rss-bridge/lib/FeedExpander.php
2023-09-21 22:05:55 +02:00

507 lines
16 KiB
PHP

<?php
/**
* This file is part of RSS-Bridge, a PHP project capable of generating RSS and
* Atom feeds for websites that don't have one.
*
* For the full license information, please view the UNLICENSE file distributed
* with this source code.
*
* @package Core
* @license http://unlicense.org/ UNLICENSE
* @link https://github.com/rss-bridge/rss-bridge
*/
/**
* An abstract class for bridges that need to transform existing RSS or Atom
* feeds.
*
* This class extends {@see BridgeAbstract} with functions to extract contents
* from existing RSS or Atom feeds. Bridges that need to transform existing feeds
* should inherit from this class instead of {@see BridgeAbstract}.
*
* Bridges that extend this class don't need to concern themselves with getting
* contents from existing feeds, but can focus on adding additional contents
* (i.e. by downloading additional data), filtering or just transforming a feed
* into another format.
*
* @link http://www.rssboard.org/rss-0-9-1 RSS 0.91 Specification
* @link http://web.resource.org/rss/1.0/spec RDF Site Summary (RSS) 1.0
* @link http://www.rssboard.org/rss-specification RSS 2.0 Specification
* @link https://tools.ietf.org/html/rfc4287 The Atom Syndication Format
*
* @todo The parsing functions should all be private. This class is complicated
* enough without having to consider children overriding functions.
*/
abstract class FeedExpander extends BridgeAbstract
{
/** Indicates an RSS 1.0 feed */
const FEED_TYPE_RSS_1_0 = 'RSS_1_0';
/** Indicates an RSS 2.0 feed */
const FEED_TYPE_RSS_2_0 = 'RSS_2_0';
/** Indicates an Atom 1.0 feed */
const FEED_TYPE_ATOM_1_0 = 'ATOM_1_0';
/**
* Holds the title of the current feed
*
* @var string
*/
private $title;
/**
* Holds the URI of the feed
*
* @var string
*/
private $uri;
/**
* Holds the icon of the feed
*
*/
private $icon;
/**
* Holds the feed type during internal operations.
*
* @var string
*/
private $feedType;
/**
* Collects data from an existing feed.
*
* Children should call this function in {@see BridgeAbstract::collectData()}
* to extract a feed.
*
* @param string $url URL to the feed.
* @param int $maxItems Maximum number of items to collect from the feed
* (`-1`: no limit).
* @return self
*/
public function collectExpandableDatas($url, $maxItems = -1)
{
if (empty($url)) {
throw new \Exception('There is no $url for this RSS expander');
}
Debug::log(sprintf('Loading from %s', $url));
/* Notice we do not use cache here on purpose:
* we want a fresh view of the RSS stream each time
*/
$mimeTypes = [
MrssFormat::MIME_TYPE,
AtomFormat::MIME_TYPE,
'*/*',
];
$httpHeaders = ['Accept: ' . implode(', ', $mimeTypes)];
$xml = getContents($url, $httpHeaders);
if ($xml === '') {
throw new \Exception(sprintf('Unable to parse xml from `%s` because we got the empty string', $url), 10);
}
// Maybe move this call earlier up the stack frames
// Disable triggering of the php error-handler and handle errors manually instead
libxml_use_internal_errors(true);
// Consider replacing libxml with https://www.php.net/domdocument
// Intentionally not using the silencing operator (@) because it has no effect here
$rssContent = simplexml_load_string(trim($xml));
if ($rssContent === false) {
$xmlErrors = libxml_get_errors();
foreach ($xmlErrors as $xmlError) {
Debug::log(trim($xmlError->message));
}
if ($xmlErrors) {
// Render only the first error into exception message
$firstXmlErrorMessage = $xmlErrors[0]->message;
}
throw new \Exception(sprintf('Unable to parse xml from `%s` %s', $url, $firstXmlErrorMessage ?? ''), 11);
}
// Restore previous behaviour in case other code relies on it being off
libxml_use_internal_errors(false);
// Commented out because it's spammy
// Debug::log(sprintf("RSS content is ===========\n%s===========", var_export($rssContent, true)));
switch (true) {
case isset($rssContent->item[0]):
Debug::log('Detected RSS 1.0 format');
$this->feedType = self::FEED_TYPE_RSS_1_0;
$this->collectRss1($rssContent, $maxItems);
break;
case isset($rssContent->channel[0]):
Debug::log('Detected RSS 0.9x or 2.0 format');
$this->feedType = self::FEED_TYPE_RSS_2_0;
$this->collectRss2($rssContent, $maxItems);
break;
case isset($rssContent->entry[0]):
Debug::log('Detected ATOM format');
$this->feedType = self::FEED_TYPE_ATOM_1_0;
$this->collectAtom1($rssContent, $maxItems);
break;
default:
Debug::log(sprintf('Unable to detect feed format from `%s`', $url));
throw new \Exception(sprintf('Unable to detect feed format from `%s`', $url));
}
return $this;
}
/**
* Collect data from an RSS 1.0 compatible feed
*
* @link http://web.resource.org/rss/1.0/spec RDF Site Summary (RSS) 1.0
*
* @param string $rssContent The RSS content
* @param int $maxItems Maximum number of items to collect from the feed
* (`-1`: no limit).
