MySociety\TheyWorkForYou\Utility\Wikipedia

Complexity

Total Complexity

24

Size/Duplication

Total Lines	182
Duplicated Lines	0 %

Test Coverage

Coverage

57.6%

Importance

Changes

0

2 Methods

Rating	Name	Duplication	Size	Complexity
C	antiTagInTag()	0	59	14
C	wikipedize()	0	109	10

<?php

namespace MySociety\TheyWorkForYou\Utility;

/**
 * Utility functions related to Wikipedia.
 *
 * Based on wikiproxy.php v0.5 04-10-2004 stefan (wholivesat) whitelabel.org
 * scripts/wikipedia-update updates the database of titles each week.
 *
 * The bits of code I didn't borrow from elsewhere (and I've credited where) is
 * licenced under the GPL. Do with it what you will, but this is my first php
 * and my first code for 7 years, so I'd appreciate feedback and suggestions via
 * comments on my blog:
 * http://www.whitelabel.org/archives/002248.html
 * (especially regex optimisations for lines 64 and 65 - ideally a way of making
 * it NOT match if we're within an IMG tag, because then I could drop the
 * antiTaginTag stuff)
 */

class Wikipedia
{

    public static function wikipedize($source) {
        global $format_date_months;
        $months = join('|', array_slice($format_date_months, 1));

        $was_array = false;
        if (is_array($source)) {
            $source = join('|||', $source);
            $was_array = true;
        }

        # Set up various variables
        $capsword = "[A-Z][a-zA-Z'0-9,-]*"; # not starting with number, as catches too much
        $fillerwords = "of|and|in|on|under|the|for";
        $middlewordsre = "(?:\s*(?:$capsword|$fillerwords))*";
        $startwordre = "(?:$capsword)"; # and, of etc. can't appear at ends
        $endwordre = "(?:$capsword|[0-9]+)"; # We might want to catch e.g. a year at the end
        $notfiller = "(?:(?!Of|And|In|On|Under|The|For)$capsword)";

        # Match either "Two Endwords" or "Endword and Some Middle Words"
        $phrasewithfiller = "$startwordre$middlewordsre\s*$endwordre";
        $greedyproperre = "/\b$phrasewithfiller\b/ms";

        # And do a match ignoring things starting with filler words
        $greedynofillerstartre = "/\b$notfiller$middlewordsre\s*$endwordre\b/ms";

        # Match without filler words (so if you have a phrase like
        # "Amnesty International and Human Rights Watch" you also get both parts
        # separately "Amnesty International" and "Human Rights Watch")
        $frugalproperre = "/\b$startwordre(?:\s*$endwordre)+\b/ms";

        # And do a greedy without the first word of a sentence
        $greedynotfirst = "/(?:\.|\?|!)\s+\S+\s+($phrasewithfiller)\b/ms";

        # And one for proper nouns in the possessive
        $greedypossessive = "/\b($phrasewithfiller)'s\b/ms";

        preg_match_all($greedyproperre, $source, $propernounphrases1);
        preg_match_all($frugalproperre, $source, $propernounphrases2);
        preg_match_all($greedynotfirst, $source, $propernounphrases3);
        preg_match_all($greedypossessive, $source, $propernounphrases4);
        preg_match_all($greedynofillerstartre, $source, $propernounphrases5);

        # Three Letter Acronyms
        preg_match_all("/\b[A-Z]{2,}/ms", $source, $acronyms);

        # We don't want no steenking duplicates
        $phrases = array_unique(array_merge($propernounphrases1[0], $propernounphrases2[0],
            $propernounphrases3[1], $propernounphrases4[1], $propernounphrases5[0], $acronyms[0]));
        foreach ($phrases as $i => $phrase) {
            # Ignore months
            if (preg_match("#^($months)\s+\d+$#", $phrase)) {
                continue;
            }
            $phrases[$i] = str_replace(' ', '_', trim($phrase));
        }

        // Assemble the resulting phrases into a parameter array
        $params = array();
        foreach ($phrases as $i => $phrase) {
            $params[':phrase' . $i] = $phrase;
        }

