/**
* check for plagiarism with copyscape
* return true if there's an issue
*/
private static function copyCheck($t)
{
$threshold = 0.05;
$result = '';
$r = Revision::newFromTitle($t);
if (!$r) {
return 'No such article';
}
$text = Wikitext::flatten($r->getText());
$res = copyscape_api_text_search_internet($text, 'ISO-8859-1', 2);
if ($res['count']) {
$words = $res['querywords'];
foreach ($res['result'] as $r) {
if (!preg_match("@^http://[a-z0-9]*.(wikihow|whstatic|youtube).com@i", $r['url'])) {
//if ($r['minwordsmatched'] / $words > $threshold) {
//we got one!
$result .= '<b>Plagiarized:</b> <a href="' . $r['url'] . '">' . $r['url'] . '</a><br />';
//}
}
}
} else {
$result = '';
}
return $result;
}
$dbr = wfGetDB(DB_SLAVE);
$sql = "SELECT page_title, page_id FROM page WHERE page_is_redirect=0 AND page_namespace=" . NS_MAIN;
$res = $dbr->query($sql, __FILE__);
$fp = fopen('short-intros.csv', 'w');
if (!$fp) {
die("could not open file for write\n");
}
fputcsv($fp, array('page_id', 'URL', 'has_template', 'intro_length', 'intro'));
foreach ($res as $row) {
$title = Title::newFromDBkey($row->page_title);
if (!$title) {
print "Can't make title out of {$row->page_title}\n";
continue;
}
$rev = Revision::newFromTitle($title);
$wikitext = $rev->getText();
$intro = Article::getSection($wikitext, 0);
$flat = Wikitext::flatten($intro);
$flat = trim($flat);
$len = mb_strlen($flat);
if ($len < 50) {
// check whether it has either the {{intro or {{introduction template
$hasTemplate = strpos(strtolower($intro), '{{intro') !== false;
$fields = array($row->page_id, 'http://www.wikihow.com/' . $title->getPartialURL(), $hasTemplate ? 'y' : 'n', $len, $flat);
fputcsv($fp, $fields);
if (@++$i % 100 == 0) {
print "article {$i}\n";
}
}
}
fclose($fp);
function getIntroText(&$t)
{
$r = Revision::newFromTitle($t);
$intro = Article::getSection($r->getText(), 0);
return Wikitext::flatten($intro);
}
public function mobileSearch($q, $start, $limit = 20)
{
global $wgOut, $wgMemc;
// Don't return more than 50 search results at a time to prevent abuse
if ($limit > 50) {
$limit = 50;
}
$key = wfMemcKey("MobileSearch", str_replace(" ", "-", $q), $start, $limit);
if ($val = $wgMemc->get($key)) {
return $val;
}
$contents = $this->googleSearchResultTitles($q, $start, $limit, 0, self::SEARCH_MOBILE);
$results = array();
foreach ($contents as $t) {
// Only return articles
if ($t->getNamespace() != NS_MAIN) {
continue;
}
$result = array();
$result['title'] = $t->getText();
$result['url'] = $t->getFullURL();
$result['imgurl'] = wfGetPad(SkinWikihowskin::getGalleryImage($t, 103, 80));
$result['intro'] = null;
if ($r = Revision::newFromid($t->getLatestRevID())) {
$intro = Wikitext::getIntro($r->getText());
$intro = trim(Wikitext::flatten($intro));
$result['intro'] = substr($intro, 0, 180);
// Put an ellipsis on the end
$len = strlen($result['intro']);
$result['intro'] .= substr($result['intro'], $len - 1, $len) == '.' ? ".." : "...";
}
if (!is_null($result['intro'])) {
