phsl / new-api

docs(web_api): 修复 Markdown 表格格式

- Add new Forbidden page at /forbidden (`web/src/pages/Forbidden/index.js`)
  - Use Semi-UI Empty with IllustrationNoAccess (250x250)
  - Update i18n description to: '您无权访问此页面，请联系管理员~'
  - Align visual style with existing 404 page
- Introduce `AdminRoute` in `web/src/helpers/auth.js`
  - Use `UserContext`/localStorage; redirect to `/forbidden` when `!user` or `user.role < 10`
- Protect console/admin routes with `AdminRoute` and register `/forbidden` in `web/src/App.js`
- Update `web/src/i18n/locales/en.json`
  - Add English translation for the new forbidden message
  - Remove legacy "没有权限" entry
- Lint passes; no runtime errors observed

- Switch Collapse from controlled (activeKey) to uncontrolled (defaultActiveKey) so user toggling works
- Add a stable key to reset Collapse state when tab/category changes
- Default all panels to collapsed via defaultActiveKey: []
- Preserve Panel itemKey for consistent behavior
- No linter errors introduced

Scope: web/src/components/table/channels/modals/ModelSelectModal.jsx

fix: zhipu_v4 thinking

- Truncate long labels via pure CSS and always show full text in a Tooltip
- Ensure the right-side Tag is never truncated and remains fully visible
- Simplify implementation: remove overflow detection and ResizeObserver
- Use minimal markup with sbg-button/sbg-inner/sbg-label to enable shrinking
- Add global rules to allow `.semi-button-content` to shrink and ellipsize

Files:
- web/src/components/common/ui/SelectableButtonGroup.jsx
- web/src/index.css

No API changes; visuals improved and code complexity reduced.

# Conflicts:
#	relay/channel/openai/adaptor.go

Frontend
- Models table (model management):
  - Render billing types with the same segmented list component (renderLimitedItems) used by tags and endpoints
  - Display quota_types as an array with capped items (maxDisplay: 3) and graceful fallback for unknown types
- Pricing view (unchanged by request):
  - Revert to single-value quota_type rendering and sorter
  - Keep ratio display logic based on quota_type only

Files
- web/src/components/table/models/ModelsColumnDefs.js
- web/src/components/table/model-pricing/view/table/PricingTableColumns.js

Notes
- This commit only adjusts the model management UI rendering; pricing views remain as-is

Backend
- Add GetBoundChannelsByModelsMap to batch-fetch bound channels via a single JOIN (Distinct), compatible with SQLite/MySQL/PostgreSQL
- Replace per-record enrichment with a single-pass enrichModels to avoid N+1 queries; compute unions for prefix/suffix/contains matches in memory
- Change Model.QuotaType to QuotaTypes []int and expose quota_types in responses
- Add GetModelQuotaTypes for cached O(1) lookups; exact models return a single-element array
- Sort quota_types for stable output order
- Remove unused code: GetModelByName, GetBoundChannels, GetBoundChannelsForModels, FindModelByNameWithRule, buildPrefixes, buildSuffixes
- Clean up redundant comments, keeping concise and readable code

Frontend
- Models table: switch to quota_types, render multiple billing modes ([0], [1], [0,1], future values supported)
- Pricing table: switch to quota_types; ratio display now checks quota_types.includes(0); array rendering for billing tags

Compatibility
- SQL uses standard JOIN/IN/Distinct; works across SQLite/MySQL/PostgreSQL
- Lint passes; no DB schema changes (quota_types is a JSON response field only)

Breaking Change
- API field renamed: quota_type -> quota_types (array). Update clients accordingly.

🔧 chore(db): drop legacy single-column UNIQUE indexes to prevent duplicate-key errors after soft-delete

Why
Previous versions created single-column UNIQUE constraints (`models.model_name`, `vendors.name`).
After introducing composite indexes on `(model_name, deleted_at)` and `(name, deleted_at)` for soft-delete support, those legacy constraints could still exist in user databases.
When a record was soft-deleted and re-inserted with the same name, MySQL raised `Error 1062 … for key 'models.model_name'`.

What
• In `migrateDB` and `migrateDBFast` paths of `model/main.go`, proactively drop:
  – `models.uk_model_name` and fallback `models.model_name`
  – `vendors.uk_vendor_name` and fallback `vendors.name`
• Keeps existing helper `dropIndexIfExists` to ensure the operation is MySQL-only and error-free when indexes are already absent.

Result
Startup migration now removes every possible legacy UNIQUE index, ensuring composite index strategy works correctly.
Users can soft-delete and recreate models/vendors with identical names without hitting duplicate-entry errors.

Summary
-------
1. **Backend**
   • `model/model_meta.go`
     – Add `GetBoundChannelsForModels([]string)` to retrieve channels for multiple models in a single SQL (`IN (?)`) and deduplicate with `GROUP BY`.

   • `controller/model_meta.go`
     – In non-exact `fillModelExtra`:
       – Remove per-model `GetBoundChannels` calls.
       – Collect matched model names, then call `GetBoundChannelsForModels` once and merge results into `channelSet`.
       – Minor cleanup on loop logic; channel aggregation now happens after quota/group/endpoint processing.

