function synthesizeSummary($wikitext, $maxSteps, $fullURL)
{
$stepsSec = Wikitext::getStepsSection($wikitext, true);
if (!$stepsSec) {
return '';
}
$stepsText = Wikitext::stripHeader($stepsSec[0]);
if (Wikitext::countAltMethods($stepsText) > 0) {
$altMethods = Wikitext::splitAltMethods($stepsText);
foreach ($altMethods as $method) {
if (Wikitext::isAltMethod($method) && Wikitext::countSteps($method) > 0) {
$stepsText = $method;
break;
}
}
}
$countSteps = Wikitext::countSteps($stepsText);
$summaryOut = '';
$steps = Wikitext::splitSteps($stepsText);
$count = 0;
foreach ($steps as $step) {
if (Wikitext::isStepSimple($step, false)) {
$summary = Wikitext::summarizeStep($step);
$summary = Wikitext::removeRefsFromFlattened($summary);
if ($summary) {
$count++;
$break = $count > 1 ? "<br>" : '';
if ($count > $maxSteps) {
$remaining = $countSteps - $maxSteps;
$text = '';
if ($remaining >= 2) {
$text = "{$remaining} more steps at wikiHow";
} elseif ($remaining == 1) {
$text = "Another step at wikiHow";
}
if ($text) {
$href = htmlspecialchars($fullURL);
$link = "<a href='{$href}'>{$text}</a>";
$summaryOut .= "{$break}{$link}";
}
break;
} else {
$summaryOut .= "{$break}{$count}. {$summary}";
}
}
}
}
return $summaryOut;
}
/**
* process a single article for the subheaders
*/
private static function processSubheaders($title)
{
list($wikitext, $stepsText, $sectionID) = self::getWikitext($title);
if ($wikitext && $stepsText) {
$subs = Wikitext::countAltMethods($stepsText);
if ($subs > 0) {
//we have some parts/methods/ways!
$count = preg_match_all('@(^\\s*===\\s*)(Method|Part)(\\s*One|\\s*1)(\\s*===\\s*)@im', $stepsText, $m);
if ($count) {
$url = 'http://www.wikihow.com/' . $title->getDBKey();
print $url . "\n";
return true;
}
}
}
return false;
}
/**
* process a single article for the subheaders
*/
private static function processSubheaders($title, $csv)
{
list($wikitext, $stepsText, $sectionID) = self::getWikitext($title);
if ($wikitext && $stepsText) {
$subs = Wikitext::countAltMethods($stepsText);
if ($subs > 1) {
//we have some parts/methods/ways!
$newstepsText = preg_replace('@(^\\s*===\\s*)(Method |Part )(.*?:|.*?\\.|\\s*===|)@im', '$1', $stepsText);
//have we made a difference?
if (strcmp($stepsText, $newstepsText) != 0) {
//it has changed; update the sub headers
$newWikitext = Wikitext::replaceStepsSection($wikitext, $sectionID, $newstepsText, true);
if ($newWikitext) {
//sub headers have been updated, add the "magic word"
list($newestWikitext, $magic_word) = self::addMagicWord($stepsText, $newWikitext);
if ($newestWikitext) {
$data = array();
Wikitext::saveWikitext($title, $newestWikitext, self::$comment);
$url = 'http://www.wikihow.com/' . $title->getDBKey();
$data[] = $url;
$data[] = $magic_word;
//rock! now a little logic to grab what changed...
preg_match('@^===.*===?@im', $stepsText, $m);
$data[] = $m[0];
preg_match('@^===.*===?@im', $newstepsText, $m);
$data[] = $m[0];
//show it
print $url . ' ' . $magic_word . "\n";
//log it
self::logIt($url . ' ' . $magic_word);
//write it to the csv
fputcsv($csv, $data, chr(9));
//good night, sweet prince...
