 1, result.ImageCount)
}
func TestGeminiForwardNative_ClientDisconnectIgnoresTruncatedTrailingLine(t *testing.T) {
complete := `data: {"candidates":[{"content":{"parts":[{"text":"hello"}],"role":"model"}}],"usageMetadata":{"promptTokenCount":1000,"candidatesTokenCount":300,"totalTokenCount":1300}}` + "\n\n"
truncated := `data: {"candidates":[{"content":{"parts":[{"text":" more"}],"role":"model"}}],"usageMetadata":{`
reqBody := []byte(`{"contents":[{"role":"user","parts":[{"text":"hi"}]}]}`)
svc, c, _ := newGeminiAbortingStreamService(t, "/v1beta/models/gemini-3-flash-preview:streamGenerateContent", reqBody, complete+truncated, true)
result, err := svc.ForwardNative(context.Background(), c, geminiVertexAPIKeyTestAccount(""), "gemini-3-flash-preview", "streamGenerateContent", true, reqBody)
require.NoError(t, err)
require.True(t, result.ClientDisconnect)
require.Equal(t, 1000, result.Usage.InputTokens)
require.Equal(t, 300, result.Usage.OutputTokens)
}
func TestGeminiForwardAsChatCompletions_ClientDisconnectImageCount(t *testing.T) {
t.Run("disconnect before finish bills no image", func(t *testing.T) {
reqBody := []byte(`{"model":"gemini-3-pro-image","stream":true,"messages":[{"role":"user","content":"draw a cat"}]}`)
svc, c, _ := newGeminiAbortingStreamService(t, "/v1/chat/completions", reqBody, geminiAbortUsageChunk, true)
result, err := svc.ForwardAsChatCompletions(context.Background(), c, geminiVertexAPIKeyTestAccount(""), reqBody)
require.NoError(t, err)
require.True(t, result.ClientDisconnect)
require.Zero(t, result.ImageCount)
require.Equal(t, 2, result.Usage.OutputTokens)
})
t.Run("disconnect after finish keeps completed-request image count", func(t *testing.T) {
reqBody := []byte(`{"model":"gemini-3-pro-image","stream":true,"messages":[{"role":"user","content":"draw a cat"}]}`)
svc, c, _ := newGeminiAbortingStreamService(t, "/v1/chat/completions", reqBody, geminiAbortImageChunk, true)
result, err := svc.ForwardAsChatCompletions(context.Background(), c, geminiVertexAPIKeyTestAccount(""), reqBody)
require.NoError(t, err)
require.True(t, result.ClientDisconnect)
require.Equal(t, 1, result.ImageCount)
})
}
// geminiFailingResponseWriter 模拟客户端已断开:所有写入失败。
type geminiFailingResponseWriter struct {
header http.Header
}
func (w *geminiFailingResponseWriter) Header() http.Header { return w.header }
func (w *geminiFailingResponseWriter) Write([]byte) (int, error) { return 0, io.ErrClosedPipe }
func (w *geminiFailingResponseWriter) WriteHeader(int) {}
func (w *geminiFailingResponseWriter) Flush() {}
func TestGeminiForwardAsChatCompletions_WriteFailureOnImageModelBillsNoImage(t *testing.T) {
gin.SetMode(gin.TestMode)
reqBody := []byte(`{"model":"gemini-3-pro-image","stream":true,"messages":[{"role":"user","content":"draw a cat"}]}`)
c, _ := gin.CreateTestContext(&geminiFailingResponseWriter{header: http.Header{}})
c.Request = httptest.NewRequest(http.MethodPost, "/v1/chat/completions", bytes.NewReader(reqBody))
svc, httpStub := newGeminiVertexAPIKeyTestService()
httpStub.response = &http.Response{
StatusCode: http.StatusOK,
Header: http.Header{"Content-Type": []string{"text/event-stream"}},
Body: io.NopCloser(bytes.NewReader([]byte(geminiAbortUsageChunk))),
}
result, err := svc.ForwardAsChatCompletions(context.Background(), c, geminiVertexAPIKeyTestAccount(""), reqBody)
require.NoError(t, err)
require.True(t, result.ClientDisconnect)