Impact
------
• Eliminates N+1 query pattern for prefix/suffix/contains rules.
• Reduces DB round-trips from *N + 1* to **1**, markedly speeding up the model-management list load.
• Keeps existing `GetBoundChannels` API intact for single-model scenarios; no breaking changes.

✨ feat: Enhance model listing and retrieval with support for Anthropic and Gemini models; refactor routes for better API key handling

Summary
-------
1. **Backend**
   • `model/model_meta.go`
     – Add `MatchedModels []string` and `MatchedCount int` (ignored by GORM) to expose matching details in API responses.
   • `controller/model_meta.go`
     – When processing prefix/suffix/contains rules in `fillModelExtra`, collect every matched model name, fill `MatchedModels`, and calculate `MatchedCount`.

2. **Frontend**
   • `web/src/components/table/models/ModelsColumnDefs.js`
     – Import `Tooltip`.
     – Enhance `renderNameRule` to:
       – Display tag text like “前缀 5个模型” for non-exact rules.
       – Show a tooltip listing all matched model names on hover.

Impact
------
Users now see the total number of concrete models aggregated under each prefix/suffix/contains rule and can inspect the exact list via tooltip, improving transparency in model management.

Summary
-------
1. **Backend**
   • `controller/model_meta.go`
     – For prefix/suffix/contains rules, aggregate endpoints, bound channels, enable groups, and quota types across all matched models.
     – When mixed billing types are detected, return `quota_type = -1` (unknown) instead of defaulting to volume-based.

2. **Frontend**
   • `web/src/helpers/utils.js`
     – `calculateModelPrice` now handles `quota_type = -1`, returning placeholder `'-'`.

   • `web/src/components/table/model-pricing/view/card/PricingCardView.jsx`
     – Billing tag logic updated: displays “按次计费” (times), “按量计费” (volume), or `'-'` for unknown.

   • `web/src/components/table/model-pricing/view/table/PricingTableColumns.js`
     – `renderQuotaType` shows “未知” for unknown billing type.

   • `web/src/components/table/models/ModelsColumnDefs.js`
     – Unified `renderQuotaType` to return `'-'` when type is unknown.

   • `web/src/components/table/model-pricing/modal/components/ModelPricingTable.jsx`
     – Group price table honors unknown billing type; pricing columns show `'-'` and neutral tag color.

3. **Utilities**
   • Added safe fallback colours/tags for unknown billing type across affected components.

Impact
------
• Ensures correct data aggregation for non-exact model matches.
• Prevents UI from implying volume billing when actual type is ambiguous.
• Provides consistent placeholder display (`'-'` or “未知”) across cards, tables and modals.

No breaking API changes; frontend gracefully handles legacy values.

Summary:
• Updated `helpers/utils.js` to display the “group” label without a colon, ensuring consistent typography with other price elements.

Details:
1. `formatPriceInfo`
   – Changed `{t('分组')}:` to `{t('分组')}` for a cleaner look.
   – Keeps spacing intact between label and selected group name.
2. No functional impact; purely visual polish.

Summary:
• Updated `PricingTopSection.jsx` to conditionally render `PricingVendorIntroWithSkeleton` only when `isMobile` is false.

Details:
1. Wrapped vendor-intro block in `!isMobile` check, preventing unnecessary content on small screens.
2. Kept desktop experience unchanged; no impact on other features.
3. Lint check passed with no new issues.

Result:
Cleaner mobile UI with improved performance and visual focus.

Overview:
• Introduced a new “Model Tag” filter across pricing screens
• Refactored `usePricingFilterCounts` to eliminate duplicated logic
• Improved tag handling to be case-insensitive and deduplicated
• Extended utilities to reset & persist the new filter

Details:
1. Added `filterTag` state to `useModelPricingData` and integrated it into all filtering paths.
2. Created reusable `PricingTags` component using `SelectableButtonGroup`.
3. Incorporated tag filter into `PricingSidebar` and mobile `PricingFilterModal`, including reset support.
4. Enhanced `resetPricingFilters` (helpers/utils) to restore tag filter defaults.
5. Refactored `usePricingFilterCounts.js`:
   • Centralized predicate `matchesFilters` to remove redundancy
   • Normalized tag parsing via `normalizeTags` helper
   • Memoized model subsets with concise filter calls
6. Updated lints – zero errors after refactor.

Result:
Users can now filter models by custom tags with consistent UX, and internal logic is cleaner, faster, and easier to extend.

Previously, the card view displayed a “-” whenever a model had no custom tags,
because `renderLimitedItems` returned a dash for an empty array.
Now the function is only invoked when `customTags.length > 0`, removing the
unwanted placeholder and keeping the UI clean.

File affected:
- web/src/components/table/model-pricing/view/card/PricingCardView.jsx

- Delay clearing selectedModel until SideSheet close animation completes
- Prevents brief display of 'AI/Unknown Model' text during closing transition
- Improves user experience by eliminating visual flicker
- Uses 300ms timeout matching Semi UI default animation duration

- Added unique keys for JSONEditor components to ensure proper re-rendering based on channelId.
- Implemented input reset to clear previous JSON field values when the modal is opened.

✨ feat: Enhance model listing and retrieval with support for Anthropic and Gemini models; refactor routes for better API key handling