porterstemmer.module in Porter-Stemmer 7

Porter 2 Stemming for Drupal.

It can either use the PECL stemming library, or a PHP implementation in the includes/standard-stemmer.inc file.

File

porterstemmer.module

<?php

/**
 * @file
 * Porter 2 Stemming for Drupal.
 *
 * It can either use the PECL stemming library, or a PHP implementation in
 * the includes/standard-stemmer.inc file.
 */

/**
 * Regular expression defining a word boundary for Porter Stemmer.
 *
 * A word boundary is anything not a letter or an apostrophe.
 */
define('PORTERSTEMMER_BOUNDARY', "[^a-zA-Z']+");

/**
 * Implements hook_search_preprocess().
 *
 * Stems the words in $text, using the Porter Stemmer 2 algorithm.
 */
function porterstemmer_search_preprocess($text) {

  // Convert text to lower case, and replace special apostrophes with regular
  // apostrophes.
  $text = drupal_strtolower(str_replace('’', "'", $text));

  // Split into words.
  $words = preg_split('/(' . PORTERSTEMMER_BOUNDARY . '+)/', $text, -1, PREG_SPLIT_DELIM_CAPTURE);
  if (!count($words)) {
    return $text;
  }
  $has_pecl_stem = _porterstemmer_pecl_loaded();
  if (!$has_pecl_stem) {
    module_load_include('inc', 'porterstemmer', 'includes/standard-stemmer');
  }

  // Process each word, skipping delimiters.
  $isword = !preg_match('/' . PORTERSTEMMER_BOUNDARY . '/', $words[0]);
  foreach ($words as $k => $word) {
    if ($isword) {
      if ($has_pecl_stem) {
        $words[$k] = stem_english($word);
      }
      else {
        $words[$k] = porterstemmer_stem($word);
      }
    }
    $isword = !$isword;
  }

  // Put it all back together (note that delimiters are in $words).
  return implode('', $words);
}

/**
 * Implements hook_help().
 */
function porterstemmer_help($path, $arg) {
  switch ($path) {
    case 'admin/help#porterstemmer':
      $output = '';
      $output .= '<h3>' . t('About') . '</h3>';
      $output .= '<p>' . t('The Porter Stemmer module implements version 2 of the <a href="@algorithm">Porter Stemmer algorithm</a>, to improve American English-language searching with the core <a href="@search-help">Search module</a>. Stemming reduces a word to its basic root or stem (e.g. "blogging" to "blog") so that variations on a word ("blogs", "blogged", "blogging", "blog") are considered equivalent when searching. This generally results in more relevant results.', array(
        '@search-help' => url('admin/help/search'),
        '@algorithm' => 'http://snowball.tartarus.org/algorithms/english/stemmer.html',
      )) . '</p>';
      return $output;
  }
}

/**
 * Implements hook_sbp_excerpt_match().
 *
 * Allows Porter Stemmer to display better search excerpts with the
 * Search by page module.
 */
function porterstemmer_sbp_excerpt_match($key, $text, $offset, $boundary) {

  // We do not check for the PECL stem here as further parsing logic is needed.
  module_load_include('inc', 'porterstemmer', 'includes/standard-stemmer');

  // Stem the keyword down to its root form.
  $key = porterstemmer_stem($key);

  // In many cases, the root word is a substring of the full word, but not
  // all. The cases where it is not, the root ends in e, i, or y, and if this
  // last letter is removed, the root is a substring of the full word.
  // So remove these letters at the end of the root.
  $didit = FALSE;
  porterstemmer_suffix($key, 'i', '', $didit, NULL, 2) or porterstemmer_suffix($key, 'e', '', $didit, NULL, 2) or porterstemmer_suffix($key, 'y', '', $didit, NULL, 2);

  // Look for this modified key at the start of a word.
  $match = array();
  if (!preg_match('/' . $boundary . '(' . $key . ')/iu', $text, $match, PREG_OFFSET_CAPTURE, $offset)) {

    // Didn't match our modified key.
    return FALSE;
  }

  // If we get here, we have a potential match. Find the end of the word we
  // actually matched, so it can be highlighted (making sure it's a real match
  // for our key).
  $newmatch = array();
  $pos = $match[1][1];

  // Note: Do not use drupal_strlen/drupal_substr here! Need the real PHP
  // string lengths/pos.
  if (preg_match('/' . $boundary . '/iu', $text, $newmatch, PREG_OFFSET_CAPTURE, $pos + strlen($key))) {
    $keyfound = substr($text, $pos, $newmatch[0][1] - $pos);
  }
  else {

    // Assume we're going to the end of the string.
    $keyfound = substr($text, $pos);
  }
  $foundstem = porterstemmer_stem($keyfound);
  porterstemmer_suffix($foundstem, 'i', '', $didit, NULL, 2) or porterstemmer_suffix($foundstem, 'e', '', $didit, NULL, 2) or porterstemmer_suffix($foundstem, 'y', '', $didit, NULL, 2);

  // Both $foundstem and $key may contain upper case.
  if (drupal_strtolower($foundstem) == drupal_strtolower($key)) {
    return array(
      'where' => $pos,
      'keyword' => $keyfound,
    );
  }

  // If we get here, then it was a false match, and we should probably
  // search again later in the string.
  return porterstemmer_sbp_excerpt_match($key, $text, $pos + strlen($keyfound), $boundary);
}

/**
 * Checks to see if the PECL stem extension has been loaded.
 *
 * @return bool
 *    TRUE if the stem_english() function from the PECL stem library can be
 *    used, FALSE if not.
 */
function _porterstemmer_pecl_loaded() {
  static $has_pecl_stem = FALSE;
  static $already_checked = FALSE;
  if ($already_checked) {
    return $has_pecl_stem;
  }
  $has_pecl_stem = extension_loaded('stem') && function_exists('stem_english');
  $already_checked = TRUE;
  return $has_pecl_stem;
}

Functions

Name	Description
porterstemmer_help	Implements hook_help().
porterstemmer_sbp_excerpt_match	Implements hook_sbp_excerpt_match().
porterstemmer_search_preprocess	Implements hook_search_preprocess().
_porterstemmer_pecl_loaded	Checks to see if the PECL stem extension has been loaded.

Constants

Name	Description
PORTERSTEMMER_BOUNDARY	Regular expression defining a word boundary for Porter Stemmer.