github/diceware
1
0
Fork 0
mirror of https://github.com/dmuth/diceware.git synced 2024-11-25 02:06:33 -05:00
Code Issues Projects Releases Packages Wiki Activity

Added function to extract words from Peter Novig's list.

Browse source
This commit is contained in:
Douglas Muth 2016-07-10 21:45:25 -04:00
parent 1cfd66a0ea
commit 9e818636f1
1 changed files with 61 additions and 5 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

66
wordlist/create-wordlist.php
View file

@ -16,7 +16,7 @@ if (php_sapi_name() != "cli") {
/**
* Read in our wordlist and return an array with all words that
* Read in our wordlist from Google and return an array with all words that
* passed validation.
*
* @param string $filename The filename
@ -24,7 +24,7 @@ if (php_sapi_name() != "cli") {
* @return array An array of words
*
*/
function readWordList($filename) {
function readWordListGoogle($filename) {
$retval = array();
@ -59,7 +59,61 @@ function readWordList($filename) {
return($retval);
} // End of readWordList()
} // End of readWordListGoogle()
/**
* Read in our wordlist from Google and return an array with all words that
* passed validation.
*
* @param string $filename The filename
*
* @return array An array of words
*
*/
function readWordListPeterNovig($filename) {
$retval = array();
$fp = @fopen($filename, "r");
if (!$fp) {
throw new Exception("Could not open '$filename' for reading");
}
$count = 0;
while ($line = fgets($fp)) {
$line = rtrim($line);
list($word, $freq) = explode("\t", $line);
$len = strlen($word);
//
// Keep all words between 5 and 8 characters
//
if ($len < 5 || $len > 8) {
continue;
}
$retval[] = $word;
$count++;
if ($count > 7776) {
break;
}
}
//
// Put the words in alphabetical order for my own sanity.
//
sort($retval);
fclose($fp);
return($retval);
} // End of readWordListPeterNovig()
/**
@ -144,8 +198,10 @@ function main() {
//
// Read our file
//
$filename = "google-10000-english.txt";
$words = readWordList($filename);
//$filename = "google-10000-english.txt";
//$words = readWordListGoogle($filename);
$filename = "count_1w.txt";
$words = readWordListPeterNovig($filename);
//
// Match words to dicerolls