Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[GBAtemp] Fix content extraction (#2313) #2314

Merged
merged 4 commits into from
Jan 29, 2022
Merged

[GBAtemp] Fix content extraction (#2313) #2314

Changes from 1 commit
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
b4cbbdf
[GBAtemp] Fix content extraction (#2313)
ORelio Oct 29, 2021
9ecc481
[GBAtemp] Switch to built-in rss-bridge functions
ORelio Oct 31, 2021
6027e30
Merge branch 'master' into patch-4
ORelio Jan 27, 2022
3158fad
Fix missing comma after merge
ORelio Jan 27, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
94 changes: 54 additions & 40 deletions bridges/GBAtempBridge.php
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -32,45 +32,62 @@ private function buildItem($uri, $title, $author, $timestamp, $thumbnail, $conte
return $item;
}

private function strEndsWith($haystack, $needle){
// str_ends_with is not available below PHP 8
$length = strlen($needle);
return $length > 0 ? substr($haystack, -$length) === $needle : true;
}

private function decodeHtmlEntities($text) {
$text = html_entity_decode($text);
$convmap = array(0x0, 0x2FFFF, 0, 0xFFFF);
return trim(mb_decode_numericentity($text, $convmap, 'UTF-8'));
}

private function cleanupPostContent($content, $site_url){
$content = str_replace(':arrow:', '➤', $content);
$content = str_replace('href="attachments/', 'href="' . $site_url . 'attachments/', $content);
$content = str_replace('src="/', 'src="' . $site_url, $content);
$content = str_replace('href="/', 'href="' . $site_url, $content);
$content = stripWithDelimiters($content, '<script', '</script>');
return $content;
$content = stripWithDelimiters($content, '<svg', '</svg>');
$content = stripRecursiveHTMLSection($content, 'div', '<div class="reactionsBar');
return $this->decodeHtmlEntities($content);
}

private function makeAbsoluteUrl($link) {
if (strpos($link, '/') === 0) {
$link = substr($link, 1);
}
return self::URI . $link;
}

private function findItemDate($item){
$time = 0;
$dateField = $item->find('abbr.DateTime', 0);
$dateField = $item->find('time', 0);
if (is_object($dateField)) {
$time = intval(
extractFromDelimiters(
$dateField->outertext,
'data-time="',
'"'
)
);
} else {
$dateField = $item->find('span.DateTime', 0);
$time = DateTime::createFromFormat(
'M j, Y \a\t g:i A',
extractFromDelimiters(
$dateField->outertext,
'title="',
'"'
)
)->getTimestamp();
$time = strtotime($dateField->datetime);
}
return $time;
}

private function findItemImage($item, $selector){
$img = extractFromDelimiters($item->find($selector, 0)->style, 'url(', ')');
$paramPos = strpos($img, '?');
if ($paramPos !== false) {
$img = substr($img, 0, $paramPos);
}
if (!$this->strEndsWith($img, '.png') && !$this->strEndsWith($img, '.jpg')) {
$img = $img . '#.image';
}
return $this->makeAbsoluteUrl($img);
}

private function fetchPostContent($uri, $site_url){
$html = getSimpleHTMLDOMCached($uri);
if(!$html) {
return 'Could not request GBAtemp: ' . $uri;
}

$content = $html->find('div.messageContent, blockquote.baseHtml', 0)->innertext;
$content = $html->find('article.message-body', 0)->innertext;
return $this->cleanupPostContent($content, $site_url);
}

Expand All @@ -81,40 +98,37 @@ public function collectData(){

switch($this->getInput('type')) {
case 'N':
foreach($html->find('li[class=news_item news full]') as $newsItem) {
$url = self::URI . $newsItem->find('a', 0)->href;
$img = $this->getURI() . extractFromDelimiters($newsItem->find('a.news_image', 0)->style, 'url(', ')') . '#.image';
foreach($html->find('li.news_item.full') as $newsItem) {
$url = $this->makeAbsoluteUrl($newsItem->find('a', 0)->href);
$img = $this->findItemImage($newsItem, 'a.news_image');
$time = $this->findItemDate($newsItem);
$author = $newsItem->find('a.username', 0)->plaintext;
$title = $newsItem->find('a', 1)->plaintext;
$title = $this->decodeHtmlEntities($newsItem->find('h3.news_title', 0)->plaintext);
$content = $this->fetchPostContent($url, self::URI);
$this->items[] = $this->buildItem($url, $title, $author, $time, $img, $content);
unset($newsItem); // Some items are heavy, freeing the item proactively helps saving memory
}
break;
case 'R':
foreach($html->find('li.portal_review') as $reviewItem) {
$url = self::URI . $reviewItem->find('a', 0)->href;
$img = $this->getURI() . extractFromDelimiters($reviewItem->find('a', 0)->style, 'image:url(', ')');
$title = $reviewItem->find('span.review_title', 0)->plaintext;
$content = getSimpleHTMLDOM($url)
$url = $this->makeAbsoluteUrl($reviewItem->find('a.review_boxart', 0)->href);
$img = $this->findItemImage($reviewItem, 'a.review_boxart');
$title = $this->decodeHtmlEntities($reviewItem->find('h2.review_title', 0)->plaintext);
$content = getSimpleHTMLDOMCached($url)
or returnServerError('Could not request GBAtemp: ' . $uri);
$author = $content->find('a.username', 0)->plaintext;
$author = $content->find('span.author--name', 0)->plaintext;
$time = $this->findItemDate($content);
$intro = '<p><b>' . ($content->find('div#review_intro', 0)->plaintext) . '</b></p>';
$intro = '<p><b>' . ($content->find('div#review_introduction', 0)->plaintext) . '</b></p>';
$review = $content->find('div#review_main', 0)->innertext;
$subheader = '<p><b>' . $content->find('div.review_subheader', 0)->plaintext . '</b></p>';
$procons = $content->find('table.review_procons', 0)->outertext;
$scores = $content->find('table.reviewscores', 0)->outertext;
$content = $this->cleanupPostContent($intro . $review . $subheader . $procons . $scores, self::URI);
$content = $this->cleanupPostContent($intro . $review, self::URI);
$this->items[] = $this->buildItem($url, $title, $author, $time, $img, $content);
unset($reviewItem); // Free up memory
}
break;
case 'T':
foreach($html->find('li.portal-tutorial') as $tutorialItem) {
$url = self::URI . $tutorialItem->find('a', 1)->href;
$title = $tutorialItem->find('a', 1)->plaintext;
$url = $this->makeAbsoluteUrl($tutorialItem->find('a', 1)->href);
$title = $this->decodeHtmlEntities($tutorialItem->find('a', 1)->plaintext);
$time = $this->findItemDate($tutorialItem);
$author = $tutorialItem->find('a.username', 0)->plaintext;
$content = $this->fetchPostContent($url, self::URI);
Expand All @@ -124,8 +138,8 @@ public function collectData(){
break;
case 'F':
foreach($html->find('li.rc_item') as $postItem) {
$url = self::URI . $postItem->find('a', 1)->href;
$title = $postItem->find('a', 1)->plaintext;
$url = $this->makeAbsoluteUrl($postItem->find('a', 1)->href);
$title = $this->decodeHtmlEntities($postItem->find('a', 1)->plaintext);
$time = $this->findItemDate($postItem);
$author = $postItem->find('a.username', 0)->plaintext;
$content = $this->fetchPostContent($url, self::URI);
Expand Down