Support Qiniu provider with OpenAI API compatibility#4256
Open
liangchaoboy wants to merge 4 commits intoQuantumNous:mainfrom
Open
Support Qiniu provider with OpenAI API compatibility#4256liangchaoboy wants to merge 4 commits intoQuantumNous:mainfrom
liangchaoboy wants to merge 4 commits intoQuantumNous:mainfrom
Conversation
Contributor
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Organization UI Review profile: CHILL Plan: Pro Run ID: 📒 Files selected for processing (3)
🚧 Files skipped from review as they are similar to previous changes (2)
WalkthroughAdds Qiniu as channel type 58: constant, base URL, name; maps it to OpenAI API type; enables stream options and upstream model fetching for Qiniu; updates frontend channel option and icon; and adds a unit test for fetching Qiniu /v1/models. Changes
Sequence Diagram(s)sequenceDiagram
participant User
participant Frontend
participant Backend
participant RelayAdaptor
participant QiniuAPI
User->>Frontend: Open channel edit / request model list
Frontend->>Backend: GET /channels/:id/models (channelType=58)
Backend->>RelayAdaptor: fetchChannelUpstreamModelIDs(channel)
RelayAdaptor->>QiniuAPI: GET /v1/models (Authorization: Bearer <key>)
QiniuAPI-->>RelayAdaptor: 200 JSON {data: [...]}
RelayAdaptor-->>Backend: parsed model ID list
Backend-->>Frontend: model list response
Frontend-->>User: display upstream models
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🧪 Generate unit tests (beta)
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
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@controller/channel_upstream_update_qiniu_test.go`:
- Around line 17-25: The handler in channel_upstream_update_qiniu_test.go is
calling t.Fatalf from the server goroutine (lines checking r.Method, r.URL.Path,
and Authorization); instead, capture the observed values (method, path, auth
header, etc.) into local variables or send them over a channel/sync struct from
the handler, return a normal HTTP response, then after invoking
fetchChannelUpstreamModelIDs in the test goroutine assert those captured values
with t.Fatalf/require. Locate the inline HTTP handler in the test and change its
t.Fatalf calls to store values (or send them on a channel) and perform the
assertions in the test goroutine after fetchChannelUpstreamModelIDs returns.
In `@relay/common/relay_info.go`:
- Line 308: Remove constant.ChannelTypeQiniu from the streamSupportedChannels
map so Qiniu isn't treated as supporting stream_options; locate the map named
streamSupportedChannels and delete the entry "constant.ChannelTypeQiniu: true"
(or conditionally exclude ChannelTypeQiniu wherever streamSupportedChannels is
used) to prevent sending StreamOptions/include_usage to Qiniu's API.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: a75ab5aa-124a-48e2-b342-39d5094ca018
📒 Files selected for processing (8)
common/api_type.goconstant/channel.gocontroller/channel_upstream_update_qiniu_test.gorelay/channel/openai/adaptor.gorelay/common/relay_info.goweb/src/components/table/channels/modals/EditChannelModal.jsxweb/src/constants/channel.constants.jsweb/src/helpers/render.jsx
Comment on lines
+17
to
+25
| if r.Method != http.MethodGet { | ||
| t.Fatalf("unexpected method: %s", r.Method) | ||
| } | ||
| if r.URL.Path != "/v1/models" { | ||
| t.Fatalf("unexpected path: %s", r.URL.Path) | ||
| } | ||
| if got := r.Header.Get("Authorization"); got != "Bearer test-key" { | ||
| t.Fatalf("unexpected Authorization header: %q", got) | ||
| } |
Contributor
There was a problem hiding this comment.
🧩 Analysis chain
🏁 Script executed:
#!/bin/bash
# Verify Fatal*/FailNow usage in httptest handler contexts (read-only check).
rg -n --type=go -C4 'httptest\.NewServer\(http\.HandlerFunc\(func' controller/channel_upstream_update_qiniu_test.go
rg -n --type=go -C3 '\bt\.(Fatal|Fatalf|FailNow)\(' controller/channel_upstream_update_qiniu_test.goRepository: QuantumNous/new-api
Length of output: 1291
🏁 Script executed:
#!/bin/bash
# Search for other httptest.NewServer patterns in test files to identify if this is systemic
rg -n --type=go 'httptest\.NewServer\(http\.HandlerFunc' --glob='*_test.go' | head -20
# Check how many test files use this pattern
rg -l --type=go 'httptest\.NewServer\(http\.HandlerFunc' --glob='*_test.go' | wc -lRepository: QuantumNous/new-api
Length of output: 207
Avoid t.Fatalf inside the HTTP handler goroutine.
At lines 18, 21, and 24, t.Fatalf is invoked from the server handler goroutine. Fatalf/FailNow are only safe from the test goroutine; calling them from other goroutines can lead to unreliable test failures. Capture request fields in the handler and assert them after fetchChannelUpstreamModelIDs returns.
