Difference between PHP regex and JavaScript regex

1) /^[\d]+$/
2) /^([\u0600-\u06FF]+\s)*[\u0600-\u06FF]+$/
3) /^([\u0600-\u06FF]+\d*\s)*[\u0600-\u06FF]+\d*$/

/u is not supported, since Javascript regexes only supports unicode in terms of codepoints. \x{???} (unicode codepoints) should be written \u???? in Javascript regex (always 4 digits 0 padded)

In these cases, the following applies to the rest of the regex:

• \s in Javascript is treated as unicode
• \d isn't, which means only ASCII digits (0-9) are accepted.

This means we specifically have to allow "foreign" numerals, e.g. Persian (codepoints 06F0-06F9):

1) /^[\d\u06F0-\u06F9]+$/
2) /^([\u0600-\u06FF]+\s)*[\u0600-\u06FF]+$/
3) /^([\u0600-\u06FF]+[\d\u06F0-\u06F9]*\s)*[\u0600-\u06FF]+[\d\u06F0-\u06F9]*$/

(Remove \d if ASCII digits shouldn't be accepted)

Not sure what the brackets are supposed to be doing in example 1, originally they could be written:

1) /^\d+$/

But to add the Persian numerals, we need them, see above.

Update

Spry character masking, however, only wants a regex to be applied on each entered character - i.e., we can't actually do pattern matching, it's just a "list" of accepted characters in all places, in which case:

1      ) /[\u06F0-\u06F9\d]/      // match 0-9 and Persian numerals
2 and 3) /[\u0600-\u06FF\d\s]/    // match all Persian characters (includes numerals), 0-9 and space

Once again, remove \d if you don't want to accept 0-9.

Update 2

Now... using regex for validation with Spry:

var checkFullName = function(value, options)
{
   // Match with the by now well-known regex:
   if (value.match(/^([\u0600-\u06FF]+\s)*[\u0600-\u06FF]+$/))
   {
      return true;
   }
   return false;
}

var sprytextfield =
     new Spry.Widget.ValidationTextField(
          "sprytextfield", 
          "custom", 
          { validation: checkFullName, validateOn: ["blur", "change"] }
     );

A similar custom function can be made for case 3.

See examples from Adobe labs