require.Zero(t, result.ImageCount)
}
func TestGeminiCompatForward_ClientDisconnectBeforeFinishBillsNoImage(t *testing.T) {
inlineWithoutFinish := `data: {"candidates":[{"content":{"parts":[{"inlineData":{"mimeType":"image/png","data":"aGVsbG8="}}],"role":"model"}}],"usageMetadata":{"promptTokenCount":10,"candidatesTokenCount":1290,"totalTokenCount":1300}}` + "\n\n"
reqBody := []byte(`{"model":"gemini-3-pro-image","max_tokens":16,"stream":true,"messages":[{"role":"user","content":"draw a cat"}]}`)
svc, c, _ := newGeminiAbortingStreamService(t, "/v1/messages", reqBody, inlineWithoutFinish, true)
result, err := svc.Forward(context.Background(), c, geminiVertexAPIKeyTestAccount(""), reqBody)
require.NoError(t, err)
require.True(t, result.ClientDisconnect)
require.Zero(t, result.ImageCount)
require.Equal(t, 1290, result.Usage.OutputTokens)
}
+128 -4
View File
@@ -8,10 +8,16 @@ import (
"fmt"
"io"
"mime"
"net"
"net/http"
"net/url"
"strconv"
"strings"
"sync"
"syscall"
"time"
"github.com/Wei-Shaw/sub2api/internal/util/urlvalidator"
)
const defaultImageMaxDownloadBytes int64 = 32 << 20 // 32 MiB
@@ -53,8 +59,123 @@ func NewImageResultUploader(storage ImageStorage, prefix string, maxDownloadByte
}
}
func defaultImageDownloadHTTPClient() *http.Client {
return &http.Client{Timeout: 60 * time.Second}
// defaultImageDownloadHTTPClient 是进程内共享的图片下载客户端,复用连接池。
var defaultImageDownloadHTTPClient = sync.OnceValue(func() *http.Client {
return newImageDownloadHTTPClient(http.ProxyFromEnvironment)
})
// newImageDownloadHTTPClient 返回只允许下载公网目的地的客户端。
// 图片 url 来自上游响应,是不可信输入:目的地为 localhost、回环、私网、链路本地、
// 云元数据等地址时一律拒绝,重定向的每一跳都经同一 RoundTrip 校验。
// 不经代理时在 socket 层校验真实连接 IP(防 DNS rebinding);经环境代理时由代理解析目标,
// 只能在发出前校验本地解析结果。
func newImageDownloadHTTPClient(proxy func(*http.Request) (*url.URL, error)) *http.Client {
direct := newImageDownloadBaseTransport(imageDownloadDialControl)
proxied := newImageDownloadBaseTransport(nil)
proxied.Proxy = proxy
return &http.Client{
Timeout: 60 * time.Second,
Transport: &imageDownloadTransport{direct: direct, proxied: proxied, proxy: proxy, resolvesPrivate: isPrivateOrLoopbackHost},
}
}
// newImageDownloadBaseTransport 返回与 http.DefaultTransport 参数一致、但不带代理的 Transport。
// control 非空时在每个 socket 连接前校验目标地址。
func newImageDownloadBaseTransport(control func(network, address string, c syscall.RawConn) error) *http.Transport {
dialer := &net.Dialer{Timeout: 30 * time.Second, KeepAlive: 30 * time.Second, Control: control}
return &http.Transport{
DialContext: dialer.DialContext,
ForceAttemptHTTP2: true,
MaxIdleConns: 100,
IdleConnTimeout: 90 * time.Second,
TLSHandshakeTimeout: 10 * time.Second,
ExpectContinueTimeout: 1 * time.Second,
}
}
// imageDownloadDialControl 拒绝连接到非公网 IP。address 是 DNS 解析后的真实连接地址,
// 多个解析结果时 Dialer 会跳过被拒地址继续尝试下一个。
func imageDownloadDialControl(_, address string, _ syscall.RawConn) error {
host, _, err := net.SplitHostPort(address)
if err != nil {
return err
}
if ip := net.ParseIP(stripIPv6Zone(host)); ip == nil || isPrivateIP(ip) {
return &net.AddrError{Err: "address not allowed for image download", Addr: address}
}
return nil
}
// imageDownloadTransport 按请求是否走环境代理分派到对应 Transport,并在发出前拒绝非公网目的地。
type imageDownloadTransport struct {
direct http.RoundTripper
proxied http.RoundTripper
proxy func(*http.Request) (*url.URL, error)
// resolvesPrivate 报告 host 的本地解析结果是否含非公网地址,仅用于经代理的请求。
