Updated batch.js for 10 words

Added import.js to import dictionary
Updated license with WordNet info
Updated puzzles.js and script for 10 words
This commit is contained in:
Russell Beattie
2013-03-12 15:10:46 -07:00
parent 7eefd5e55e
commit b8e7d91a3c
10 changed files with 33707 additions and 23581 deletions

34
utils/README.md Normal file
View File

@@ -0,0 +1,34 @@
# Utils
The utils directory contain Node.js Javascript files to create a word dictionary and generate a puzzle.
* import.js - Imports Princeton's WordNet open dictionary database files into a words.txt file
* batch.js - Creates a new puzzle.js file containing 999 word puzzles from the words.txt file created above.
# import.js instructions
Imports and cleans dictionary files provided by WordNet 3.0
Download dictionary file (look for "DATABASE FILES ONLY" package):
http://wordnet.princeton.edu/wordnet/download/current-version/
Decompress and copy data.adj, data.adv, data.noun and data.verb files
to same directory as this script, then run:
$ node import.js
This will produce a words.txt dictionary file of 4-10 letter words
to be used by batch.js puzzle generator.
# batch.js instructions
After creating a words.txt dictionary, run:
$ node batch.js
This will create a puzzles.js file in the root directory of the project.

136
utils/batch.js Normal file
View File

@@ -0,0 +1,136 @@
'use strict';
var fs = require('fs');
// List out the possible ways to divide a word into squares with 2 or 3 letters.
var squaresList = {};
squaresList[2] = [
{wordlen: 4, squares: [2,2] },
{wordlen: 5, squares: [2,3] },
{wordlen: 5, squares: [3,2] },
{wordlen: 6, squares: [3,3] }
];
squaresList[3] = [
{wordlen: 6, squares: [2,2,2] },
{wordlen: 7, squares: [2,2,3] },
{wordlen: 7, squares: [2,3,2] },
{wordlen: 7, squares: [3,2,2] },
{wordlen: 8, squares: [3,3,2] },
{wordlen: 8, squares: [3,2,3] },
{wordlen: 8, squares: [2,3,3] },
{wordlen: 9, squares: [3,3,3] },
];
squaresList[4] = [
{wordlen: 8, squares: [2,2,2,2] },
{wordlen: 9, squares: [2,2,2,3] },
{wordlen: 9, squares: [2,2,3,2] },
{wordlen: 9, squares: [2,3,2,2] },
{wordlen: 9, squares: [3,2,2,2] },
{wordlen: 10, squares: [3,3,2,2]},
{wordlen: 10, squares: [3,2,3,2]},
{wordlen: 10, squares: [3,2,2,3]},
{wordlen: 10, squares: [2,2,3,3]},
{wordlen: 10, squares: [2,3,2,3]}
];
// To fill 30 squares with
var squaresWordLengths = [
[4, 4, 3, 3, 3, 3, 3, 3, 2, 2],
[4, 4, 4, 3, 3, 3, 3, 2, 2, 2]
];
// pull in the word list, divide into lists by word length
var lines = fs.readFileSync('./words.txt').toString().split('\n');
var words = {};
for (var i = 0; i < lines.length; i++) {
var line = lines[i].trim();
if(line !== ''){
var ln = line.split(' - ');
var s = ln[0].length;
if(!words[s]){
words[s] = [];
}
words[s].push({
word: ln[0],
def : ln[1].charAt(0).toUpperCase() + ln[1].substr(1) + '.'
});
}
}
// generate puzzles
var puzzles = [];
for(var p = 1; p <= 1000; p++){
var puzzle = {};
var cwords = [];
var csquares = [];
var squaresNeeded = squaresWordLengths[Math.floor(Math.random() * squaresWordLengths.length)];
for (var i = 0; i < squaresNeeded.length; i++) {
var squareTypes = squaresList[squaresNeeded[i]];
var squareType = squareTypes[Math.floor((Math.random() * squareTypes.length))];
var wordlist = words[squareType.wordlen];
var entry = wordlist[Math.floor((Math.random() * wordlist.length))];
cwords.push({
word: rot13(entry.word),
def: entry.def
});
var start = 0;
var squares = squareType.squares;
for (var x = 0; x < squares.length; x++) {
var letnum = squares[x];
csquares.push(entry.word.substr(start, letnum));
start += letnum;
}
}
cwords = shuffle(cwords);
csquares = shuffle(csquares);
puzzle['clues'] = cwords;
puzzle['squares'] = csquares;
puzzles.push(puzzle);
}
var json = JSON.stringify(puzzles);
var js = 'var puzzles = ' + json;
fs.writeFileSync('../puzzles.js', js);
function rot13(str) {
return str.replace(/[a-zA-Z]/g, function(c) {
return String.fromCharCode((c <= 'Z' ? 90 : 122) >= (c = c.charCodeAt(0) + 13) ? c : c - 26);
});
}
function shuffle(arr) {
for(var j, x, i = arr.length; i; j = parseInt(Math.random() * i), x = arr[--i], arr[i] = arr[j], arr[j] = x);
return arr;
}

