mirror of
https://github.com/purrgrammer/grimoire.git
synced 2026-04-10 15:36:53 +02:00
* feat: add preview routes for Nostr identifiers (npub, nevent, note, naddr)
This commit adds dedicated preview routes for Nostr identifiers at the root level:
- /npub... - Shows a single profile view for npub identifiers
- /nevent... - Shows a single event detail view for nevent identifiers
- /note... - Shows a single event detail view for note identifiers
- /naddr... - Redirects spellbooks (kind 30777) to /:actor/:identifier route
Key changes:
- Created PreviewProfilePage component for npub identifiers
- Created PreviewEventPage component for nevent/note identifiers
- Created PreviewAddressPage component for naddr redirects
- Added hideBottomBar prop to AppShell to hide tabs in preview mode
- Added routes to root.tsx for all identifier types
Preview pages don't show bottom tabs and don't affect user's workspace layout.
* chore: update package-lock.json
* refactor: create reusable useNip19Decode hook and improve preview pages
This commit makes the preview pages production-ready by:
1. Created useNip19Decode hook (src/hooks/useNip19Decode.ts):
- Reusable hook for decoding NIP-19 identifiers (npub, note, nevent, naddr, nprofile)
- Type-safe with discriminated union for decoded entities
- Comprehensive error handling with retry functionality
- Loading states and error messages
- Well-documented with JSDoc comments and usage examples
2. Comprehensive test coverage (src/hooks/useNip19Decode.test.ts):
- 11 tests covering all entity types (npub, note, nevent, naddr)
- Tests for error handling (missing identifier, invalid format, corrupted bech32)
- Tests for retry functionality and state changes
- Uses jsdom environment for React hook testing
- All tests passing ✓
3. Refactored preview pages to use the hook:
- PreviewProfilePage: Simplified from 80 to 81 lines with cleaner logic
- PreviewEventPage: Improved type safety and error handling
- PreviewAddressPage: Better separation of concerns
- All pages now have consistent error handling and retry functionality
- Better user experience with improved error messages
4. Dependencies added:
- @testing-library/react for React hook testing
- @testing-library/dom for DOM testing utilities
- jsdom and happy-dom for browser environment simulation in tests
Benefits:
- Code deduplication: Preview pages share decoding logic
- Type safety: Discriminated union prevents type errors
- Testability: Hook can be tested independently
- Maintainability: Single source of truth for NIP-19 decoding
- User experience: Consistent error handling and retry across all preview pages
- Production-ready: Comprehensive tests and error handling
* refactor: simplify useNip19Decode to synchronous with memoization
NIP-19 decoding is synchronous - removed unnecessary async complexity:
Hook changes (src/hooks/useNip19Decode.ts):
- Removed loading states (isLoading, setIsLoading)
- Removed retry functionality (unnecessary for sync operations)
- Now uses useMemo for efficient memoization
- Returns { decoded, error } instead of { decoded, isLoading, error, retry }
- Same string always yields same result (memoized)
- Went from ~120 lines to ~115 lines, but much simpler
Preview page changes:
- Removed loading spinners and states
- Removed retry buttons
- Simplified error handling
- Cleaner, more readable code
- PreviewProfilePage: 55 lines (down from 81)
- PreviewEventPage: 83 lines (down from 105)
- PreviewAddressPage: 83 lines (down from 117)
Test changes (src/hooks/useNip19Decode.test.ts):
- Removed waitFor and async/await (not needed for sync)
- Tests run faster (39ms vs 77ms - 49% improvement)
- Added memoization tests to verify caching works
- Simplified from 11 async tests to 11 sync tests
- All 11 tests passing ✓
Benefits:
- Simpler mental model: decode happens instantly
- Better performance: no state updates, just memoization
- Easier to test: synchronous tests are simpler
- More correct: matches the actual synchronous nature of nip19.decode()
- Less code: removed ~150 lines of unnecessary complexity
* feat: show detail view for all addressable events in naddr preview
Previously, PreviewAddressPage only handled spellbooks (kind 30777) and
showed errors for other addressable events. Now:
- Spellbooks (kind 30777): Redirect to /:actor/:identifier (existing behavior)
- All other addressable events: Show in EventDetailViewer
This enables previewing any addressable event (long-form articles, live
events, community posts, etc.) via naddr links.
