Multiple occurences of same character in a string regexp - Python

The following should work:

^([A-Z])(?!.?\1)([A-Z])(?!\2)([A-Z])[a-z]\1\1\3$

For example:

>>> regex = re.compile(r'^([A-Z])(?!.?\1)([A-Z])(?!\2)([A-Z])[a-z]\1\1\3$')
>>> regex.match('ABAaAAA')  # fails: first three are not different
>>> regex.match('ABCaABC')  # fails: first two of second three are not first char
>>> regex.match('ABCaAAB')  # fails: last char is not last of first three
>>> regex.match('ABCaAAC')  # matches!
<_sre.SRE_Match object at 0x7fe09a44a880>

Explanation:

^          # start of string
([A-Z])    # match any uppercase character, place in \1
(?!.?\1)   # fail if either of the next two characters are the previous character
([A-Z])    # match any uppercase character, place in \2
(?!\2)     # fail if next character is same as the previous character
([A-Z])    # match any uppercase character, place in \3
[a-z]      # match any lowercase character
\1         # match capture group 1
\1         # match capture group 1
\3         # match capture group 3
$          # end of string

If you want to pull these matches out from a larger chunk of text, just get rid of the ^ and $ and use regex.search() or regex.findall().

You may however find the following approach easier to understand, it uses regex for the basic validation but then uses normal string operations to test all of the extra requirements:

def validate(s):
    return (re.match(r'^[A-Z]{3}[a-z][A-Z]{3}$', s) and s[4] == s[0] and 
            s[5] == s[0] and s[-1] == s[2] and len(set(s[:3])) == 3)

>>> validate('ABAaAAA')
False
>>> validate('ABCaABC')
False
>>> validate('ABCaAAB')
False
>>> validate('ABCaAAC')
True