fragcount.py

from __future__ import absolute_import, print_function, division, unicode_literals

import re
from collections import defaultdict, Counter


INITIAL_SYMBOLS = b'abc'


def count_fragments(lines, initial_symbols=None):
    initial_symbols = initial_symbols or INITIAL_SYMBOLS

    # # @KK: Built-in 're' module has wonky support for repeated groups, so this pattern (which
    # # would match the whole thing in one go and makes more sense) doesn't really work.
    # pattern = re.compile(r'^([' + re.escape(initial_symbols) + '][^' + re.escape(initial_symbols) + ']*)*$')

    pattern = re.compile(r'^([' + re.escape(initial_symbols) + '][^' + re.escape(initial_symbols) + ']*)(.*)$')
    
    fragments = defaultdict(Counter)
    for i, line in enumerate(lines):
        while line:
            m = pattern.match(line)
            if not m:
                raise AssertionError('Line {} does not match regex!'.format(i+1))
            frag, line = m.groups()
            fragments[frag[0]].update((frag,))
    
    return fragments
    
def format_counts(fragments):    
    for initial, frag_counts in fragments.items():
        print('{}'.format(initial))
        for qty, frag in sorted(((qty, frag) for frag, qty in frag_counts.items()), reverse=True):
            print('  {:>5}  {}'.format(qty, frag))
        
def main():
    import sys
    with open(sys.argv[1], 'rb') as f:
        format_counts(count_fragments(line for line in f))

        
if __name__ == '__main__':
    main()

fragcount_test.py

from __future__ import absolute_import, print_function, division, unicode_literals

import pytest

from fragcount import count_fragments


@pytest.mark.parametrize(('lines', 'expected'), [
    ([b'axyzbyazyzz'],
     dict(a=dict(axyz=1, azyzz=1), b=dict(by=1))),
])
def test_count_fragments(lines, expected):
    result = count_fragments(lines, initial_symbols=b'abc')
    result_pod = {k: {x: y for x, y in v.items()} for k, v in result.items()}
    assert result_pod == expected