Commit Graph

3642 Commits

Author SHA1 Message Date
t0ng7u
d96f846648 🐛 fix(model, web): robust JSON handling; remove datatypes dep; stabilize JSONEditor manual mode
- Why:
  - Eliminate `gorm.io/datatypes` for a single field and fix scan errors when drivers return JSON as string.
  - Prevent JSONEditor manual mode from locking on invalid JSON and from appending stray characters after “Fill Template”.

- What:
  - Backend (`model/prefill_group.go`):
    - Replaced `datatypes.JSON` with `JSONValue` (based on `json.RawMessage`) for `PrefillGroup.Items`.
    - Implemented `sql.Scanner` and `driver.Valuer` to accept both `[]byte` and `string`.
    - Implemented `MarshalJSON`/`UnmarshalJSON` to preserve raw JSON in API without base64.
    - Converted comments to Chinese.
  - Frontend (`web/src/components/common/ui/JSONEditor.js`):
    - Added `manualText` buffer for manual mode to avoid input being overridden by external value.
    - Only propagate `onChange` when manual text is valid JSON; otherwise show error but do not block typing.
    - Safe manual-mode rendering: derive rows from `manualText` and avoid calling `split` on non-strings.
    - Improved mode toggle: populate `manualText` from visual data; validate before switching back to visual.
    - Fixed “Fill Template” to sync `manualText`, `jsonData`, and `onChange` to avoid stray trailing characters.

- Impact:
  - Resolves: “unsupported Scan, storing driver.Value type string into type *json.RawMessage”.
  - Resolves: `value.split is not a function` in manual mode.
  - Resolves: extra `s` appended after inserting template.
  - API shape and DB column type remain the same (`gorm:"type:json"`); no `go.mod` changes.
  - Lints pass for modified files.

Files changed:
- model/prefill_group.go
- web/src/components/common/ui/JSONEditor.js
2025-08-08 04:21:50 +08:00
t0ng7u
473f3b6f3e ♻️ refactor(model): replace gorm.io/datatypes with JSONValue for PrefillGroup.Items; fix JSON scan across drivers
- Why:
  - Avoid introducing `gorm.io/datatypes` for a single field.
  - Align with existing pattern (`ChannelInfo`, `Properties`) using `Scanner`/`Valuer`.
  - Fix runtime error when drivers return JSON as string.

- What:
  - Introduced `JSONValue` (based on `json.RawMessage`) implementing `sql.Scanner` and `driver.Valuer`, with `MarshalJSON`/`UnmarshalJSON` to preserve raw JSON in API.
  - Updated `PrefillGroup.Items` to use `JSONValue` with `gorm:"type:json"`.
  - Localized comments in `model/prefill_group.go` to Chinese.

- Impact:
  - Resolves “unsupported Scan, storing driver.Value type string into type *json.RawMessage”.
  - Works with MySQL/Postgres/SQLite whether JSON is returned as `[]byte` or `string`.
  - API and DB schema remain unchanged; no `go.mod` changes; lints pass.

Files changed:
- model/prefill_group.go
2025-08-08 04:09:53 +08:00
t0ng7u
7f1a471751 ♻️ refactor(model): replace gorm.io/datatypes with json.RawMessage for PrefillGroup.Items
- Why: Avoid adding `gorm.io/datatypes` for a single field; the rest of the codebase does not use it, and using the standard library keeps dependencies lean.
- What:
  - Switched `PrefillGroup.Items` from `datatypes.JSON` to `json.RawMessage`.
  - Updated imports in `model/prefill_group.go` to use `encoding/json` and removed the unused `gorm.io/datatypes`.
  - Preserved `gorm:"type:json"` so DB column behavior remains the same.
- Impact:
  - API response/request shape for `items` remains unchanged (still JSON).
  - DB schema behavior is unchanged; GORM migration continues to handle the field as JSON.
  - No other references to `datatypes` exist; no `go.mod` changes needed.
  - Lints pass for the modified file.