Suggested fix
func TestFetchChannelUpstreamModelIDs_Qiniu(t *testing.T) {
t.Parallel()
+ type requestSnapshot struct {
+ method string
+ path string
+ auth string
+ }
+ reqCh := make(chan requestSnapshot, 1)
+
srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
- if r.Method != http.MethodGet {
- t.Fatalf("unexpected method: %s", r.Method)
- }
- if r.URL.Path != "/v1/models" {
- t.Fatalf("unexpected path: %s", r.URL.Path)
- }
- if got := r.Header.Get("Authorization"); got != "Bearer test-key" {
- t.Fatalf("unexpected Authorization header: %q", got)
- }
+ reqCh <- requestSnapshot{
+ method: r.Method,
+ path: r.URL.Path,
+ auth: r.Header.Get("Authorization"),
+ }
w.Header().Set("Content-Type", "application/json")
_, _ = w.Write([]byte(`{"object":"list","data":[{"id":"deepseek/deepseek-v3.1-terminus-thinking"},{"id":"gpt-4"}]}`))
}))
defer srv.Close()
@@
got, err := fetchChannelUpstreamModelIDs(ch)
if err != nil {
t.Fatalf("fetchChannelUpstreamModelIDs returned error: %v", err)
}
+ req := <-reqCh
+ if req.method != http.MethodGet {
+ t.Fatalf("unexpected method: %s", req.method)
+ }
+ if req.path != "/v1/models" {
+ t.Fatalf("unexpected path: %s", req.path)
+ }
+ if req.auth != "Bearer test-key" {
+ t.Fatalf("unexpected Authorization header: %q", req.auth)
+ }
if len(got) != 2 || got[0] != "deepseek/deepseek-v3.1-terminus-thinking" || got[1] != "gpt-4" {
t.Fatalf("unexpected models: %#v", got)
}
}🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In `@controller/channel_upstream_update_qiniu_test.go` around lines 17 - 25, The
handler in channel_upstream_update_qiniu_test.go is calling t.Fatalf from the
server goroutine (lines checking r.Method, r.URL.Path, and Authorization);
instead, capture the observed values (method, path, auth header, etc.) into
local variables or send them over a channel/sync struct from the handler, return
a normal HTTP response, then after invoking fetchChannelUpstreamModelIDs in the
test goroutine assert those captured values with t.Fatalf/require. Locate the
inline HTTP handler in the test and change its t.Fatalf calls to store values
(or send them on a channel) and perform the assertions in the test goroutine
after fetchChannelUpstreamModelIDs returns.
| // 定义支持流式选项的通道类型 | ||
| var streamSupportedChannels = map[int]bool{ | ||
| constant.ChannelTypeOpenAI: true, | ||
| constant.ChannelTypeQiniu: true, |
Contributor
There was a problem hiding this comment.
🧩 Analysis chain
🌐 Web query:
Qiniu (七牛) OpenAI-compatible chat/completions API: does streaming support stream_options(especiallyinclude_usage)?
💡 Result:
No, Qiniu's OpenAI-compatible chat/completions API does not support the stream_options parameter (including include_usage) for streaming responses. Official documentation lists stream as supported but omits stream_options. OpenAPI specs and parameter tables confirm standard OpenAI parameters like model, messages, max_tokens, temperature are included, but stream_options is absent. Examples show streaming with stream: true without usage in final chunk. Separate usage query APIs exist (/v2/stat/usage). Some providers reject stream_options, causing 422 errors (similar to Mistral issues).
Citations:
- 1: https://developer.qiniu.com/aitokenapi/13390/chat-completions
- 2: https://developer.qiniu.com/aitokenapi/12882/ai-inference-api?category=api
- 3: https://apidocs.qnaigc.com/233004161d0
- 4: https://apidocs.qnaigc.com/421308051e0
- 5: https://apidocs.qnaigc.com/397197786e0
Remove Qiniu from streamSupportedChannels—it does not support the stream_options parameter.
Qiniu's OpenAI-compatible API supports stream but explicitly omits stream_options (including include_usage). Including Qiniu at line 308 will cause upstream failures when StreamOptions are passed. Separate usage APIs exist instead.
