[Bug] CSV import crashes on duplicate header names

[thisisprabha]

Summary
The CSV import path can crash before it reaches error handling when a file contains duplicate header names.

Evidence

• ios/NetWorthIOS/Services/CSVService.swift:50-55 normalizes header names and immediately builds columnIndex with Dictionary(uniqueKeysWithValues:).
• Dictionary(uniqueKeysWithValues:) traps on duplicate keys instead of returning a recoverable error.
• This means malformed files with repeated columns such as values or updatedAt bypass the intended ImportError.invalidHeader path and can terminate the import flow.
• Source basis: dirty workspace change in ios/NetWorthIOS/Services/CSVService.swift.

Why This Matters
A malformed backup should fail gracefully. Crashing the import flow turns a user-data recovery action into an app-stability issue and makes it harder to explain what was wrong with the file.

Suggested Direction
Validate the normalized header row for duplicates before constructing the dictionary, and surface a controlled import error when duplicates are present.

Acceptance Criteria

• Duplicate header names are detected without crashing.
• The importer returns a controlled error for duplicate-column CSVs.
• The UI shows a normal import failure notice instead of terminating the flow.
• Add a regression test covering duplicate headers after BOM/whitespace normalization.

Deduping Notes
Checked open issues #1 and #2; neither covers CSV parsing or import crash handling.