Files changed:
- model/prefill_group.go

No breaking changes.
2025-08-08 03:52:07 +08:00
t0ng7u
bbac342f3a 🤔 revert: revert go.sum & go.mod version 2025-08-08 03:35:31 +08:00
t0ng7u
4b3702987f 🤔 revert: revert go.sum & go.mod version 2025-08-08 03:22:25 +08:00
t0ng7u
6341847203 🎨 refactor(ui): merge “Content Configuration” into “Basic Information” card in EditPrefillGroupModal
- Move `items` field (`JSONEditor` for endpoint type, `Form.TagInput` otherwise) into the first “Basic Information” card
- Remove the second “Content Configuration” card and its header; consolidate to a single-card layout
- Preserve form initialization, validation, and submit logic; API payload structure remains unchanged
- Improves clarity and reduces visual clutter without altering behavior
- Lint passes

Affected file:
- `web/src/components/table/models/modals/EditPrefillGroupModal.jsx`

No breaking changes.
2025-08-08 03:04:51 +08:00
t0ng7u
4e75a9b3b3 feat: Improve models UX and robustness: add JSONEditor extraFooter, fix endpoints rendering, and clean up deps
- Why
  - Needed to separate help text from action buttons in JSONEditor for better layout and UX.
  - Models table should robustly render both new object-based endpoint mappings and legacy arrays.
  - Columns should re-render when vendor map changes.
  - Minor import cleanups for consistency.

- What
  - JSONEditor.js
    - Added optional prop extraFooter to render content below the extraText divider.
    - Kept extraText rendered via Divider; extraFooter appears on the next line for clear separation.
  - EditModelModal.jsx
    - Moved endpoint group buttons from extraText into extraFooter to display under the helper text.
    - Kept merge-logic: group items are merged into current endpoints JSON with key override semantics.
    - Consolidated lucide-react imports into a single line.
  - ModelsColumnDefs.js
    - Made endpoint renderer resilient:
      - Supports object-based JSON (keys as endpoint types) and legacy array format.
      - Displays keys/items as tags and limits the number shown; uses stringToColor for visual consistency.
    - Consolidated Semi UI imports into a single line.
  - ModelsTable.jsx
    - Fixed columns memoization dependency to include vendorMap, ensuring re-render when vendor data changes.

- Notes
  - Backward-compatible: extraFooter is additive; existing JSONEditor usage remains unchanged.
  - No API changes to backend.
  - No linter errors introduced.

- Files touched
  - web/src/components/common/ui/JSONEditor.js
  - web/src/components/table/models/modals/EditModelModal.jsx
  - web/src/components/table/models/ModelsColumnDefs.js
  - web/src/components/table/models/ModelsTable.jsx

- Impact
  - Clearer UI for endpoint editing (buttons now below helper text).
  - Correct endpoints display for object-based mappings in models list.
  - More reliable reactivity when vendor data updates.
2025-08-08 02:59:45 +08:00
t0ng7u
26f44b8d4b Merge remote-tracking branch 'origin/alpha' into refactor/model-pricing 2025-08-08 02:34:44 +08:00
t0ng7u
8fba0017c7 feat(pricing+endpoints+ui): wire custom endpoint mapping end‑to‑end and overhaul visual JSON editor
Backend (Go)
- Include custom endpoints in each model’s SupportedEndpointTypes by parsing Model.Endpoints (JSON) and appending keys alongside native endpoint types.
- Build a global supportedEndpointMap map[string]EndpointInfo{path, method} by:
  - Seeding with native defaults.
  - Overriding/adding from models.endpoints (accepts string path → default POST, or {path, method}).
- Expose supported_endpoint at the top level of /api/pricing (vendors-like), removing per-model duplication.
- Fix default path for EndpointTypeOpenAIResponse to /v1/responses.
- Keep concurrency/caching for pricing retrieval intact.

Frontend (React)
- Fetch supported_endpoint in useModelPricingData and propagate to PricingPage → ModelDetailSideSheet → ModelEndpoints.
- ModelEndpoints
  - Resolve path+method via endpointMap; replace {model} with actual model name.
  - Fix mobile visibility; always show path and HTTP method.
- JSONEditor
  - Wrap with Form.Slot to inherit form layout; simplify visual styles.
  - Use Tabs for “Visual” / “Manual” modes.
  - Unify editors: key-value editor now supports nested JSON:
    - “+” to convert a primitive into an object and add nested fields.
    - Add “Convert to value” for two‑way toggle back from object.
    - Stable key rename without reordering rows; new rows append at bottom.
    - Use Row/Col grid for clean alignment; region editor uses Form.Slot + grid.
- Editing flows
  - EditModelModal / EditPrefillGroupModal use JSONEditor (editorType='object') for endpoint mappings.
  - PrefillGroupManagement renders endpoint group items by JSON keys.