resolvesPrivate func(ctx context.Context, host string) (bool, error)
}
func (t *imageDownloadTransport) RoundTrip(req *http.Request) (*http.Response, error) {
next, err := t.route(req)
if err != nil {
if req.Body != nil {
_ = req.Body.Close()
}
return nil, err
}
return next.RoundTrip(req)
}
func (t *imageDownloadTransport) route(req *http.Request) (http.RoundTripper, error) {
if req.URL == nil {
return nil, errors.New("image url is empty")
}
host := req.URL.Hostname()
if isBlockedImageDownloadHost(host) {
return nil, fmt.Errorf("image url host is not allowed: %s", host)
}
var proxyURL *url.URL
if t.proxy != nil {
var err error
if proxyURL, err = t.proxy(req); err != nil {
return nil, err
}
}
if proxyURL == nil {
return t.direct, nil
}
if t.resolvesPrivate == nil {
return nil, errors.New("image download proxy host check is not configured")
}
blocked, err := t.resolvesPrivate(req.Context(), host)
if err != nil {
return nil, fmt.Errorf("resolve image url host: %w", err)
}
if blocked {
return nil, fmt.Errorf("image url host is not allowed: %s", host)
}
return t.proxied, nil
}
// isBlockedImageDownloadHost 报告 host 是否为空、被拒绝的主机名,或非公网 IP 字面量。
// 域名的解析结果由 socket 层或代理分支再校验。
func isBlockedImageDownloadHost(host string) bool {
host = strings.TrimSuffix(strings.ToLower(strings.TrimSpace(host)), ".")
if urlvalidator.IsBlockedHost(host) || isBlockedHostname(host) {
return true
}
if ip := net.ParseIP(stripIPv6Zone(host)); ip != nil {
return isPrivateIP(ip)
}
return false
}
// stripIPv6Zone 去掉 IPv6 字面量的 zone(如 fe80::1%eth0),使其可被 net.ParseIP 解析。
func stripIPv6Zone(host string) string {
if i := strings.LastIndexByte(host, '%'); i >= 0 {
return host[:i]
}
return host
}
// Rewrite 将 result(上游生图响应 JSON)里的每张图片转存到对象存储,
@@ -215,9 +336,12 @@ func (u *ImageResultUploader) download(ctx context.Context, rawURL string) ([]by
if int64(len(data)) > limit {
return nil, "", fmt.Errorf("downloaded image exceeds %d bytes", limit)
}
contentType := strings.TrimSpace(strings.Split(resp.Header.Get("Content-Type"), ";")[0])
contentType := strings.ToLower(strings.TrimSpace(strings.Split(resp.Header.Get("Content-Type"), ";")[0]))
if !strings.HasPrefix(contentType, "image/") {
contentType = detectImageContentType(data)
// 声明与嗅探都不是图片时拒绝,避免把任意响应体(内部接口的 JSON/HTML 等)转存给租户。
if contentType = detectedImageContentType(data); contentType == "" {
return nil, "", errors.New("download image: content is not an image")
}
}
return data, contentType, nil
}
+258 -7
View File
@@ -1,12 +1,17 @@
package service
import (
"bytes"
"context"
"encoding/base64"
"encoding/json"
"errors"
"io"
"net"
"net/http"
"net/http/httptest"
"net/url"
"sync/atomic"
"testing"
"time"
@@ -73,18 +78,24 @@ func TestImageResultUploaderRewritesB64JSON(t *testing.T) {
}
func TestImageResultUploaderRewritesURL(t *testing.T) {
upstream := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, _ *http.Request) {
w.Header().Set("Content-Type", "image/png")
_, _ = w.Write(pngBytes)
}))
defer upstream.Close()
var requested string
client := &http.Client{Transport: roundTripFunc(func(req *http.Request) (*http.Response, error) {
requested = req.URL.String()
return &http.Response{
StatusCode: http.StatusOK,
Header: http.Header{"Content-Type": []string{"image/png"}},
Body: io.NopCloser(bytes.NewReader(pngBytes)),
Request: req,
}, nil
})}
storage := &fakeImageStorage{}
uploader := NewImageResultUploader(storage, "images/", 0, nil)