usleep(self::SLEEPTIME);
return true;
}
}
}
}
}
return false;
}
private static function genTitle($title, $test, $custom)
{
$titleTxt = $title->getText();
$howto = wfMsg('howto', $titleTxt);
list($wikitext, $stepsText) = self::getWikitext($title);
switch ($test) {
case self::TITLE_CUSTOM:
// Custom
$title = $custom;
break;
case self::TITLE_SITE_PREVIOUS:
// How to XXX: N Steps (with Pictures) - wikiHow
list($numSteps, $withPictures) = self::getTitleExtraInfo($wikitext, $stepsText, $test);
$inner = $numSteps > 0 ? "{$howto}: {$numSteps} Steps" : $howto;
$inner = $withPictures ? "{$inner} (with Pictures)" : $inner;
$title = wfMsg('pagetitle', $inner);
break;
default:
// How to XXX: N Steps (with Pictures) - wikiHow
// From Chris's Mar 25 email
// How to XXX: N Steps (with Pictures) - wikiHow
// From Chris's Mar 25 email
case 5:
// default, but not "with Pictures"
// default, but not "with Pictures"
case 6:
// n Tips on How to ... "with Pictures"
// n Tips on How to ... "with Pictures"
case 7:
// n Tips on How to ... but not "with Pictures"
// n Tips on How to ... but not "with Pictures"
case 8:
// How to ...: Step-by-Step Instructions "with Pictures"
// How to ...: Step-by-Step Instructions "with Pictures"
case 9:
// How to ...: Step-by-Step Instructions but not "with Pictures"
$methods = Wikitext::countAltMethods($stepsText);
if ($methods >= 3) {
$inner = "{$methods} Ways to {$titleTxt}";
$title = wfMsg('pagetitle', $inner);
if (strlen($title) > self::MAX_TITLE_LENGTH) {
$title = $inner;
}
} else {
list($numSteps, $withPictures) = self::getTitleExtraInfo($wikitext, $stepsText, $test);
$forceNoWithPictures = in_array($test, array(5, 7, 9));
$withPictures = !$forceNoWithPictures ? $withPictures : false;
if ($test == 6 || $test == 7) {
$inner = $numSteps > 0 ? "{$numSteps} Tips on {$howto}" : $howto;
} elseif ($test == 8 || $test == 9) {
$inner = $numSteps > 0 ? "{$howto}: Step-by-Step Instructions" : $howto;
} else {
$inner = $numSteps > 0 ? "{$howto}: {$numSteps} Steps" : $howto;
}
$inner = $withPictures ? "{$inner} (with Pictures)" : $inner;
$title = wfMsg('pagetitle', $inner);
// first, try articlename + metadata + wikihow
if (strlen($title) > self::MAX_TITLE_LENGTH) {
// next, try articlename + metadata
$title = $inner;
if ($numSteps > 0 && strlen($title) > self::MAX_TITLE_LENGTH) {
// next, try articlename + steps
if ($test == 6 || $test == 7) {
$inner = "{$numSteps} Tips on {$howto}";
} elseif ($test == 8 || $test == 9) {
$inner = "{$howto}: Step-by-Step Instructions";
} else {
$title = "{$howto}: {$numSteps} Steps";
}
}
if (strlen($title) > self::MAX_TITLE_LENGTH) {
// next, try articlename + wikihow
$title = wfMsg('pagetitle', $howto);
if (strlen($title) > self::MAX_TITLE_LENGTH) {
// lastly, set title just as articlename
$title = $howto;
}
}
}
}
break;
// start of new Title Tests from Chris's March 29 email
//case 12: // How to XXX: N Tips - wikiHow
//case 13: // N Tips on How to XXX - wikiHow
//case 14: // How to XXX: Step-by-Step Instructions
//case 15: // How to XXX: N Methods - wikiHow
//case 16: // N Ways to XXX - wikiHow
//case 17: // How to XXX with Step-by-Step Pictures
// start of new title tests from Chris's Oct 2 email
/*case 18: // How to XXX with Step-by-Step Pictures
$inner = '';
$methods = Wikitext::countAltMethods($stepsText);
if ($methods >= 4) {
$inner = "$methods Ways to $titleTxt";
} else {
$steps = Wikitext::countSteps($stepsText);
if (3 <= $steps && $steps < 15) {
$inner = "$steps Tips on $howto";
}
}
if (!$inner) {
$inner = "$howto: Step-by-Step Instructions";
}
$title = wfMsg('pagetitle', $inner);
if (strlen($title) > self::MAX_TITLE_LENGTH) {
$title = $inner;
}
break;*/
}
return $title;
}