Data expectations / compatibility
- models.endpoints should be a JSON object mapping endpoint type → string path or {path, method}. Strings default to POST.
- No schema changes; existing TEXT field continues to store JSON.

QA
- /api/pricing now returns custom endpoint types and global supported_endpoint.
- UI shows both native and custom endpoints; paths/methods render on mobile; nested editing works and preserves order.
2025-08-08 02:34:15 +08:00
Calcium-Ion
677a02c632 Merge pull request #1517 from RedwindA/fix/gemini-fetch-models
Fix/gemini-fetch-models
2025-08-07 20:44:34 +08:00
Calcium-Ion
177b891905 Merge pull request #1518 from RedwindA/fix/disable-gemini-sse
Fix/disable-gemini-ping
2025-08-07 20:44:24 +08:00
CaIon
c4dcc6df9c feat: enhance Adaptor to support multiple relay modes in request handling 2025-08-07 19:30:42 +08:00
CaIon
7ddd314015 feat: implement ConvertClaudeRequest method in baidu_v2 Adaptor 2025-08-07 19:19:59 +08:00
Calcium-Ion
ba7325c884 Merge pull request #1522 from QuantumNous/support-deepseek-claude
feat: support deepseek claude format (convert)
2025-08-07 19:04:05 +08:00
Calcium-Ion
3c4b1ef127 Merge pull request #1521 from QuantumNous/support-qwen-claude
feat: support qwen claude format
2025-08-07 19:03:40 +08:00
CaIon
18c630e5e4 feat: support deepseek claude format (convert) 2025-08-07 19:01:49 +08:00
CaIon
0ea0a432bf feat: support qwen claude format 2025-08-07 18:32:31 +08:00
IcedTangerine
8a964efbed Merge pull request #1519 from feitianbubu/pr/fix-qwen3-thinking-test
feat: enable thinking mode on ali thinking model
2025-08-07 17:39:27 +08:00
CaIon
865bb7aad8 Revert "feat: update Usage struct to support dynamic token handling with ceil function #1503"
This reverts commit 71c39c9893.
2025-08-07 16:22:40 +08:00
CaIon
d9c1fb5244 feat: update MaxTokens handling 2025-08-07 16:15:59 +08:00
CaIon
71c39c9893 feat: update Usage struct to support dynamic token handling with ceil function #1503 2025-08-07 15:40:12 +08:00
feitianbubu
38067f1ddc feat: enable thinking mode on ali thinking model 2025-08-07 11:59:54 +08:00
t0ng7u
7cfeb6e87c Merge remote-tracking branch 'origin/alpha' into refactor/model-pricing 2025-08-07 11:09:28 +08:00
t0ng7u
0a231a8acc 🎨 feat(models): add row styling for disabled models in ModelsTable
Add visual distinction for enabled/disabled models by applying different
background colors to table rows based on model status. This implementation
follows the same pattern used in ChannelsTable for consistent user experience.

Changes:
- Modified handleRow function in useModelsData.js to include row styling
- Disabled models (status !== 1) now display with gray background using
  --semi-color-disabled-border CSS variable
- Enabled models (status === 1) maintain normal background color
- Preserved existing row click selection functionality