* @return void
*
* @todo Instead of passing $maxItems to all functions, just add all items
* and remove excessive items later.
*/
protected function collectRss1($rssContent, $maxItems)
{
$this->loadRss2Data($rssContent->channel[0]);
foreach ($rssContent->item as $item) {
$tmp_item = $this->parseItem($item);
if (!empty($tmp_item)) {
$this->items[] = $tmp_item;
}
if ($maxItems !== -1 && count($this->items) >= $maxItems) {
break;
}
}
}
/**
* Collect data from a RSS 2.0 compatible feed
*
* @link http://www.rssboard.org/rss-specification RSS 2.0 Specification
*
* @param int $maxItems Maximum number of items to collect from the feed (`-1`: no limit).
* @return void
*
* @todo Instead of passing $maxItems to all functions, just add all items and remove excessive items later.
*/
protected function collectRss2(\SimpleXMLElement $rssContent, $maxItems)
{
$rssContent = $rssContent->channel[0];
$this->loadRss2Data($rssContent);
foreach ($rssContent->item as $item) {
$tmp_item = $this->parseItem($item);
if (!empty($tmp_item)) {
$this->items[] = $tmp_item;
}
if ($maxItems !== -1 && count($this->items) >= $maxItems) {
break;
}
}
}
/**
* Collect data from a Atom 1.0 compatible feed
*
* @link https://tools.ietf.org/html/rfc4287 The Atom Syndication Format
*
* @param object $content The Atom content
* @param int $maxItems Maximum number of items to collect from the feed
* (`-1`: no limit).
* @return void
*
* @todo Instead of passing $maxItems to all functions, just add all items
* and remove excessive items later.
*/
protected function collectAtom1($content, $maxItems)
{
$this->loadAtomData($content);
foreach ($content->entry as $item) {
$tmp_item = $this->parseItem($item);
if (!empty($tmp_item)) {
$this->items[] = $tmp_item;
}
if ($maxItems !== -1 && count($this->items) >= $maxItems) {
break;
}
}
}
/**
* Load RSS 2.0 feed data into RSS-Bridge
*
* @param object $rssContent The RSS content
* @return void
*
* @todo set title, link, description, language, and so on
*/
protected function loadRss2Data($rssContent)
{
$this->title = trim((string)$rssContent->title);
$this->uri = trim((string)$rssContent->link);
if (!empty($rssContent->image)) {
$this->icon = trim((string)$rssContent->image->url);
}
}
/**
* Load Atom feed data into RSS-Bridge
*
* @param object $content The Atom content
* @return void
*/
protected function loadAtomData($content)
{
$this->title = (string)$content->title;
// Find best link (only one, or first of 'alternate')
if (!isset($content->link)) {
$this->uri = '';
} elseif (count($content->link) === 1) {
$this->uri = (string)$content->link[0]['href'];
} else {
$this->uri = '';
foreach ($content->link as $link) {
if (strtolower($link['rel']) === 'alternate') {
$this->uri = (string)$link['href'];
break;
}
}
}
if (!empty($content->icon)) {
$this->icon = (string)$content->icon;
} elseif (!empty($content->logo)) {
$this->icon = (string)$content->logo;
}
}
/**
* Parse the contents of a single Atom feed item into a RSS-Bridge item for
* further transformation.
*
* @param object $feedItem A single feed item
* @return object The RSS-Bridge item
*
* @todo To reduce confusion, the RSS-Bridge item should maybe have a class
* of its own?
*/
protected function parseATOMItem($feedItem)
{
// Some ATOM entries also contain RSS 2.0 fields
$item = $this->parseRss2Item($feedItem);
if (isset($feedItem->id)) {
$item['uri'] = (string)$feedItem->id;
}
if (isset($feedItem->title)) {
$item['title'] = html_entity_decode((string)$feedItem->title);
}
if (isset($feedItem->updated)) {
$item['timestamp'] = strtotime((string)$feedItem->updated);
}
if (isset($feedItem->author)) {
$item['author'] = (string)$feedItem->author->name;
}
if (isset($feedItem->content)) {
$contentChildren = $feedItem->content->children();
if (count($contentChildren) > 0) {
$content = '';
foreach ($contentChildren as $contentChild) {
$content .= $contentChild->asXML();
}
$item['content'] = $content;
} else {
$item['content'] = (string)$feedItem->content;
}
}
//When "link" field is present, URL is more reliable than "id" field
if (count($feedItem->link) === 1) {
$item['uri'] = (string)$feedItem->link[0]['href'];
} else {
foreach ($feedItem->link as $link) {
if (strtolower($link['rel']) === 'alternate') {
$item['uri'] = (string)$link['href'];
}
if (strtolower($link['rel']) === 'enclosure') {
$item['enclosures'][] = (string)$link['href'];
}
}
}
return $item;
}
/**
* Parse the contents of a single RSS 0.91 feed item into a RSS-Bridge item
* for further transformation.