Changes:
- Import EventDetailViewer
- Removed error state for non-spellbook kinds
- Show EventDetailViewer with AddressPointer for all other kinds
- Simplified from 83 lines to 77 lines
* fix: correct route patterns for NIP-19 identifier previews
The previous route patterns (/npub:identifier) conflicted with the catch-all
/:actor/:identifier route and didn't properly match NIP-19 identifiers.
Fixed by:
1. Using wildcard routes with correct prefixes:
- /npub1* (not /npub:identifier)
- /nevent1* (not /nevent:identifier)
- /note1* (not /note:identifier)
- /naddr1* (not /naddr:identifier)
2. Updated preview components to use params['*'] for wildcard capture:
- Reconstruct full identifier as prefix + captured part
- e.g., npub1 + params['*'] = npub107jk7htfv...
This ensures routes properly match before the catch-all /:actor/:identifier
route and correctly capture the full bech32-encoded identifier.
Test URL: /npub107jk7htfv243u0x5ynn43scq9wrxtaasmrwwa8lfu2ydwag6cx2quqncxg
* style: apply prettier formatting
* fix: use loader-based routing for NIP-19 identifiers in React Router v7
Previous attempts using wildcard routes didn't work properly in React Router v7.
Solution:
- Single /:identifier route with a loader that validates NIP-19 prefixes
- Loader throws 404 if identifier doesn't start with npub1/note1/nevent1/naddr1
- Created Nip19PreviewRouter component that routes to correct preview page
- Routes are properly ordered: /:identifier before /:actor/:identifier catch-all
This ensures /npub107jk... routes to profile preview, not spellbook route.
Benefits:
- Simpler routing configuration (1 route vs 4 duplicate routes)
- Proper validation via loader
- Clean separation of concerns with router component
- Works correctly in React Router v7
---------
Co-authored-by: Claude <noreply@anthropic.com>
63 lines
1.5 KiB
TypeScript
63 lines
1.5 KiB
TypeScript
import { createBrowserRouter, RouterProvider } from "react-router";
|
|
import { AppShell } from "./components/layouts/AppShell";
|
|
import DashboardPage from "./components/pages/DashboardPage";
|
|
import SpellbookPage from "./components/pages/SpellbookPage";
|
|
import Nip19PreviewRouter from "./components/pages/Nip19PreviewRouter";
|
|
|
|
const router = createBrowserRouter([
|
|
{
|
|
path: "/",
|
|
element: (
|
|
<AppShell>
|
|
<DashboardPage />
|
|
</AppShell>
|
|
),
|
|
},
|
|
{
|
|
path: "/preview/:actor/:identifier",
|
|
element: (
|
|
<AppShell>
|
|
<SpellbookPage />
|
|
</AppShell>
|
|
),
|
|
},
|
|
// NIP-19 identifier preview route - must come before /:actor/:identifier catch-all
|
|
{
|
|
path: "/:identifier",
|
|
element: (
|
|
<AppShell hideBottomBar>
|
|
<Nip19PreviewRouter />
|
|
</AppShell>
|
|
),
|
|
// Only match single-segment paths that look like NIP-19 identifiers
|
|
loader: ({ params }) => {
|
|
const id = params.identifier;
|
|
if (
|
|
!id ||
|
|
!(
|
|
id.startsWith("npub1") ||
|
|
id.startsWith("note1") ||
|
|
id.startsWith("nevent1") ||
|
|
id.startsWith("naddr1")
|
|
)
|
|
) {
|
|
throw new Response("Not Found", { status: 404 });
|
|
}
|
|
return null;
|
|
},
|
|
},
|
|
// Catch-all for two-segment paths (spellbooks, etc.)
|
|
{
|
|
path: "/:actor/:identifier",
|
|
element: (
|
|
<AppShell>
|
|
<SpellbookPage />
|
|
</AppShell>
|
|
),
|
|
},
|
|
]);
|
|
|
|
export default function Root() {
|
|
return <RouterProvider router={router} />;
|
|
}
|