This enhancement improves the visual feedback for users to quickly identify
which models are active vs inactive in the models management interface.
2025-08-07 10:54:05 +08:00
RedwindA
1cea7a0314 fix: 调整Disable Ping标志的设置位置 2025-08-07 06:18:22 +08:00
RedwindA
ed95a9f2b2 fix a typo in comment 2025-08-07 01:11:01 +08:00
RedwindA
76d71a032a fix: 修复 FetchUpstreamModels 函数中 AuthHeader 的使用,确保正确处理 多key聚合的情况 2025-08-07 01:01:45 +08:00
RedwindA
38bff1a0e0 refactor: 移除 GoogleOpenAI 兼容模型相关结构体,简化 FetchUpstreamModels 函数逻辑 2025-08-07 00:54:48 +08:00
Xyfacai
0c0caad827 refactor: 调整模型匹配 2025-08-06 20:09:22 +08:00
Xyfacai
4445e5891f fix: error code 显示问题 2025-08-06 19:40:26 +08:00
CaIon
f46cefbd39 fix: update budget calculation logic in relay-gemini to use clamping function 2025-08-06 16:25:48 +08:00
CaIon
feef022303 feat: enhance ThinkingAdaptor with effort-based budget clamping and extra body handling 2025-08-06 16:20:38 +08:00
CaIon
6a80c18189 feat: add reasoning support for Openrouter requests with "-thinking" suffix 2025-08-06 12:50:26 +08:00
Calcium-Ion
6616bb4048 Merge pull request #1508 from wzxjohn/feature/aws_new_apikey_support
feat: support aws bedrock apikey
2025-08-06 12:04:28 +08:00
Calcium-Ion
ac5f51c3d5 Merge pull request #1510 from RedwindA/fix/manual-price-edit-modelName-check
fix:修复添加模型倍率时的输入框锁定
2025-08-06 12:03:44 +08:00
Calcium-Ion
587888a688 Merge pull request #1511 from neotf/feat-05
feat: add support for claude-opus-4-1 model and update ratios
2025-08-06 12:03:33 +08:00
Calcium-Ion
7370b4fbcd Merge pull request #1509 from QuantumNous/responses-input-cache-token
fix: responses cache token 未计费
2025-08-06 11:22:14 +08:00
t0ng7u
94506bee99 feat(models): Revamp EditModelModal UI and UX
This commit significantly refactors the `EditModelModal` component to streamline the user interface and enhance usability, aligning it with the interaction patterns found elsewhere in the application.

- **Consolidated Layout:** Merged the "Vendor Information" and "Feature Configuration" sections into a single "Basic Information" card. This simplifies the form, reduces clutter, and makes all settings accessible in one view.

- **Improved Prefill Groups:** Replaced the separate `Select` dropdowns for tag and endpoint groups with a more intuitive button-based system within the `extraText` of the `TagInput` components.

- **Additive Button Logic:** The prefill group buttons now operate in an additive mode. Users can click multiple group buttons to incrementally add tags or endpoints, with duplicates being automatically handled.

- **Clear Functionality:** Added "Clear" buttons for both tags and endpoints, allowing users to easily reset the fields.

- **Code Cleanup:** Removed the unused `endpointOptions` constant and unnecessary icon imports (`Building`, `Settings`) to keep the codebase clean.
2025-08-06 03:29:45 +08:00
t0ng7u
7c814a5fd9 🚀 refactor: migrate vendor-count aggregation to model layer & align frontend logic
Summary
• Backend
  – Moved duplicate-name validation and total vendor-count aggregation from controllers (`controller/model_meta.go`, `controller/vendor_meta.go`, `controller/prefill_group.go`) to model layer (`model/model_meta.go`, `model/vendor_meta.go`, `model/prefill_group.go`).
  – Added `GetVendorModelCounts()` and `Is*NameDuplicated()` helpers; controllers now call these instead of duplicating queries.
  – API response for `/api/models` now returns `vendor_counts` with per-vendor totals across all pages, plus `all` summary.
  – Removed redundant checks and unused imports, eliminating `go vet` warnings.

• Frontend
  – `useModelsData.js` updated to consume backend-supplied `vendor_counts`, calculate the `all` total once, and drop legacy client-side counting logic.
  – Simplified initial data flow: first render now triggers only one models request.
  – Deleted obsolete `updateVendorCounts` helper and related comments.
  – Ensured search flow also sets `vendorCounts`, keeping tab badges accurate.

