an_interesting_problem.py

# def ngrams(tokens, n):
#     num_tokens = len(tokens)
#     if n > num_tokens:
#         return
#     ngram_list = []
#     for i in xrange(num_tokens):
#         token_groups = []
#         for j in xrange(i, i+n):
#             if j < num_tokens:
#                 token_groups.append(tokens[j])
#         ngram_list.append(token_groups)
#     return ngram_list
    
_map = {
    ('A'): 'X1',
    ('B'): 'X2',
    ('C'): 'X3',
    ('D'): 'X4',
    ('E'): 'X5',
    ('A', 'B'): 'Y1',
    ('B', 'C', 'D'): 'Y2',
    ('C', 'D', 'F'): 'Y3',
    ('C', 'D', 'G'): 'Y4'
 }

def get_similar_sequence_keys(_map, sequence):
    similar_sequence_keys = []
    for k in _map:
        if ''.join(k).startswith(''.join(sequence)):
            similar_sequence_keys.append(k)
    return similar_sequence_keys


def run():
    import re
    token_regex = re.compile(r'\S+')
    tokens = token_regex.findall(raw_input().strip())

    token_length = len(tokens)
    output = []
    index = 0
    while index < len(tokens):
        token = tokens[index]
        similar_sequence_keys = sorted(get_similar_sequence_keys(_map, token), reverse=True)
        if not similar_sequence_keys:
            break
        for sequence_key in similar_sequence_keys:
            longest_sequence_length = len(sequence_key)
            last_index_limit = index + longest_sequence_length
            token_sequence_list = tokens[index:last_index_limit]
            if len(token_sequence_list) == 1:
                token_tuple = token_sequence_list[0]
            else:
                token_tuple = tuple(token_sequence_list)
            if token_tuple in similar_sequence_keys:
                output.append(_map.get(token_tuple))
                index = last_index_limit
                break

    return output

if __name__ == "__main__":
    output = run()
    print output