        # Open up a db connection, and whittle our list down even further, against
        # the real titles.
        $matched = array();
        $db = new \ParlDB;
        $source = explode('|||', $source);
        $q = $db->query("SELECT titles.title FROM titles LEFT JOIN titles_ignored ON titles.title=titles_ignored.title WHERE titles.title IN (" . join(',', array_keys($params)) . ") AND titles_ignored.title IS NULL", $params);
        $phrases = array();
        foreach ($q as $row) {
            $phrases[] = $row['title'];
        }

        # Sort into order, largest first
        usort($phrases, function($a, $b) {
            return strlen($a) < strlen($b);
        });

        foreach ($phrases as $wikistring) {
            $phrase = str_replace('_', ' ', $wikistring);
            $wikistring = str_replace("'", "%27", $wikistring);
            $phrase_re = str_replace('_', '[ ]', $wikistring);

            # See if already matched a string this one is contained within
            foreach ($matched as $got) {
                if (strstr($got, $phrase))
                    continue 2;
            }

            twfy_debug("WIKIPEDIA", "Matched '$phrase'");
            # 1 means only replace one match for phrase per paragraph
            $source = preg_replace ('{
            \b(' . $phrase_re . ')\b # Match the phrase itself
            (?!                      # Match as long as the following does *not* apply:
                (?:                  #   Match, possessively, as many strings of:
                 [^<]+               #     non-"<" characters,
                 |                   #     or
                 <(?!/?a\b)          #     a "<" as long as it is not followed by "a"
                )*+                  #     as a word on its own (ie. "<a " or "</a>")
                </a>                 #   Match a "</a>"
            )                        # ie. match as long as we do not find a </a> and have not found a <a>
            }x', "<a href=\"https://en.wikipedia.org/wiki/$wikistring\">\\1</a>", $source, 1);
            array_push($matched, $phrase);
        }

        if (!$was_array)
            $source = join('|||', $source);

        return $source;

    }

    /**
     * Anti Tag-In-Tag
     *
     * Credit: isaac schlueter (lifted from http://uk2.php.net/strip-tags)
     *
     * @todo Remove this, it seems to be redundant.
     */

    public static function antiTagInTag($content = '', $format = 'htmlhead')
    {
      $tagend = -1;
      for( $tagstart = strpos( $content, '<', $tagend + 1 ) ; $tagstart !== false && $tagstart < strlen( $content ); $tagstart = strpos( $content, '<', $tagend ) )
        {
          // got the start of a tag.  Now find the proper end!
          $walker = $tagstart + 1;
          $open = 1;
          while( $open != 0 && $walker < strlen( $content ) )
        {
          $nextopen = strpos( $content, '<', $walker );
          $nextclose = strpos( $content, '>', $walker );
          if( $nextclose === false )
            {    // ERROR! Open waka without close waka!
              // echo '<code>Error in antiTagInTag - malformed tag!</code> ';
              return $content;
            }
          if( $nextopen === false || $nextopen > $nextclose )
            { // No more opens, but there was a close; or, a close happens before the next open.
              // walker goes to the close+1, and open decrements
              $open --;
              $walker = $nextclose + 1;
            }
          elseif( $nextopen < $nextclose )
            { // an open before the next close
              $open ++;
              $walker = $nextopen + 1;
            }
        }
          $tagend = $walker;
          if( $tagend > strlen( $content ) )
        $tagend = strlen( $content );
          else
        {
          $tagend --;
          $tagstart ++;
        }
          $tag = substr( $content, $tagstart, $tagend - $tagstart );
          $tags[] = '<' . $tag . '>';

          if (function_exists('format_to_output')) {
            $newtag = format_to_output($tag, $format);
          } else {
            $newtag = strip_tags($tag);
          }

          $newtags[] = '<' . $newtag . '>';

          if (function_exists('format_to_output')) {
            $newtag = format_to_output($tag, $format);

The assignment to $newtag is dead and can be removed.

          } else {
            $newtag = strip_tags($tag);
          }
        }
      if (isset($tags)&&isset($newtags)) {
      $content = str_replace($tags, $newtags, $content);
      }

    return $content;

    }

}