Why
This refactor enforces single-responsibility (aggregation in model layer), delivers consistent totals irrespective of pagination, and removes redundant client queries, leading to cleaner code and better performance.
2025-08-06 01:40:08 +08:00
neotf
24aa29598a feat: add support for claude-opus-4-1 model and update ratios 2025-08-06 00:58:46 +08:00
t0ng7u
d61a862fa2 Merge remote-tracking branch 'origin/alpha' into refactor/model-pricing 2025-08-05 23:19:24 +08:00
RedwindA
e29c6b44c7 fix(web): 修复模型倍率设置中添加新模型时输入框锁定的问题 2025-08-05 23:18:42 +08:00
t0ng7u
327a0ca323 🚀 refactor: refine pricing refresh logic & hide disabled models
Summary
-------
1. Pricing generation
   • `model/pricing.go`: skip any model whose `status != 1` when building
     `pricingMap`, ensuring disabled models are never returned to the
     front-end.

2. Cache refresh placement
   • `controller/model_meta.go`
     – Removed `model.RefreshPricing()` from pure read handlers
       (`GetAllModelsMeta`, `SearchModelsMeta`).
     – Kept refresh only in mutating handlers
       (`Create`, `Update`, `Delete`), guaranteeing data is updated
       immediately after an admin change while avoiding redundant work
       on every read.

Result
------
Front-end no longer receives information about disabled models, and
pricing cache refreshes occur exactly when model data is modified,
improving efficiency and consistency.
2025-08-05 23:18:12 +08:00
creamlike1024
a746309a8e fix: responses 流 cache token 未计费 2025-08-05 23:08:08 +08:00
wzxjohn
d247f90571 feat: support aws bedrock apikey 2025-08-05 23:01:30 +08:00
creamlike1024
edbe18b157 fix: responses cache token 未计费 2025-08-05 22:56:27 +08:00
t0ng7u
d951485431 feat: enhance soft-delete handling & boost pricing cache performance
Summary
-------
This commit unifies soft-delete behaviour across meta tables and
introduces an in-memory cache for model pricing look-ups to improve
throughput under high concurrency.

Details
-------
Soft-delete consistency
• PrefillGroup / Vendor / Model
  – Added `gorm.DeletedAt` field with `json:"-" gorm:"index"`.
  – Replaced plain `uniqueIndex` with partial unique indexes
    `uniqueIndex:<name>,where:deleted_at IS NULL`
    allowing duplicate keys after logical deletion while preserving
    uniqueness for active rows.
• Imports updated to include `gorm.io/gorm`.
• JSON output now hides `deleted_at`, matching existing tables.

High-throughput pricing cache
• model/pricing.go
  – Added thread-safe maps `modelEnableGroups` & `modelQuotaTypeMap`
    plus RW-mutex for O(1) access.
  – `updatePricing()` now refreshes these maps alongside `pricingMap`.
• model/model_extra.go
  – Rewrote `GetModelEnableGroups` & `GetModelQuotaType` to read from
    the new maps, falling back to automatic refresh via `GetPricing()`.

Misc
• Retained `RefreshPricing()` helper for immediate cache invalidation
  after admin actions.
• All modified files pass linter; no breaking DB migrations required
  (handled by AutoMigrate).

Result
------
– Soft-delete logic is transparent, safe, and allows record “revival”.
– Pricing-related queries are now constant-time, reducing CPU usage and
  latency under load.
2025-08-05 22:26:19 +08:00
Calcium-Ion
306a1a3f57 Merge pull request #1507 from QuantumNous/multi-key-manage
feat: implement channel-specific locking for thread-safe polling
2025-08-05 20:40:26 +08:00
CaIon
2431de78fa fix: reorder request URL handling for relay formats in Adaptor 2025-08-05 20:40:00 +08:00
t0ng7u
f3a1f98add 🐛 fix(models): eliminate vendor column flicker by loading vendors before models
Why:
• The vendor list API is separate from the models API, causing the “Vendor” column in `ModelsTable` to flash (rendering `'-'` first, then updating) after the table finishes loading.
• This visual jump degrades the user experience.

What:
• Updated `web/src/hooks/models/useModelsData.js`
  – In the initial `useEffect`, vendors are fetched first with `loadVendors()` and awaited.
  – Only after vendors are ready do we call `loadModels()`, ensuring `vendorMap` is populated before the table renders.

Outcome:
• The table now renders with complete vendor data on first paint, removing the flicker and providing a smoother UI.
2025-08-04 22:11:13 +08:00