$results[] = array('article' => $result);
}
}
$searchResults['results'] = $results;
$json = json_encode($searchResults);
$wgMemc->set($key, $json, 3600);
// 1 hour
header("Content-type: application/json");
$wgOut->disable(true);
echo $json;
}
function formatRelated($t)
{
global $wgUser, $wgParser;
$result = '';
if ($t && $t->exists()) {
$r = Revision::newFromTitle($t);
$intro = $wgParser->getSection($r->getText(), 0);
$intro = Wikitext::flatten($intro);
if (strlen($intro) > 250) {
$intro = substr($intro, 0, 250) . '...';
}
$sk = $wgUser->getSkin();
$img = SkinWikihowskin::getGalleryImage($t, 238, 139);
$result .= "<div class='slide_related'>\n\t\t\t\t\t\t<a href='{$t->getFullURL()}'><img src='{$img}' alt='' width='238' height='139' class='gs_img' /></a>\n\t\t\t\t\t\t<h3><a href='{$t->getFullURL()}'>{$t->getText()}</a></h3>\n\t\t\t\t\t\t<p>{$intro}</p>\n\t\t\t\t\t\t</div>";
}
return $result;
}
public function parseArticle_02($article)
{
global $wgWikiHowSections, $wgTitle, $wgUser;
$ads = $wgUser->getID() == 0;
$sk = new SkinWikihowskin();
$sectionMap = array(wfMsg('Intro') => 'intro', wfMsg('Ingredients') => 'ingredients', wfMsg('Steps') => 'steps', wfMsg('Video') => 'video', wfMsg('Tips') => 'tips', wfMsg('Warnings') => 'warnings', wfMsg('relatedwikihows') => 'relatedwikihows', wfMsg('sourcescitations') => 'sources', wfMsg('thingsyoullneed') => 'thingsyoullneed');
foreach ($wgWikiHowSections as $section) {
$reverse_msgs[wfMsg($section)] = $section;
}
$parts = preg_split("@(<h2.*</h2>)@im", $article, 0, PREG_SPLIT_DELIM_CAPTURE);
$body = '';
$intro_img = '';
for ($i = 0; $i < sizeof($parts); $i++) {
if ($i == 0) {
//intro
preg_match("/Image:(.*)\">/", $parts[$i], $matches);
if (count($matches) > 0) {
$img = $matches[1];
$img = preg_replace('@%27@', "'", $img);
$image = Title::makeTitle(NS_IMAGE, $img);
if ($image) {
$file = wfFindFile($image);
if ($file) {
$thumb = $file->getThumbnail(200, -1, true, true);
$intro_img = '<a href="' . $image->getFullUrl() . '"><img border="0" width="200" class="mwimage101" src="' . wfGetPad($thumb->url) . '" alt="" /></a>';
}
}
}
if ($intro_img == '') {
$intro_img = '<img border="0" width="200" class="mwimage101" src="' . wfGetPad('/skins/WikiHow/images/wikihow_sq_200.png') . '" alt="" />';
}
$r = Revision::newFromTitle($wgTitle);
$intro_text = Wikitext::getIntro($r->getText());
$intro_text = trim(Wikitext::flatten($intro_text));
$body .= '<br /><div id="color_div"></div><br />';
$body .= '<div id="article_intro">' . $intro_text . '</div>';
if ($ads) {
$body .= '<div class="ad_noimage intro_ad">' . wikihowAds::getAdUnitPlaceholder('intro') . '</div>';
}
} else {
if (stripos($parts[$i], "<h2") === 0 && $i < sizeof($parts) - 1) {
preg_match("@<span>.*</span>@", $parts[$i], $matches);
$rev = "";
if (sizeof($matches) > 0) {
$h2 = trim(strip_tags($matches[0]));
$rev = isset($reverse_msgs[$h2]) ? $reverse_msgs[$h2] : "";
}
if ($rev !== 'steps') {
$body .= $parts[$i];
}
$i++;
if ($rev == "steps") {
$body .= "\n<div id=\"steps\" class='editable'>{$parts[$i]}</div>\n";
