Fix: Gap analysis weak-link deduplication checks wrong dictionary level, inflating extra count

[PRAteek-singHWY]

Issue

What is the issue?

In application/database/db.py, the gap_analysis() function has a dict-level lookup bug on line 2224 that breaks weak-link deduplication.

Line 2224 checks end_key in extra_paths_dict[key], but extra_paths_dict[key] is {"paths": {}} - so end_key (a node ID) can never match the only key "paths". The condition is always False. Line 2211 (strong-path branch) already does this correctly: end_key in extra_paths_dict[key]["paths"].

Expected Behaviour

• grouped_paths[key]["extra"] should only count unique weak-link destinations
• When multiple paths reach the same destination, the shortest-scoring path should be kept

Actual Behaviour

• extra count is inflated - every weak path increments the counter, even duplicates to the same destination
• The "Show average and weak links (N)" button on the frontend shows a wildly higher N than actual unique destinations
• Deduplication is broken - the last path processed wins instead of the shortest

Impact

This causes the "Show average and weak links (N)" button on the Map Analysis page to display wildly inflated numbers - far higher than the actual number of unique destinations. Every duplicate weak path is counted and stored separately instead of being deduplicated, meaning users see misleading gap analysis results and the shortest-scoring path is not preserved.

Fix: extra_paths_dict[key] → extra_paths_dict[key]["paths"] on line 2224, consistent with line 2211.

[PRAteek-singHWY]

Hi @Pa04rth

Raised a PR with a one-line fix for this. Even though the fix is small (extra_paths_dict[key] → extra_paths_dict[key]["paths"]), the impact is significant - it affects every gap analysis result by inflating weak-link counts and breaking shortest-path selection. The strong-path branch 13 lines above (line 2211) already does this correctly, so the fix just makes both branches consistent.

Thank You.

[Pa04rth]

Assigned!