My customer phone list has the same number in Arabic digits and in Western digits; why aren't the duplicates removed?

Where the mix comes from

A customer who types their number into your order form on a phone with an Arabic keyboard may send Arabic-Indic digits; another types Western ones. When you combine store orders with WhatsApp chats and a list from an exhibition, you end up with both styles. Your eye reads ٠٥٥١٢٣٤٥٦٧ and 0551234567 as one number; the computer sees ten completely different characters.

A worked example

Note that none of the matching options solves this. Case sensitivity has nothing to do with digits, ignoring spaces does not turn one character into another, and ignoring diacritics is about Arabic vowel marks only.

Steps

1. Paste the list into the Arabic/Persian/Western digit converter, set 'Target digit style' to 'Western 0123456789' (the default target is Arabic-Indic), and copy the output.
2. Paste that output into the Duplicate-line remover.
3. Check the number of removed lines, and open 'Lines that repeated' to see which numbers were repeated and how often.
4. Copy the clean list or download it as unique-lines.txt.

Other spellings of the same number that are not caught

Unifying digits fixes one kind of hidden repeat. Others survive it, because the tool does not know a line is a phone number:

• The country code: we tried 0551234567, 966551234567 and +966551234567 together, and all three stayed with nothing removed. Pick one format and bring every number to it before removing duplicates.
• Spaces inside the number: 055 123 4567 is a different line from 0551234567. The ignore-spaces option covers the start and end of a line only, not the middle.
• Dashes or brackets: 055-123-4567 is also a different line.

Which copy stays?

After conversion the two copies are identical, so it does not matter which one stays. But watch the order of steps: if you remove duplicates first and convert afterwards, the list comes back with the number twice. Convert first, then remove.