[Tutor] word challenge

[gyro funch]

Raymond Hettinger wrote:
>>Since the number of paths through the phrase is very large,
>>a good bet is to start with a dictionary of available words
>>and test to see if they can be found in the phrase (while
>>paying attention to letter order and the requirement that
>>each group be represented).

<snip excellent comments and inspiring code>


Excellent Raymond!
Thanks for the great analysis and well though out code.

On the other end of the spectrum is my code below.
This uninspiring, brute-force code relies on the 'batteries included' aspect of Python and makes the regular expression engine do all of the work. We basically construct a regular expression satisfying the criteria outlined earlier and 'let it rip' on the word list. As Raymond and others pointed out, the word list could be pre-processed to eliminate words clearly not satisfying the criteria.

Thanks to all for your great code and efforts!

-g


My apologies if my mailer screws up the code formatting.


class FindWords(object):
    def __init__(self, wfile = '/usr/dict/words'):                                                            
        self.wfile = wfile
        self.wlist = []
        self._make_word_list()

    def _make_word_list(self):
        import re
        wrd_re = re.compile(r'^[a-z]+$')
        fh = open(self.wfile,'r')
        all_lines = fh.readlines()
        fh.close()
        for line in all_lines:
            word = line.strip().lower()
            if wrd_re.search(word):
                self.wlist.append(word)
        return self.wlist

    def _build_regex(self, phrase):
        mlist = []
        for word in phrase.split():
            wr,wlist = range(len(word)),[]
            for j in wr:
                wre = ''
                for i in wr:
                    if i == j:
                        char = ''
                    else:
                        char = '?'
                    wre += '%s%s' % (word[i],char)
                wlist.append(wre)
            mlist.append('(%s)' % '|'.join(wlist))
        full_re = '^%s$' % ''.join(mlist)
        return full_re

    def get_words(self, phrase):
        import re
        found_words = []
        p_regex = re.compile(self._build_regex(phrase))
        for word in self.wlist:
            if p_regex.search(word):
                found_words.append(word)
        return found_words
                                         
def main():
    mw = FindWords()
    phrase = 'handy cool phrase'
    wl = mw.get_words(phrase)
    # let's get the longest and shortest words
    wdict = {}
    for word in wl:
        wdict.setdefault(len(word),[]).append(word)
    wrds = wdict.keys()
    mn,mx = min(wrds),max(wrds)
    print wdict[mn]
    print wdict[mx]
    return



_______________________________________________
Express Yourself - Share Your Mood in Emails!
Visit www.SmileyCentral.com - the happiest place on the Web.