Find word by given characters

[Ben Bacarisse]

Bischoop <Bischoop at vimart.net> writes:

> Let me clarify what I want to do:
> We all know Scrabble game.
> there's a file with Dictionary containing word in each line, the idea is
> to input some letters for example mentioned earlier: att, the script
> supposed to find all words in which letters: 'a','t','t' occur and print
> these words. It supposed not to print word: 'auto' because there's only
> one 't' but words such as: 'toast', 'toasty', 'tolerant' are meeting the
> criteria becase we have in these words one 'a' and two 't' as user has
> input.

In Scrabble, you will usually know the order you want for the known
letters, so you could just make a regular expression: t.*a.*t

In fact, you often know the gaps, so you might even be able to match
something more specific like t.a..t instead.

If you don't know the order, you can still make a chain of matches.  For
example, matching for t.*t and then also a.  Every distinct letter needs
a match, and repeated letters need a single match with .* between them.

And since Python REs have (?=...) look ahead syntax you could even make
a single pattern like this:

  (?=.*t.*t)(?=.*a)

While there are probably better ways in Python, this is what I'd do on
the command line.  For example

$ grep -P '(?=.*t.*t)(?=.*a)' word-list | wc -l
45677
$ grep 't.*t' word-list | grep a | wc -l 
45677

-- 
Ben.