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In `@relay/common/relay_info.go` at line 308, Remove constant.ChannelTypeQiniu
from the streamSupportedChannels map so Qiniu isn't treated as supporting
stream_options; locate the map named streamSupportedChannels and delete the
entry "constant.ChannelTypeQiniu: true" (or conditionally exclude
ChannelTypeQiniu wherever streamSupportedChannels is used) to prevent sending
StreamOptions/include_usage to Qiniu's API.
liangchaoboy
force-pushed
the
feat/qiniu-provider
branch
3 times, most recently
from
b8002ff to
89e5896
Compare
added 2 commits
April 15, 2026 19:43
Introduce Qiniu (OpenAI-compatible) channel type, enable upstream /v1/models fetching in UI, and add a unit test for model listing. Made-with: Cursor
Replace hardcoded OpenAI/Azure channel type check with the SupportStreamOptions flag so all channels registered in streamSupportedChannels (including Qiniu) correctly preserve StreamOptions in upstream requests.
liangchaoboy
force-pushed
the
feat/qiniu-provider
branch
from
89e5896 to
f4c3225
Compare
liangchaoboy
changed the title
Contributor
There was a problem hiding this comment.
🧹 Nitpick comments (1)
web/src/components/table/channels/modals/EditChannelModal.jsx (1)
133-136: Consolidate fetchable-type source of truth.Line 133 redefines a fetchable-type set that is already centralized in
MODEL_FETCHABLE_CHANNEL_TYPES. Keeping both lists invites future drift when new types are added. Prefer reusing the shared constant directly.♻️ Suggested simplification
-const MODEL_FETCHABLE_TYPES = new Set([ - 1, 4, 14, 34, 17, 26, 27, 24, 47, 25, 20, 23, 31, 40, 42, 48, 43, - 58, -]); +const MODEL_FETCHABLE_TYPES = MODEL_FETCHABLE_CHANNEL_TYPES;🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@web/src/components/table/channels/modals/EditChannelModal.jsx` around lines 133 - 136, The local constant MODEL_FETCHABLE_TYPES in EditChannelModal.jsx duplicates the centralized MODEL_FETCHABLE_CHANNEL_TYPES and should be removed; replace the local definition by importing and using MODEL_FETCHABLE_CHANNEL_TYPES (the shared source of truth) wherever MODEL_FETCHABLE_TYPES is referenced (e.g., in validation or conditional checks inside the EditChannelModal component), and ensure you add the appropriate import for MODEL_FETCHABLE_CHANNEL_TYPES from its module and update any variable name references to use the imported symbol.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Nitpick comments:
In `@web/src/components/table/channels/modals/EditChannelModal.jsx`:
- Around line 133-136: The local constant MODEL_FETCHABLE_TYPES in
EditChannelModal.jsx duplicates the centralized MODEL_FETCHABLE_CHANNEL_TYPES
and should be removed; replace the local definition by importing and using
MODEL_FETCHABLE_CHANNEL_TYPES (the shared source of truth) wherever
MODEL_FETCHABLE_TYPES is referenced (e.g., in validation or conditional checks
inside the EditChannelModal component), and ensure you add the appropriate
import for MODEL_FETCHABLE_CHANNEL_TYPES from its module and update any variable
name references to use the imported symbol.
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: 97a36106-6b29-4e81-8319-1f9324de2a6b
📒 Files selected for processing (8)
common/api_type.goconstant/channel.gocontroller/channel_upstream_update_qiniu_test.gorelay/channel/openai/adaptor.gorelay/common/relay_info.goweb/src/components/table/channels/modals/EditChannelModal.jsxweb/src/constants/channel.constants.jsweb/src/helpers/render.jsx
✅ Files skipped from review due to trivial changes (1)
- web/src/constants/channel.constants.js
🚧 Files skipped from review as they are similar to previous changes (6)
- common/api_type.go
- web/src/helpers/render.jsx
- controller/channel_upstream_update_qiniu_test.go
- relay/channel/openai/adaptor.go
- relay/common/relay_info.go
- constant/channel.go
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
1 participant
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.
变更描述 / Description
本 PR 为 Qiniu（七牛） 增加了一个独立的渠道类型（
type=58），并将其按 OpenAI 兼容供应商接入现有转发链路：后端将ChannelTypeQiniu映射到APITypeOpenAI，因此请求转发与鉴权逻辑复用 OpenAI adaptor（Authorization: Bearer <key>），无需额外的请求体/响应体转换。同时，补齐了“从上游拉取模型列表”的能力：前端将
type=58加入可拉取模型列表的白名单集合，使渠道编辑弹窗可以调用现有的/v1/models拉取流程；后端沿用统一的/v1/models解析逻辑，返回data[].id作为模型列表。新增单元测试使用httptest模拟上游/v1/models，验证请求路径与 Bearer 鉴权头正确，并能正确解析模型 ID。变更类型 / Type of change
🔗 关联任务 / Related Issue
✅ 提交前检查项 / Checklist
Bug fix，我已提交或关联对应 Issue，且不会将设计取舍、预期不一致或理解偏差直接归类为 bug。运行证明 / Proof of Work
本地测试通过（只跑本次新增覆盖项）：
go test ./controller -run Qiniu ok github.com/QuantumNous/new-api/controller 0.456sSummary by CodeRabbit
New Features
Tests