uploader := NewImageResultUploader(storage, "images/", 0, client)
result := json.RawMessage(`{"created":1,"data":[{"url":"` + upstream.URL + `/pic.png"}]}`)
result := json.RawMessage(`{"created":1,"data":[{"url":"https://images.example.test/pic.png"}]}`)
out, err := uploader.Rewrite(context.Background(), "imgtask_xyz", result)
require.NoError(t, err)
require.Equal(t, "https://images.example.test/pic.png", requested)
require.Len(t, storage.saved, 1)
require.Equal(t, pngBytes, storage.saved[0].data)
@@ -244,3 +255,243 @@ func TestImageTaskServiceCompleteOffloadFailureMarksFailed(t *testing.T) {
require.Contains(t, string(got.Error), "object storage")
require.NotContains(t, string(got.Result), "b64_json", "failed offload must not persist base64 to Redis")
}
func TestImageResultUploaderDefaultClientRejectsPrivateURL(t *testing.T) {
var hits atomic.Int32
internal := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, _ *http.Request) {
hits.Add(1)
_, _ = w.Write([]byte("GIF89a-payload"))
}))
defer internal.Close()
storage := &fakeImageStorage{}
uploader := NewImageResultUploader(storage, "images/", 0, nil)
result := json.RawMessage(`{"data":[{"url":"` + internal.URL + `/a.gif"}]}`)
_, err := uploader.Rewrite(context.Background(), "imgtask_private", result)
require.ErrorContains(t, err, "image url host is not allowed")
require.Zero(t, hits.Load(), "internal service must not be contacted")
require.Empty(t, storage.saved)
}
func TestImageTaskServiceCompletePrivateURLMarksFailed(t *testing.T) {
var hits atomic.Int32
internal := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, _ *http.Request) {
hits.Add(1)
_, _ = w.Write(pngBytes)
}))
defer internal.Close()
store := &imageTaskMemoryStore{}
storage := &fakeImageStorage{}
svc := NewImageTaskServiceWithUploader(store, NewImageResultUploader(storage, "images/", 0, nil), time.Hour, time.Minute)
owner := ImageTaskOwner{UserID: 1, APIKeyID: 2}
created, err := svc.Create(context.Background(), owner)
require.NoError(t, err)
result := json.RawMessage(`{"data":[{"url":"` + internal.URL + `/a.gif"}]}`)
require.NoError(t, svc.Complete(context.Background(), created.ID, http.StatusOK, result))
got, err := svc.Get(context.Background(), owner, created.ID)
require.NoError(t, err)
require.Equal(t, ImageTaskStatusFailed, got.Status)
require.Empty(t, got.ImageURL)
require.NotContains(t, string(got.Result), internal.URL)
require.Zero(t, hits.Load())
require.Empty(t, storage.saved)
}
func newCountingImageDownloadTransport(proxyURL *url.URL) (*imageDownloadTransport, *atomic.Int32, *atomic.Int32) {
var directCalls, proxiedCalls atomic.Int32
respond := func(counter *atomic.Int32) http.RoundTripper {
return roundTripFunc(func(req *http.Request) (*http.Response, error) {
counter.Add(1)
return &http.Response{StatusCode: http.StatusOK, Body: io.NopCloser(bytes.NewReader(pngBytes)), Request: req}, nil
})
}
return &imageDownloadTransport{
direct: respond(&directCalls),
proxied: respond(&proxiedCalls),
proxy: func(*http.Request) (*url.URL, error) { return proxyURL, nil },
resolvesPrivate: func(ctx context.Context, host string) (bool, error) {
if host == "intranet.example.test" {
return true, nil
}
if host == "unresolvable.example.test" {
return false, errors.New("no such host")
}
return isPrivateOrLoopbackHost(ctx, host)
},
}, &directCalls, &proxiedCalls
}
func TestImageDownloadTransportRejectsNonPublicHosts(t *testing.T) {
