Resolves the same "Failed to fetch" cancellation issue that was occurring
in the "modify single event in a series" flow by eliminating concurrent
HTTP requests from the frontend.
## Problem:
The RecurringEditAction::ThisEvent handler was making two concurrent requests:
1. UPDATE request via update_callback.emit()
2. CREATE request via create_callback.emit()
This caused the same race condition and HTTP cancellation (~700-900ms) that
we previously fixed in the "this_and_future" flow.
## Solution:
- **Remove concurrent CREATE request** from frontend
- **Use single UPDATE request** with "this_only" scope
- **Backend handles both operations** atomically:
1. Add EXDATE to original series (exclude occurrence)
2. Create exception event with RECURRENCE-ID (user modifications)
## Implementation:
- Frontend sends single request with occurrence_date and dragged times
- Backend update_single_occurrence() already handled both operations
- Added comprehensive RFC 5545 documentation for single occurrence modification
- Cleaned up unused imports and variables
## Benefits:
- No more HTTP request cancellation for single event modifications
- Proper RFC 5545 EXDATE + RECURRENCE-ID exception handling
- Atomic operations ensure data consistency
- Matches the pattern used in this_and_future fix
The "modify single event" drag operations now work reliably without
network errors, completing the fix for all recurring event modification flows.
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
Connor Johnstone
b9e8778f8f