Check word prefixes to reduce problem space

Produce a prefix wordlist from a dictionary wordlist using the
`filter_prefix.sh` script.

boggler.py

@@ -9,7 +9,7 @@ def readBoggleFile(file):
 
 def readWordlist(file):
     with open(file, 'r') as f:
-        return {k:0 for k in f.read().split()}
+        return {k:1 for k in f.read().split()}
         
 def getAdjacentIndexes(board, row, col):
     maxHeight = len(board) - 1
@@ -34,22 +34,24 @@ def getAdjacentIndexes(board, row, col):
 
     return indexes 
 
-def findWordsByIndex(board, wordlist, row, col, depth, adjacentIndexes, prefix = '', path = [], foundWords = {}):
-    if depth == 0:
+def findWordsByIndex(board, wordlist, row, col, depth, adjacentIndexes, prefix = '', path = [], foundWords = {}, prefixes = None, maxDepth = 5, prefixLen = 0):
+    if depth == maxDepth:
         return {}
 
+    if prefixes != None and depth == prefixLen and prefix not in prefixes:
+            return {}
+    prefix = prefix + board[row][col]
+
     foundWords = deepcopy(foundWords)
     path = deepcopy(path)
 
-    prefix = prefix + board[row][col]
     path.append((row, col))
     if prefix in wordlist:
         foundWords[prefix] = 1
 
     adjacents = [x for x in adjacentIndexes[(row, col)] if x not in path]
-    #print('adj:', adjacents, 'p:', (row,col), 'pre:', prefix)
     for a in adjacents:
-        newWords = findWordsByIndex(board, wordlist, a[0], a[1], depth-1, adjacentIndexes, prefix, path, foundWords)
+        newWords = findWordsByIndex(board, wordlist, a[0], a[1], depth+1, adjacentIndexes, prefix, path, foundWords, prefixes, maxDepth, prefixLen)
         for k,v in newWords.items():
             if k in wordlist and k in foundWords:
                 # Add and increment word counts
@@ -60,13 +62,13 @@ def findWordsByIndex(board, wordlist, row, col, depth, adjacentIndexes, prefix =
 
     return foundWords
 
-def findWords(board, adjacentIndexes, maxDepth = 5):
+def findWords(board, adjacentIndexes, maxDepth = 5, prefixes = None, prefixLen = 0):
     words = {}
     for row in range(len(board)):
         for col in range(len(board[0])):
             # Find words for each row/col starting points
             wordlist = readWordlist(WORDLISTS_PATH + 'words_' + board[row][col] + '.txt')
-            newWords = findWordsByIndex(board, wordlist, row, col, maxDepth, adjacentIndexes)
+            newWords = findWordsByIndex(board, wordlist, row, col, 0, adjacentIndexes, prefixes = prefixes, maxDepth = maxDepth, prefixLen = prefixLen)
             for w,v in newWords.items():
                 # Add and increment word counts
                 words[w] = words[w] + v if w in words.keys() else v
@@ -75,8 +77,8 @@ def findWords(board, adjacentIndexes, maxDepth = 5):
 
 
 if __name__ == '__main__':
-    if len(sys.argv) != 2:
-        print('Usage: python3 blogger.py boggle_board_file.txt')
+    if len(sys.argv) != 3:
+        print('Usage: python3 blogger.py BOARD_FILE MAX_WORD_LENGTH')
     else:
         board = readBoggleFile(sys.argv[1])
         print(board)
@@ -85,14 +87,14 @@ if __name__ == '__main__':
             for c in range(len(board)):
                 adjacentIndexes[(r,c)] = getAdjacentIndexes(board, r, c)
 
-
         adjacentLetters = {k:[board[pos[0]][pos[1]] for pos in v] for (k,v) in adjacentIndexes.items()}
 
-
-        maxSize = 8
-        words = findWords(board, adjacentIndexes, maxSize)
-        words = [w for w in words if len(w) > 2]
+        maxSize = int(sys.argv[2], 10)
+        prefixes = readWordlist('./wordlists/all/prefix4.txt')
+        words = findWords(board, adjacentIndexes, maxSize, prefixes, 4)
+        #words = findWords(board, adjacentIndexes, maxSize)
         for w in words:
             print(w)
-        #print([w for w in words if len(w) == 3])
 
+        #words = [w for w in words if len(w) > 2]
+        #print([w for w in words if len(w) == 3])

filter_prefix.sh

@@ -0,0 +1,9 @@
+#!/bin/sh
+
+if [ -z $1 ] || [ -z $2 ]; then
+    echo "Usage: ./filter_prefix.sh WORDLIST PREFIX_LENGTH"
+    exit 0
+else
+    cat "$1" | pcregrep -o1 "^([a-zA-Z]{$2})" | uniq
+    exit 0
+fi