} else {
if ($rev != "") {
$body .= "\n<div id=\"{$rev}\" class='article_inner editable'>{$parts[$i]}</div>\n";
} else {
$body .= "\n<div class='article_inner editable'>{$parts[$i]}</div>\n";
}
}
} else {
$body .= $parts[$i];
}
}
}
$punct = "!\\.\\?\\:";
# valid ways of ending a sentence for bolding
$i = strpos($body, '<div id="steps"');
if ($i !== false) {
$j = strpos($body, '<div id=', $i + 5);
}
if ($j === false) {
$j = strlen($body);
}
if ($j !== false && $i !== false) {
$steps = substr($body, $i, $j - $i);
$parts = preg_split("@(<[/]?ul>|<[/]?ol>|<[/]?li>)@im", $steps, 0, PREG_SPLIT_DELIM_CAPTURE | PREG_SPLIT_NO_EMPTY);
$numsteps = preg_match_all('/<li>/m', $steps, $matches);
$level = 0;
$steps = "";
$upper_tag = "";
$levelstack = array();
$tagstack = array();
$current_tag = "";
$current_li = 1;
$donefirst = false;
// used for ads to tell when we've put the ad after the first step
if ($numsteps < 100) {
while ($p = array_shift($parts)) {
switch (strtolower($p)) {
case "<ol>":
$level++;
if ($level == 1) {
$p = '<ol class="steps_list">';
$upper_tag = "ol";
} else {
$p = " <div class='listbody'>{$p}";
}
if ($current_tag != "") {
$tagstack[] = $current_tag;
}
$current_tag = "ol";
$levelstack[] = $current_li;
$current_li = 1;
break;
case "<ul>":
if ($current_tag != "") {
$tagstack[] = $current_tag;
}
$current_tag = "ul";
$levelstack[] = $current_li;
$level++;
break;
case "</ol>":
$p .= '<div id="steps_end"></div>';
case "</ul>":
$level--;
if ($level == 0) {
$upper_tag = "";
}
$current_tag = array_pop($tagstack);
$current_li = array_pop($levelstack);
break;
case "<li>":
$closecount = 0;
if ($level == 1 && $upper_tag == "ol") {
$li_number = $current_li++;
//$p = '<li>'. str_pad($li_number,2,'0',STR_PAD_LEFT);
$p = '<li>';
# this is where things get interesting. Want to make first sentence bold!
# but we need to handle cases where there are tags in the first sentence
# split based on HTML tags
$next = array_shift($parts);
$htmlparts = preg_split("@(<[^>]*>)@im", $next, 0, PREG_SPLIT_DELIM_CAPTURE | PREG_SPLIT_NO_EMPTY);
$dummy = 0;
$incaption = false;
$apply_b = false;
while ($x = array_shift($htmlparts)) {
# if it's a tag, just append it and keep going
if (preg_match("@(<[^>]*>)@im", $x)) {
//tag
$p .= $x;
if ($x == "<span class='caption'>") {
$incaption = true;
} else {
if ($x == "</span>" && $incaption) {
$incaption = false;
}
}
continue;
}
# put the closing </b> in if we hit the end of the sentence
if (!$incaption) {
if (!$apply_b && trim($x) != "") {
$p .= '<p class="step_head"><span>' . str_pad($li_number, 2, '0', STR_PAD_LEFT) . '</span>';
$apply_b = true;
}
if ($apply_b) {
$x = preg_replace("@([{$punct}])@im", "\$1</p>", $x, 1, &$closecount);
}
}
$p .= $x;
if ($closecount > 0) {
break;
} else {
#echo "\n\n-----$x----\n\n";
}
$dummy++;
}
# get anything left over
$p .= implode("", $htmlparts);
if ($closecount == 0) {
$p .= "</b>";
}
// close the bold tag if we didn't already
if ($level == 1 && $current_li == 2 && $ads && !$donefirst) {
