When syncing official models, clicking "Apply overwrite" with zero selected
conflict fields resulted in no request being sent and the modal simply closing.
This blocked creation of missing models/vendors even though the backend
supports an empty `overwrite` array and will still create missing items.
Changes:
- Remove the early-return guard in `UpstreamConflictModal.handleOk`
- Always call `onSubmit(payload)` even when `payload` is empty
- Keep closing behavior when the request succeeds
Behavior:
- Users can now proceed with upstream sync without selecting any conflict fields
- Missing models/vendors are created as expected
- Existing models are not overwritten unless fields are explicitly selected
Affected:
- web/src/components/table/models/modals/UpstreamConflictModal.jsx
Quality:
- Lint passes
- No breaking changes
- No visual/UI changes beyond the intended behavior
Test plan:
1) Open official models sync and trigger a conflicts preview
2) Click "Apply overwrite" without selecting any fields
3) Expect the sync to proceed and a success toast indicating created models
4) Re-try with some fields selected to confirm overwrites still work
Ensure UpstreamConflictModal submits { overwrite: payload, locale } instead of spreading an array into an object
Remove numeric-key fallback from applyUpstreamOverwrite for simpler and explicit logic
Effect: selected fields are now actually updated; success message shows updated model count
Refs: backend SyncUpstreamModels expects overwrite: overwriteField[]
Frontend (web)
- ModelsActions.jsx
- Replace “Sync Official” with “Sync” and open a new two-step SyncWizard.
- Pass selected locale through to preview, sync, and overwrite flows.
- Keep conflict resolution flow; inject locale into overwrite submission.
- New: models/modals/SyncWizardModal.jsx
- Two-step wizard: (1) method selection (config-sync disabled for now), (2) language selection (en/zh/ja).
- Horizontal, centered Radio cards; returns { option, locale } via onConfirm.
- UpstreamConflictModal.jsx
- Add search input (model fuzzy search) and native pagination.
- Column header checkbox now only applies to rows in the current filtered result.
- Fix “Cannot access ‘filteredDataSource’ before initialization”.
- Refactor with useMemo/useCallback; extract helpers to remove duplicated logic:
- getPresentRowsForField, getHeaderState, applyHeaderChange
- Minor code cleanups and stability improvements.
- i18n (en.json)
- Add strings for the sync wizard and related actions (Sync, Sync Wizard, Select method/source/language, etc.).
- Adjust minor translations.
Hooks
- useModelsData.jsx
- Extend previewUpstreamDiff, syncUpstream, applyUpstreamOverwrite to accept options with locale.
- Send locale via query/body accordingly.
Backend (Go)
- controller/model_sync.go
- Accept locale from query/body and resolve i18n upstream URLs.
- Add SYNC_UPSTREAM_BASE for upstream base override (default: https://basellm.github.io/llm-metadata).
- Make HTTP timeouts/retries/limits configurable:
- SYNC_HTTP_TIMEOUT_SECONDS, SYNC_HTTP_RETRY, SYNC_HTTP_MAX_MB
- Add ETag-based caching and support both envelope and pure array JSON formats.
- Concurrently fetch vendors and models; improve error responses with locale and source URLs.
- Include source meta (locale, models_url, vendors_url) in success payloads.
Notes
- No breaking changes expected.
- Lint passes for touched files.
- Set sidebar skeleton background to use theme variable (--semi-color-bg-0) instead of hardcoded white for better dark mode compatibility
- Apply consistent background to both collapsed and expanded skeleton states
- Ensure sidebar container uses theme background when skeleton is loading
- Remove duplicate margin-right classes from skeleton wrapper components that conflicted with CSS definitions
- Simplify navigation text structure by removing unnecessary div wrappers to improve text truncation
- Add proper text layout styles for better truncation handling when menu items have long names
- Standardize icon-to-text spacing across all sidebar navigation items
The Linux/Vite build failed with:
“Could not resolve "./headerbar" from "src/components/layout/PageLayout.jsx"”
On Linux and with stricter ESM/rollup resolution, directory index files (index.jsx)
may not be auto-resolved reliably. Explicitly importing the index file ensures
cross-platform stability.
