Crucial Hints - The Spoiler Page

In the last page I presented your programming challenge and asked you to devise ways to do the Countdown Words Game.

What it all comes down to is this: given 9 random letters, find all 9, 8 and 7 letter words which use combinations of those 9 letters.

The Logic

Nine letter words are the easiest: hits will contain all 9 of the input letters.

Eight letter words get a little more difficult. There will be 9 searches, each one will in essence exclude one of the 9 letters.

Seven letter words are the most complex. There will be 36 searches, ie, the number of combinations of 7 from 9, leaving out two letters at a time.

So we're going to be doing a total of 46 searches.

The Challenge

The challenge is simple: how can you tell if 'dosarmn' equals 'randoms' in the words list?

The answer: you don't! You need to do something different.

What we have to do:

• The text file of words is no good as it stands. We need a massaged version, so:
  We write a program which looks at each word in the dictionary and makes a sorted version of it. We then store the sorted version along with the proper word. This could take the form of a database table with two columns, sorted and proper. However, in Python, in which I wrote my solution, things are just as fast, if not faster, with a plain text file. Here is a sample from my file, which is saved as a CSV file:
  
  directing,cdegiinrt
  direction,cdeiinort
  directive,cdeeiirtv
  directly,cdeilrty
  director,cdeiorrt
  directors,cdeiorrst
  directory,cdeiorrty
  Directory,cdeiorrty
  directrix,cdeiirrtx
  directs,cdeirst
  Diredawa,aaddeirw
  direful,defilru
  direfully,defillruy
  direly,deilry
  dirempt,deimprt
  direness,deeinrss
  direption,deiinoprt
  direst,deirst
  direxit,deiirtx
  dirged,ddegir
  dirgeful,defgilru
  dirgelike,deegiiklr
  dirgeman,adegimnr
  dirges,degirs
  dirgie,degiir
  dirging,dggiinr
  dirgler,degilrr
  dirham,adhimr
  
  I also have my dictionary in a database table. A sample:
  
  abnegator aabegnort
  abnerval aabelnrv
  abneural aabelnru
  abnormal aablmnor
  abnormals aablmnors
  abnormity abimnorty
  abnormous abmnoorsu
  aboardage aaabdegor
  abococket abccekoot
  abodement abdeemnot
  aboding abdgino
  abogado aabdgoo
  abogados aabdgoos
  aboideau aabdeiou
  aboideaus aabdeiosu
  aboideaux aabdeioux
  aboiteau aabeiotu
  aboiteaus aabeiostu
  aboiteaux aabeiotux
  
  That's two separate columns.


• The second crucial point is that once the user has input the nine random letters this nine-letter word has to be sorted. Thus, from our example, regesnate becomes aeeegnrst.
  Now, the dark clouds are lifted! All we are looking at now is a simple word-matching exercise! Look for aeeegnrst to find the nines, then remove one letter at a time from aeeegnrst and look for the resulting pattern. Finally remove two letters at a time . . . etc etc.


• Finally - no duplicates in the displayed results. How? Well in my Python solution I use lists. Every time a hit is found the appropriate list - 9s, 8s or 7s - is checked to see if that word is already there. If so, ignore. The list is automatically sorted so this approach also satisfies our ordered results requirement.


And that's it! I'll leave the details to you. The above is not the coding answer; it's the secrets of how to get started, turning what at first may seem to be enormously difficult into a fairly simple text searching operation.

Discussion, comments, questions - all very welcome at alansworld {at} gmail.com