$p .= '<br class="clearall" />' . wikihowAds::getAdUnitPlaceholder(0);
$donefirst = true;
}
}
break;
case "</li>":
$p = "<div class='clearall'></div>{$p}";
//changed BR to DIV b/c IE doesn't work with the BR clear tag
break;
}
// switch
$steps .= $p;
}
// while
} else {
$steps = substr($body, $i, $j - $i);
$steps = "<div id='steps_notmunged'>\n" . $steps . "\n</div>\n";
}
// we have to put the final_li in the last OL LI step, so reverse the walk of the tokens
$parts = preg_split("@(<[/]?ul>|<[/]?ol>|<[/]?li>)@im", $steps, 0, PREG_SPLIT_DELIM_CAPTURE);
$parts = array_reverse($parts);
$steps = "";
$level = 0;
$gotit = false;
$donelast = false;
foreach ($parts as $p) {
$lp = strtolower($p);
if ($lp == "</ol>") {
$level++;
$gotit = false;
} else {
if ($lp == "</ul>") {
$level++;
} else {
if (strpos($lp, "<li") !== false && $level == 1 && !$gotit) {
/// last OL step list f****r
//$p = preg_replace("@<li[^>]*>@i", '<li class="steps_li final_li">', $p);
$gotit = true;
} else {
if (strpos($lp, "<ul") !== false) {
$level--;
} else {
if (strpos($lp, "<ol") !== false) {
$level--;
} else {
if ($lp == "</li>" && !$donelast) {
// ads after the last step
if ($ads) {
if (substr($body, $j) == "") {
$p = "<script>missing_last_ads = true;</script>" . wikihowAds::getAdUnitPlaceholder(1) . $p;
$no_third_ad = true;
} else {
$p = '<br />' . wikihowAds::getAdUnitPlaceholder(2) . $p;
}
}
$donelast = true;
}
}
}
}
}
}
$steps = $p . $steps;
}
$body = substr($body, 0, $i) . $steps . substr($body, $j);
}
/// if numsteps == 100?
/// ads below tips, walk the sections and put them after the tips
if ($ads) {
$foundtips = false;
$anchorTag = "";
foreach ($wgWikiHowSections as $s) {
$isAtEnd = false;
if ($s == "ingredients" || $s == "steps") {
continue;
}
// we skip these two top sections
$i = strpos($body, '<div id="' . $s . '"');
if ($i !== false) {
$j = strpos($body, '<h2>', $i + strlen($s));
} else {
continue;
// we didnt' find this section
}
if ($j === false) {
$j = strlen($body);
// go to the end
$isAtEnd = true;
}
if ($j !== false && $i !== false) {
$section = substr($body, $i, $j - $i);
if ($s == "video") {
// special case for video
$newsection = "<div id='video'><center>{$section}</center></div>";
$body = str_replace($section, $newsection, $body);
continue;
} else {
if ($s == "tips") {
//tip ad is now at the bottom of the tips section
//need to account for the possibility of no sections below this and therefor
//no anchor tag
if ($isAtEnd) {
$anchorTag = "<p></p>";
}
$body = str_replace($section, $section . $anchorTag . wikihowAds::getAdUnitPlaceholder('2a') . '<p><br /></p>', $body);
$foundtips = true;
break;
} else {
$foundtips = true;
if ($isAtEnd) {
$anchorTag = "<p></p>";
}
$body = str_replace($section, $section . $anchorTag . wikihowAds::getAdUnitPlaceholder(2), $body);
break;
}
}
}
}
if (!$foundtips && !$no_third_ad) {
//must be the video section
//need to put in the empty <p> tag since all the other sections have them for the anchor tags.
