rss-bridge/bridges/FilterBridge.php
2024-04-06 18:07:45 +02:00

168 lines
5.4 KiB
PHP

<?php
class FilterBridge extends FeedExpander
{
const MAINTAINER = 'Frenzie, ORelio';
const NAME = 'Filter';
const CACHE_TIMEOUT = 3600; // 1h
const DESCRIPTION = 'Filters a feed of your choice';
const URI = 'https://github.com/RSS-Bridge/rss-bridge';
const PARAMETERS = [[
'url' => [
'name' => 'Feed URL',
'type' => 'text',
'exampleValue' => 'https://lorem-rss.herokuapp.com/feed?unit=day',
'required' => true,
],
'filter' => [
'name' => 'Filter (regular expression!!!)',
'required' => false,
],
'filter_type' => [
'name' => 'Filter type',
'type' => 'list',
'required' => false,
'values' => [
'Keep matching items' => 'permit',
'Hide matching items' => 'block',
],
'defaultValue' => 'permit',
],
'case_insensitive' => [
'name' => 'Case-insensitive filter',
'type' => 'checkbox',
'required' => false,
],
'fix_encoding' => [
'name' => 'Attempt Latin1/UTF-8 fixes when evaluating filter',
'type' => 'checkbox',
'required' => false,
],
'target_author' => [
'name' => 'Apply filter on author',
'type' => 'checkbox',
'required' => false,
],
'target_content' => [
'name' => 'Apply filter on content',
'type' => 'checkbox',
'required' => false,
],
'target_title' => [
'name' => 'Apply filter on title',
'type' => 'checkbox',
'required' => false,
'defaultValue' => 'checked'
],
'target_uri' => [
'name' => 'Apply filter on URI/URL',
'type' => 'checkbox',
'required' => false,
],
'title_from_content' => [
'name' => 'Generate title from content (overwrite existing title)',
'type' => 'checkbox',
'required' => false,
],
'length_limit' => [
'name' => 'Max length analyzed by filter (-1: no limit)',
'type' => 'number',
'required' => false,
'defaultValue' => -1,
],
]];
public function collectData()
{
$url = $this->getInput('url');
if (!Url::validate($url)) {
throw new \Exception('The url parameter must either refer to http or https protocol.');
}
$this->collectExpandableDatas($this->getURI());
}
protected function parseItem(array $item)
{
// Generate title from first 50 characters of content?
if ($this->getInput('title_from_content') && array_key_exists('content', $item)) {
$content = str_get_html($item['content']);
$plaintext = $content->plaintext;
if (mb_strlen($plaintext) < 51) {
$item['title'] = $plaintext;
} else {
$pos = strpos($item['content'], ' ', 50);
$item['title'] = substr($plaintext, 0, $pos);
if (strlen($plaintext) >= $pos) {
$item['title'] .= '...';
}
}
}
$filter = $this->getInput('filter');
if (! str_contains($filter, '#')) {
$delimiter = '#';
} elseif (! str_contains($filter, '/')) {
$delimiter = '/';
} else {
throw new \Exception('Cannot use both / and # inside filter');
}
$regex = $delimiter . $filter . $delimiter;
if ($this->getInput('case_insensitive')) {
$regex .= 'i';
}
// Retrieve fields to check
$filter_fields = [];
if ($this->getInput('target_author')) {
$filter_fields[] = $item['author'] ?? null;
}
if ($this->getInput('target_content')) {
$filter_fields[] = $item['content'] ?? null;
}
if ($this->getInput('target_title')) {
$filter_fields[] = $item['title'] ?? null;
}
if ($this->getInput('target_uri')) {
// todo: maybe consider 'http' and 'https' equivalent? Also maybe optionally .www subdomain?
$filter_fields[] = $item['uri'] ?? null;
}
// Apply filter on item
$keep_item = false;
$length_limit = intval($this->getInput('length_limit'));
foreach ($filter_fields as $field) {
if ($length_limit > 0) {
$field = substr($field, 0, $length_limit);
}
$result = preg_match($regex, $field);
if ($result === false) {
// todo: maybe notify user about the error here?
}
$keep_item |= boolval($result);
if ($this->getInput('fix_encoding')) {
$keep_item |= boolval(preg_match($regex, utf8_decode($field)));
$keep_item |= boolval(preg_match($regex, utf8_encode($field)));
}
}
// Reverse result? (keep everything but matching items)
if ($this->getInput('filter_type') === 'block') {
$keep_item = !$keep_item;
}
return $keep_item ? $item : null;
}
public function getURI()
{
$url = $this->getInput('url');
if (empty($url)) {
$url = parent::getURI();
}
return $url;
}
}