122
utils/import.js Normal file
View File

@@ -0,0 +1,122 @@
/*
import.js
Imports and cleans dictionary files provided by WordNet 3.0
Download dictionary file (look for "DATABASE FILES ONLY" package):
http://wordnet.princeton.edu/wordnet/download/current-version/
Decompress and copy data.adj, data.adv, data.noun and data.verb files
to same directory as this script, then run:
$ node import.js
This will produce a words.txt dictionary file of 4-10 letter words
to be used by batch.js puzzle generator.
*/
var fs = require('fs');
var util = require('util');
var files = ['data.adj','data.adv','data.verb','data.noun'];
var text = '';
util.log('Starting import');
for(var f = 0; f < files.length; f++){
text += fs.readFileSync('./' + files[f]).toString()
}
var lines = text.split('\n');
var wordList = [];
var defList = [];
var perc = Math.floor(lines.length / 100);
for (var i = 0; i < lines.length; i++) {
if(lines[i].substr(0, 2) == ' '){
continue;
}
if(i % perc == 0){
util.log(Math.floor((i/lines.length) * 100) + '%');
}
var line = lines[i].trim();
var ln = line.split('|');
var wordPartStr = ln[0].trim();
var wordPart = wordPartStr.split(' ');
if(wordPart[4]){
var word = wordPart[4].trim();
var def = ln[1].trim();
if(wordList.indexOf(word) !== -1 || word.charAt(0).toUpperCase() == word.charAt(0) || word.indexOf('\'') !== -1 || word.indexOf('-') !== -1 || word.indexOf('_') !== -1 || word.indexOf('(') !== -1){
continue;
}
if(def.indexOf(';') !== -1){
def = def.substr(0, def.indexOf(';'));
}
def = def.charAt(0).toUpperCase() + def.substr(1) + '.';
wordList.push(word);
defList.push(def);
}
}
var words = '';
for(var x = 4; x < 11; x++){
for(var w = 0; w < wordList.length; w++){
if(wordList[w].length == x){
//console.log(wordList[w] + ' - ' + defList[w]);
words += wordList[w] + ' - ' + defList[w] + '\n';
}
}
}
fs.writeFileSync('./words.txt', words);
/*
var words = {};
for(var x = 4; x < 11; x++){
var wordList = [];
for (var i = 0; i < lines.length; i++) {
var line = lines[i].trim();
if(line !== ''){
var ln = line.split('|');
var word = ln[0].trim();
var def = ln[1].trim();
if(wordList.indexOf(word) !== -1){
continue;
}
if(def.indexOf(';') !== -1){
def = def.substr(0, def.indexOf(';'));
}
def = def.charAt(0).toUpperCase() + def.substr(1) + '.';
var s = word.length;
if(s == x){
wordList.push(word);
console.log(word + ' - ' + def);
}
}
}
}
*/

33505
utils/words.txt Normal file

File diff suppressed because it is too large Load Diff