$body .= "<p class='video_spacing'></p>" . wikihowAds::getAdUnitPlaceholder(2);
}
}
$catlinks = $sk->getCategoryLinks($false);
$authors = $sk->getAuthorFooter();
if ($authors != "" || is_array($this->data['language_urls']) || $catlinks != "") {
//k, now grab the bottom stuff
$article_bottom .= '<br />' . wfGetSuggestedTitles($wgTitle) . '<br />
<h2 class="section_head" id="article_info_header">' . wfMsg('article_info') . '</h2>
<div id="article_info" class="article_inner">
<p>' . self::getLastEdited() . '</p>
<p>' . wfMsg('categories') . ':<br/>' . $catlinks . '</p>
<p>' . $authors . '</p>
</div><!--end article_info-->';
}
if ($wgUser->getID() == 0 && !$isMainPage && $action != 'edit' && $wgTitle->getNamespace() == NS_MAIN) {
$article_bottom .= '<div class="final_ad">' . wikihowAds::getAdUnitPlaceholder(7) . '</div>';
}
$article_bottom .= '
<div id="final_question">
' . $userstats . '
<p><b>' . $sk->pageStats() . '</b></p>
<div id="page_rating">' . RateArticle::showForm() . '</div>
<p></p>
</div> <!--end last_question-->
</div> <!-- article -->';
//share buttons
$url = urlencode($wgServer . "/" . $wgTitle->getPrefixedURL());
$fb = '<div class="like_button"><fb:like href="' . $url . '" send="false" layout="button_count" width="86" show_faces="false"></fb:like></div>';
$gp1 = '<div class="gplus1_button"><g:plusone size="medium" callback="plusone_vote"></g:plusone></div>';
// $fb_share = '<div class="like_button like_tools"><fb:like href="' . $url . '" send="false" layout="button_count" width="86" show_faces="false"></fb:like></div>';
$tb_admin = '<div class="admin_state"><a href="http://twitter.com/share" data-lang="' . $wgLanguageCode . '" style="display:none; background-image: none; color: #ffffff;" class="twitter-share-button" data-count="horizontal" data-via="wikiHow" data-text="How to ' . htmlspecialchars($wgTitle->getText()) . '" data-related="JackHerrick:Founder of wikiHow">Tweet</a></div>';
$tb = '<a href="http://twitter.com/share" data-lang="' . $wgLanguageCode . '" style="display:none; background-image: none; color: #ffffff;" class="twitter-share-button" data-count="horizontal" data-via="wikiHow" data-text="How to ' . htmlspecialchars($wgTitle->getText()) . '" data-related="JackHerrick:Founder of wikiHow">Tweet</a>';
$the_buttons = '<div id="share_buttons_top">' . $fb;
if ($wgUser->isSysop() && $wgTitle->userCan('delete')) {
$the_buttons .= $tb_admin;
} else {
$the_buttons .= $tb;
}
$the_buttons .= $gp1 . '</div>';
$title = '<h1>How to ' . $wgTitle->getText() . '</h1>';
$edited = $sk->getAuthorHeader();
$sidebar = '<div id="sidenav"><div id="showslideshow"></div><div id="pp_big_space">' . $intro_img . '</div></div>';
$main = '<div id="article_main">' . $title . $the_buttons . $edited . $body . $article_bottom . '</div>';
$article = '<div id="article_layout_' . self::ARTICLE_LAYOUT . '">' . $sidebar . $main . '</div>';
return $article;
}
/**
* Sets the meta description in the database to be part of the intro, part
* of the first step, or 'original' which is something like "wikiHow
* article on How to <title>".
*/
private function buildDescription($style)
{
if (self::DESC_STYLE_ORIGINAL == $style) {
return array(true, '');
}
if (self::DESC_STYLE_EDITED == $style) {
return array(true, $this->row['ami_desc']);
}
$wikitext = $this->getArticleWikiText();
if (!$wikitext) {
return array(false, '');
}
if (self::DESC_STYLE_INTRO == $style || self::DESC_STYLE_INTRO_NO_TITLE == $style) {
// grab intro
$desc = Wikitext::getIntro($wikitext);
// append first step to intro if intro maybe isn't long enough
if (strlen($desc) < 2 * self::MAX_DESC_LENGTH) {
list($steps, ) = Wikitext::getStepsSection($wikitext);
if ($steps) {
$desc .= ' ' . Wikitext::cutFirstStep($steps);
}
}
} elseif (self::DESC_STYLE_STEP1 == $style) {
// grab steps section
list($desc, ) = Wikitext::getStepsSection($wikitext);
// pull out just the first step
if ($desc) {
$desc = Wikitext::cutFirstStep($desc);
} else {
$desc = Wikitext::getIntro($wikitext);
}
} else {
//throw new Exception('ArticleMetaInfo: unknown style');
return array(false, '');
}
$desc = Wikitext::flatten($desc);
$howto = wfMsg('howto', $this->titleText);
if ($desc) {
if (self::DESC_STYLE_INTRO_NO_TITLE != $style) {
$desc = $howto . '. ' . $desc;
}
} else {
$desc = $howto;
}
$desc = self::trimDescription($desc);
return array(true, $desc);
}