hosts := []string{
"http://127.0.0.1:18082/a.gif",
"http://localhost/a.png",
"http://LOCALHOST./a.png",
"http://foo.localhost/a.png",
"http://10.0.0.8/a.png",
"http://172.16.3.4/a.png",
"http://192.168.1.1/a.png",
"http://100.64.0.1/a.png",
"http://169.254.169.254/latest/meta-data/",
"http://metadata.google.internal/computeMetadata/v1/",
"http://0.0.0.0/a.png",
"http://[::1]/a.png",
"http://[::ffff:127.0.0.1]/a.png",
"http://[fd00::1]/a.png",
"http://[fe80::1%25eth0]/a.png",
"http:///a.png",
}
for _, proxied := range []bool{false, true} {
for _, rawURL := range hosts {
var proxyURL *url.URL
if proxied {
proxyURL = &url.URL{Scheme: "http", Host: "127.0.0.1:7890"}
}
transport, directCalls, proxiedCalls := newCountingImageDownloadTransport(proxyURL)
req, err := http.NewRequest(http.MethodGet, rawURL, nil)
require.NoError(t, err)
_, err = transport.RoundTrip(req)
require.Error(t, err, "proxied=%v url=%s", proxied, rawURL)
require.Zero(t, directCalls.Load(), rawURL)
require.Zero(t, proxiedCalls.Load(), rawURL)
}
}
}
func TestImageDownloadTransportRoutesPublicHosts(t *testing.T) {
transport, directCalls, proxiedCalls := newCountingImageDownloadTransport(nil)
req, err := http.NewRequest(http.MethodGet, "https://203.0.113.10/a.png", nil)
require.NoError(t, err)
resp, err := transport.RoundTrip(req)
require.NoError(t, err)
_ = resp.Body.Close()
require.EqualValues(t, 1, directCalls.Load())
require.Zero(t, proxiedCalls.Load())
transport, directCalls, proxiedCalls = newCountingImageDownloadTransport(&url.URL{Scheme: "http", Host: "127.0.0.1:7890"})
req, err = http.NewRequest(http.MethodGet, "https://203.0.113.10/a.png", nil)
require.NoError(t, err)
resp, err = transport.RoundTrip(req)
require.NoError(t, err)
_ = resp.Body.Close()
require.Zero(t, directCalls.Load())
require.EqualValues(t, 1, proxiedCalls.Load(), "a loopback env proxy must still be usable for public targets")
}
func TestImageDownloadClientRejectsRedirectToPrivateHost(t *testing.T) {
var calls atomic.Int32
transport := &imageDownloadTransport{
direct: roundTripFunc(func(req *http.Request) (*http.Response, error) {
calls.Add(1)
return &http.Response{
StatusCode: http.StatusFound,
Header: http.Header{"Location": []string{"http://169.254.169.254/latest/meta-data/"}},
Body: io.NopCloser(bytes.NewReader(nil)),
Request: req,
}, nil
}),
}
uploader := NewImageResultUploader(&fakeImageStorage{}, "images/", 0, &http.Client{Transport: transport})
_, err := uploader.Rewrite(context.Background(), "imgtask_redirect", json.RawMessage(`{"data":[{"url":"https://images.example.test/a.png"}]}`))
require.ErrorContains(t, err, "image url host is not allowed: 169.254.169.254")
require.EqualValues(t, 1, calls.Load(), "the private redirect hop must not be sent")
}
func TestImageDownloadClientDialsOnlyPublicAddresses(t *testing.T) {
var hits atomic.Int32
internal := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, _ *http.Request) {
hits.Add(1)
_, _ = w.Write(pngBytes)
}))
defer internal.Close()
client := newImageDownloadHTTPClient(nil)
direct, ok := client.Transport.(*imageDownloadTransport).direct.(*http.Transport)
require.True(t, ok)
require.Nil(t, direct.Proxy, "the direct transport must not pick up an environment proxy")
// 绕过 RoundTrip 预检直接 dial:IP 字面量与经 DNS 解析到回环的域名都必须在 socket 层被拒。
conn, err := direct.DialContext(context.Background(), "tcp", internal.Listener.Addr().String())
if conn != nil {
