mexon/friendica
1
0
Fork 0
Code Pull requests Activity

Add new Network\Probe::ensureAbsoluteLinkFromHTMLDoc method

Browse source 
- Add supports for relative URL in href attributes to probe feed URL
This commit is contained in:
Hypolite Petovan 2020-05-21 00:24:23 -04:00
parent c2105f93d1
commit 3ef987e4e1
1 changed files with 58 additions and 16 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

74
src/Network/Probe.php
View file

@ -1771,30 +1771,72 @@ class Probe
$xpath = new DOMXPath($doc); $xpath = new DOMXPath($doc);
//$feeds = $xpath->query("/html/head/link[@type='application/rss+xml']"); $feedUrl = $xpath->evaluate('string(/html/head/link[@type="application/rss+xml" and @rel="alternate"]/@href)');
$feeds = $xpath->query("/html/head/link[@type='application/rss+xml' and @rel='alternate']");
if (!is_object($feeds)) { $feedUrl = $feedUrl ? self::ensureAbsoluteLinkFromHTMLDoc($feedUrl, $url, $xpath) : '';
return false;
return $feedUrl;
}
/**
* Return an absolute URL in the context of a HTML document retrieved from the provided URL.
*
* Loosely based on RFC 1808
*
* @see https://tools.ietf.org/html/rfc1808
*
* @param string $href The potential relative href found in the HTML document
* @param string $base The HTML document URL
* @param DOMXPath $xpath The HTML document XPath
* @return string
*/
private static function ensureAbsoluteLinkFromHTMLDoc(string $href, string $base, DOMXPath $xpath)
{
if (filter_var($href, FILTER_VALIDATE_URL, FILTER_FLAG_SCHEME_REQUIRED | FILTER_FLAG_HOST_REQUIRED)) {
return $href;
} }
if ($feeds->length == 0) { $base = $xpath->evaluate('string(/html/head/base/@href)') ?: $base;
return false;
}
$feed_url = ""; $baseParts = parse_url($base);
foreach ($feeds as $feed) { // Naked domain case (scheme://basehost)
$attr = []; $path = $baseParts['path'] ?? '/';
foreach ($feed->attributes as $attribute) {
$attr[$attribute->name] = trim($attribute->value); // Remove the filename part of the path if it exists (/base/path/file)
$path = implode('/', array_slice(explode('/', $path), 0, -1));
$hrefParts = parse_url($href);
// Root path case (/path) including relative scheme case (//host/path)
if ($hrefParts['path'] && $hrefParts['path'][0] == '/') {
$path = $hrefParts['path'];
} else {
$path = $path . '/' . $hrefParts['path'];
// Resolve arbitrary relative path
// Lifted from https://www.php.net/manual/en/function.realpath.php#84012
$parts = array_filter(explode('/', $path), 'strlen');
$absolutes = array();
foreach ($parts as $part) {
if ('.' == $part) continue;
if ('..' == $part) {
array_pop($absolutes);
} else {
$absolutes[] = $part;
}
} }
if (empty($feed_url) && !empty($attr['href'])) { $path = '/' . implode('/', $absolutes);
$feed_url = $attr["href"];
}
} }
return $feed_url; // Relative scheme case (//host/path)
$baseParts['host'] = $hrefParts['host'] ?? $baseParts['host'];
$baseParts['path'] = $path;
unset($baseParts['query']);
unset($baseParts['fragment']);
return Network::unparseURL($baseParts);
} }
/** /**