Question

For example, I want to validate a string containing a phone number to enforce first a ‘+’ followed by 2 digits, then a sequence of either digits or a single instance of either a hyphen or a space. The string must end with a digit.

I currently have:

<xs:pattern value="\+\d{2}[- ]?[\d -]+[\d]"/>

but this does not restrict repetition of the spaces or hyphens within the string. How can this be achieved?

Answer 1

You say “… then a sequence of either digits or a single instance of either a hyphen or a space” — take literally, that suggests that the following are all acceptable (I quote the strings when they end with a space):

"+01 "
+01-
+4312345678912345678900

Looking at your code, I guess that you were speaking informally, and what you want could better be described as a plus, two digits, and then a sequence of digits interrupted by at most one hyphen or space, which must not be final. If that’s an accurate paraphrase, you might try

\+\d{2}\d*[- ]?\d+

If you want to allow up to one hyphen and up to one space, it gets a lot more complicated, but it’s still expressible. Assuming you don’t want an empty sequence of digits to be acceptable:

\+\d{2}(\d+|(\d*(-\d+| \d+)?))

[Not tested.]