L-electron-Rare/moodle
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

MDL-31301 use static textlib methods

Browse Source
This commit is contained in:
Petr Skoda
2012-03-03 11:46:13 +01:00
parent 5bbf3cb72b
commit f8311defeb
53 changed files with 191 additions and 329 deletions
Download Patch File Download Diff File Expand all files Collapse all files
lib/tokeniserlib.php
+7 -14
View File
@@ -164,8 +164,6 @@ define('PREG_CLASS_CJK', '\x{3041}-\x{30ff}\x{31f0}-\x{31ff}\x{3400}-\x{4db5}'.
*/
function tokenise_text($text, $stop_words = array(), $overlap_cjk = false, $join_numbers = false) {
$textlib = textlib_get_instance();
// Multipliers for scores of words inside certain HTML tags.
// Note: 'a' must be included for link ranking to work.
$tags = array('h1' => 25,
@@ -204,7 +202,7 @@ function tokenise_text($text, $stop_words = array(), $overlap_cjk = false, $join
if ($tag) {
// Increase or decrease score per word based on tag
list($tagname) = explode(' ', $value, 2);
$tagname = $textlib->strtolower($tagname);
$tagname = textlib::strtolower($tagname);
// Closing or opening tag?
if ($tagname[0] == '/') {
$tagname = substr($tagname, 1);
@@ -243,7 +241,7 @@ function tokenise_text($text, $stop_words = array(), $overlap_cjk = false, $join
$accum .= $word .' ';
$num = is_numeric($word);
// Check word length
if ($num || $textlib->strlen($word) >= MINIMUM_WORD_SIZE) {
if ($num || textlib::strlen($word) >= MINIMUM_WORD_SIZE) {
// Normalize numbers
if ($num && $join_numbers) {
$word = (int)ltrim($word, '-0');
@@ -319,13 +317,11 @@ function tokenise_split($text, $stop_words, $overlap_cjk, $join_numbers) {
*/
function tokenise_simplify($text, $overlap_cjk, $join_numbers) {
$textlib = textlib_get_instance();
// Decode entities to UTF-8
$text = $textlib->entities_to_utf8($text, true);
$text = textlib::entities_to_utf8($text, true);
// Lowercase
$text = $textlib->strtolower($text);
$text = textlib::strtolower($text);
// Simple CJK handling
if ($overlap_cjk) {
@@ -377,10 +373,8 @@ function tokenise_simplify($text, $overlap_cjk, $join_numbers) {
*/
function tokenise_expand_cjk($matches) {
$textlib = textlib_get_instance();
$str = $matches[0];
$l = $textlib->strlen($str);
$l = textlib::strlen($str);
// Passthrough short words
if ($l <= MINIMUM_WORD_SIZE) {
return ' '. $str .' ';
@@ -391,7 +385,7 @@ function tokenise_expand_cjk($matches) {
// Begin loop
for ($i = 0; $i < $l; ++$i) {
// Grab next character
$current = $textlib->substr($str, 0, 1);
$current = textlib::substr($str, 0, 1);
$str = substr($str, strlen($current));
$chars[] = $current;
if ($i >= MINIMUM_WORD_SIZE - 1) {
@@ -408,6 +402,5 @@ function tokenise_expand_cjk($matches) {
*/
function tokenise_truncate_word(&$text) {
$textlib = textlib_get_instance();
$text = $textlib->substr($text, 0, MAXIMUM_WORD_SIZE);
$text = textlib::substr($text, 0, MAXIMUM_WORD_SIZE);
}