based on (slightly) adopted regex from https://stackoverflow.com/a/20971688/1104502

It parses three groups:

• international code
• phone-number
• extension

Helpful if you want to strip national-code for feeding a normalized "base phone number" into a search index. Additional replace all non-digits [^\d] in a 2nd phase too, to get a digit-only field, which you can search with wildcards or n-grams.