Changes:
- Update PageLayout import from "./headerbar" to "./headerbar/index.jsx"
Impact:
- Fixes build on Linux
- No runtime behavior changes
Verification:
- Linter passes for web/src/components/layout/PageLayout.jsx
Notes:
- Prefer explicit index file imports (and extensions) to avoid platform differences.
- Delete dropIndexIfExists helper from `model/main.go`
- Remove all calls to dropIndexIfExists in `migrateDB` and `migrateDBFast`
- Drop related comments and MySQL-only DROP INDEX code paths
- Keep GORM AutoMigrate as the sole migration path for `Model` and `Vendor`
Why:
- Simplifies migrations and avoids destructive index drops at startup
- Prevents noisy MySQL 1091 errors and vendor-specific branches
- Aligns with composite unique indexes (uk_model_name_delete_at, uk_vendor_name_delete_at)
Impact:
- No expected runtime behavior change; schema remains managed by GORM
- Legacy single-column unique indexes (if any) will no longer be auto-dropped
- Safe across MySQL/PostgreSQL/SQLite; MySQL Chinese charset checks remain intact
Verification:
- Lint passed for `model/main.go`
- Confirmed no remaining `DROP INDEX` or `dropIndexIfExists` references
Backend (controller/ratio_sync.go):
- Add built‑in official upstream to GetSyncableChannels (ID: -100, BaseURL: https://basellm.github.io)
- Support absolute endpoint URLs; otherwise join BaseURL + endpoint (defaults to /api/ratio_config)
- Harden HTTP client:
- IPv4‑first with IPv6 fallback for github.io
- Add ResponseHeaderTimeout
- 3 attempts with exponential backoff (200/400/800ms)
- Validate Content-Type and limit response body to 10MB (safe decode via io.LimitReader)
- Robust parsing: support type1 ratio_config map and type2 pricing list
- Use net.SplitHostPort for host parsing
- Use float tolerance in differences comparison to avoid false mismatches
- Remove unused code (tryDirect) and improve warnings
Frontend:
- UpstreamRatioSync.jsx: auto-assign official endpoint to /llm-metadata/api/newapi/ratio_config-v1-base.json
- ChannelSelectorModal.jsx:
- Pin the official source at the top of the list
- Show a green “官方” tag next to the status
- Refactor status renderer to accept the full record
Notes:
- Backward compatible; no API surface changes
- Official ratio_config reference: https://basellm.github.io/llm-metadata/api/newapi/ratio_config-v1-base.json
- Apply canonical Go formatting to all .go files
- No functional changes; whitespace/import/struct layout only
- Improves consistency, reduces diff noise, and aligns with standard tooling
- Model: rename `uk_model_name` -> `uk_model_name_delete_at`
(composite on `model_name` + `deleted_at`)
- Vendor: rename `uk_vendor_name` -> `uk_vendor_name_delete_at`
(composite on `name` + `deleted_at`)
- Keep legacy cleanup in `model/main.go` to drop old index names
(`uk_model_name`, `model_name`, `uk_vendor_name`, `name`) for compatibility.
Result: idempotent GORM migrations and no unnecessary index churn on MySQL restarts.
Files:
- `model/model_meta.go`
- `model/vendor_meta.go`
- Ran: bun run eslint:fix && bun run lint:fix
- Inserted AGPL license header via eslint-plugin-header
- Enforced no-multiple-empty-lines and other lint rules
- Formatted code using Prettier v3 (@so1ve/prettier-config)
- No functional changes; formatting-only baseline across JS/JSX files
- Add defaultVendorIcons mapping for major AI vendors
- Update getOrCreateVendor to automatically set vendor icons
- Add getDefaultVendorIcon helper function
- Support LobeHub icons Color variants (e.g., Claude.Color, Gemini.Color)
- Fix issue where default vendors were created without icons
This ensures that when new models are encountered, their vendors
will be created with appropriate colored icons for better UI display.
Affected vendors include:
- OpenAI, Anthropic, Google, Moonshot, 智谱, 阿里巴巴
- DeepSeek, MiniMax, 百度, 讯飞, 腾讯, Cohere
- Cloudflare, 360, 零一万物, Jina, Mistral, xAI
- Meta, 字节跳动, 快手, 即梦, Vidu, Microsoft/Azure