_ = conn.Close()
}
require.ErrorContains(t, err, "address not allowed for image download")
_, port, err := net.SplitHostPort(internal.Listener.Addr().String())
require.NoError(t, err)
conn, err = direct.DialContext(context.Background(), "tcp", net.JoinHostPort("localhost", port))
if conn != nil {
_ = conn.Close()
}
require.ErrorContains(t, err, "address not allowed for image download")
require.Zero(t, hits.Load())
}
func newStaticImageDownloadClient(contentType string, body []byte) *http.Client {
return &http.Client{Transport: roundTripFunc(func(req *http.Request) (*http.Response, error) {
header := http.Header{}
if contentType != "" {
header.Set("Content-Type", contentType)
}
return &http.Response{StatusCode: http.StatusOK, Header: header, Body: io.NopCloser(bytes.NewReader(body)), Request: req}, nil
})}
}
func TestImageResultUploaderDownloadContentTypes(t *testing.T) {
avif := []byte("\x00\x00\x00\x1cftypavif-not-sniffable")
tests := []struct {
name string
contentType string
body []byte
wantType string
wantErr string
}{
{name: "declared png", contentType: "image/png", body: pngBytes, wantType: "image/png"},
{name: "octet-stream sniffed png", contentType: "application/octet-stream", body: pngBytes, wantType: "image/png"},
{name: "missing header sniffed png", body: pngBytes, wantType: "image/png"},
{name: "declared avif not sniffable", contentType: "Image/AVIF; charset=binary", body: avif, wantType: "image/avif"},
{name: "json body", contentType: "application/json", body: []byte(`{"ok":true}`), wantErr: "content is not an image"},
{name: "html body", contentType: "text/html", body: []byte("<html>admin</html>"), wantErr: "content is not an image"},
{name: "empty body", body: nil, wantErr: "content is not an image"},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
storage := &fakeImageStorage{}
uploader := NewImageResultUploader(storage, "images/", 0, newStaticImageDownloadClient(tt.contentType, tt.body))
_, err := uploader.Rewrite(context.Background(), "imgtask_ct", json.RawMessage(`{"data":[{"url":"https://images.example.test/a"}]}`))
if tt.wantErr != "" {
require.ErrorContains(t, err, tt.wantErr)
require.Empty(t, storage.saved)
return
}
require.NoError(t, err)
require.Len(t, storage.saved, 1)
require.Equal(t, tt.wantType, storage.saved[0].contentType)
})
}
}
func TestImageDownloadDialControl(t *testing.T) {
for _, address := range []string{"127.0.0.1:443", "10.1.2.3:80", "100.100.100.200:80", "169.254.169.254:80", "[::1]:443", "[fe80::1%eth0]:443", "[::ffff:192.168.0.1]:443", "bad"} {
require.Error(t, imageDownloadDialControl("tcp", address, nil), address)
}
for _, address := range []string{"203.0.113.10:443", "[2001:db8::1]:443"} {
require.NoError(t, imageDownloadDialControl("tcp", address, nil), address)
}
}
func TestImageDownloadTransportProxiedChecksResolvedHost(t *testing.T) {
for _, host := range []string{"intranet.example.test", "unresolvable.example.test"} {
transport, directCalls, proxiedCalls := newCountingImageDownloadTransport(&url.URL{Scheme: "http", Host: "127.0.0.1:7890"})
req, err := http.NewRequest(http.MethodGet, "https://"+host+"/a.png", nil)
require.NoError(t, err)
_, err = transport.RoundTrip(req)
require.Error(t, err, host)
require.Zero(t, directCalls.Load(), host)
require.Zero(t, proxiedCalls.Load(), "a proxied request must fail closed when the target is private or unresolvable: %s", host)
}
}