CODEWITHSUNDEEP
pythonregexpython-3.4
user1602492
user1602492

Reputation:

Fail to understand to me strange re behaviour

Can someone please explain to me, why I am getting a different result when I remove the lines containing '# duplicate'?

import re    

def nysiis(term: str) -> str:
    """
    returns New York State Identification and Intelligence Algorithm (NYSIIS) code for given term
    """

    if not len(term):
        return ''
    else:
        term = term.upper()

    table = {
        r'\W+': '',  # remove whitespace and non-word characters

        r'^MAC': 'MCC',
        r'^KN': 'NN',
        r'K': 'C',
        r'PH|PF': 'FF',
        r'SCH': 'SSS',

        r'(EE|IE)$': 'Y',
        r'(DT|ND|NT|RD|RT)$': 'D',

        # From now on first letter must no longer change.
        r'(?<!^)EV': 'AF',
        r'(?<!^)[AEIOU]': 'A',
        r'(?<!^)Q': 'G',
        r'(?<!^)Z': 'S',
        r'(?<!^)(?:M|KN)': 'N',

        r'(?<!^)([^AEIOUY])H': r'\1',
        r'(?<!^)(.)H[^AEIOUY]': r'\1',
        r'(?<!^)([AEIOUY])W': r'\1',

        r'AY$': r'Y',
        r'S$': r'',

        r'(\w)\1+': r'\1',  # original
        r'A+$': r''  # original
    }

    for k, v in table.items():
        term = re.sub(k, v, term)

    table = {  # duplicate
        r'(\w)\1+': r'\1',  # duplicate
        r'A+$': r''  # duplicate
    }  # duplicate

    for k, v in table.items():  # duplicate
        term = re.sub(k, v, term)  # duplicate

    return term

if __name__ == '__main__':
    names = [
        'Bishop', 'Carlson', 'Carr', 'Chapman', 'Franklin',
        'Greene', 'Harper', 'Jacobs', 'Larson', 'Lawrence',
        'Lawson', 'Louis, XVI', 'Lynch', 'Mackenzie', 'Matthews',
        'McCormack', 'McDaniel', 'McDonald', 'Mclaughlin', 'Morrison',
        "O'Banion", "O'Brien", 'Richards', 'Silva', 'Watkins',
        'Wheeler', 'Willis', 'brown, sr', 'browne, III', 'browne, IV',
        'knight', 'mitchell', "o'daniel",
    ]

    for name in names:
        print('%15s: %s' % (name, nysiis(name)))

Upvotes: 1

Views: 67

Answers (1)

Ned Batchelder
Ned Batchelder

Reputation: 375574

You don't want to use a dict for your substitutions: the order of iteration is not the same as the order you listed them in. If you change your dicts to a list of pairs instead, then it works as you expect.

table = [
    (r'\W+', ''),
    #...
]

for k, v in table:
    ...

Upvotes: 1

Related Questions