*
* @param object $feedItem A single feed item
* @return object The RSS-Bridge item
*
* @todo To reduce confusion, the RSS-Bridge item should maybe have a class
* of its own?
*/
protected function parseRss091Item($feedItem)
{
$item = [];
if (isset($feedItem->link)) {
$item['uri'] = (string)$feedItem->link;
}
if (isset($feedItem->title)) {
$item['title'] = html_entity_decode((string)$feedItem->title);
}
// rss 0.91 doesn't support timestamps
// rss 0.91 doesn't support authors
// rss 0.91 doesn't support enclosures
if (isset($feedItem->description)) {
$item['content'] = (string)$feedItem->description;
}
return $item;
}
/**
* Parse the contents of a single RSS 1.0 feed item into a RSS-Bridge item
* for further transformation.
*
* @param object $feedItem A single feed item
* @return object The RSS-Bridge item
*
* @todo To reduce confusion, the RSS-Bridge item should maybe have a class
* of its own?
*/
protected function parseRss1Item($feedItem)
{
// 1.0 adds optional elements around the 0.91 standard
$item = $this->parseRss091Item($feedItem);
$namespaces = $feedItem->getNamespaces(true);
if (isset($namespaces['dc'])) {
$dc = $feedItem->children($namespaces['dc']);
if (isset($dc->date)) {
$item['timestamp'] = strtotime((string)$dc->date);
}
if (isset($dc->creator)) {
$item['author'] = (string)$dc->creator;
}
}
return $item;
}
/**
* Parse the contents of a single RSS 2.0 feed item into a RSS-Bridge item
* for further transformation.
*
* @param object $feedItem A single feed item
* @return object The RSS-Bridge item
*
* @todo To reduce confusion, the RSS-Bridge item should maybe have a class
* of its own?
*/
protected function parseRss2Item($feedItem)
{
// Primary data is compatible to 0.91 with some additional data
$item = $this->parseRss091Item($feedItem);
$namespaces = $feedItem->getNamespaces(true);
if (isset($namespaces['dc'])) {
$dc = $feedItem->children($namespaces['dc']);
}
if (isset($namespaces['media'])) {
$media = $feedItem->children($namespaces['media']);
}
if (isset($feedItem->guid)) {
foreach ($feedItem->guid->attributes() as $attribute => $value) {
if (
$attribute === 'isPermaLink'
&& (
$value === 'true' || (
filter_var($feedItem->guid, FILTER_VALIDATE_URL)
&& (empty($item['uri']) || !filter_var($item['uri'], FILTER_VALIDATE_URL))
)
)
) {
$item['uri'] = (string)$feedItem->guid;
break;
}
}
}
if (isset($feedItem->pubDate)) {
$item['timestamp'] = strtotime((string)$feedItem->pubDate);
} elseif (isset($dc->date)) {
$item['timestamp'] = strtotime((string)$dc->date);
}
if (isset($feedItem->author)) {
$item['author'] = (string)$feedItem->author;
} elseif (isset($feedItem->creator)) {
$item['author'] = (string)$feedItem->creator;
} elseif (isset($dc->creator)) {
$item['author'] = (string)$dc->creator;
} elseif (isset($media->credit)) {
$item['author'] = (string)$media->credit;
}
if (isset($feedItem->enclosure) && !empty($feedItem->enclosure['url'])) {
$item['enclosures'] = [(string)$feedItem->enclosure['url']];
}
return $item;
}
/**
* Parse the contents of a single feed item, depending on the current feed
* type, into a RSS-Bridge item.
*
* @param object $item The current feed item
* @return object A RSS-Bridge item, with (hopefully) the whole content
*/
protected function parseItem($item)
{
switch ($this->feedType) {
case self::FEED_TYPE_RSS_1_0:
return $this->parseRss1Item($item);
case self::FEED_TYPE_RSS_2_0:
return $this->parseRss2Item($item);
case self::FEED_TYPE_ATOM_1_0:
return $this->parseATOMItem($item);
default:
throw new \Exception(sprintf('Unknown version %s!', $this->getInput('version')));
}
}
/** {@inheritdoc} */
public function getURI()
{
if (!empty($this->uri)) {
return $this->uri;
}
return parent::getURI();
}
/** {@inheritdoc} */
public function getName()
{
if (!empty($this->title)) {
return $this->title;
}
return parent::getName();
}
/** {@inheritdoc} */
public function getIcon()
{
if (!empty($this->icon)) {
return $this->icon;
}
return parent::getIcon();
}
}