The clean & modern RSS server that doesn't give you any crap. https://thearsse.com/
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

170 lines
6.2 KiB

<?php
/** @license MIT
* Copyright 2017 J. King, Dustin Wilson et al.
* See LICENSE and AUTHORS files for details */
declare(strict_types=1);
namespace JKingWeb\Arsse\Misc;
/**
* A collection of functions for manipulating URLs
*/
class URL {
/** Returns whether a URL is absolute i.e. has a scheme */
public static function absolute(string $url): bool {
return (bool) strlen((string) parse_url($url, \PHP_URL_SCHEME));
}
/** Normalizes a URL
*
* Normalizations performed are:
*
* - Lowercasing scheme
* - Lowercasing ASCII host names
* - IDN normalization
* - IPv6 address normalization
* - Resolution of relative path segments
* - Discarding empty path segments
* - Discarding empty queries
* - Generic percent-encoding normalization
* - Fragment discarding
*
* It does NOT drop trailing slashes from paths, nor does it perform Unicode normalization or context-aware percent-encoding normalization
*
* @param string $url The URL to normalize
* @param string $u Username to add to the URL, replacing any existing credentials
* @param string $p Password to add to the URL, if a username is specified
*/
public static function normalize(string $url, string $u = null, string $p = null): string {
extract(parse_url($url));
$out = "";
if (isset($scheme)) {
$out .= strtolower($scheme).":";
}
if (isset($host)) {
$out .= "//";
if (strlen($u ?? "")) {
$out .= self::normalizeEncoding(rawurlencode($u));
if (strlen($p ?? "")) {
$out .= ":".self::normalizeEncoding(rawurlencode($p));
}
$out .= "@";
} elseif (strlen($user ?? "")) {
$out .= self::normalizeEncoding($user);
if (strlen($pass ?? "")) {
$out .= ":".self::normalizeEncoding($pass);
}
$out .= "@";
}
$out .= self::normalizeHost($host);
$out .= isset($port) ? ":$port" : "";
}
$out .= self::normalizePath($path ?? "", isset($host));
if (isset($query) && strlen($query)) {
$out .= "?".self::normalizeEncoding($query);
}
return $out;
}
/** Perform percent-encoding normalization for a given URL component */
protected static function normalizeEncoding(string $part): string {
$pos = 0;
$end = strlen($part);
$out = "";
// process each character in sequence
while ($pos < $end) {
$c = $part[$pos];
if ($c === "%") {
// the % character signals an encoded character...
$d = substr($part, $pos+1, 2);
if (!preg_match("/^[0-9a-fA-F]{2}$/", $d)) {
// unless there are fewer than two characters left in the string or the two characters are not hex digits
$d = ord($c);
} else {
$d = hexdec($d);
$pos += 2;
}
} else {
$d = ord($c);
}
$dc = chr($d);
if ($d < 0x21 || $d > 0x7E || $d == 0x25) {
// these characters are always encoded
$out .= "%".strtoupper(dechex($d));
} elseif (preg_match("/[a-zA-Z0-9\._~-]/", $dc)) {
// these characters are never encoded
$out .= $dc;
} else {
// these characters are passed through as-is
if ($c === "%") {
$out .= "%".strtoupper(dechex($d));
} else {
$out .= $c;
}
}
$pos++;
}
return $out;
}
/** Normalizes a hostname per IDNA:2008 */
protected static function normalizeHost(string $host): string {
if ($host[0] === "[" && substr($host, -1) === "]") {
// normalize IPv6 addresses
$addr = @inet_pton(substr($host, 1, strlen($host) - 2));
if ($addr !== false) {
return "[".inet_ntop($addr)."]";
}
}
$idn = idn_to_ascii($host, \IDNA_NONTRANSITIONAL_TO_ASCII, \INTL_IDNA_VARIANT_UTS46);
return $idn !== false ? idn_to_utf8($idn, \IDNA_NONTRANSITIONAL_TO_UNICODE, \INTL_IDNA_VARIANT_UTS46) : $host;
}
/** Normalizes the whole path segment to remove empty segments and relative segments */
protected static function normalizePath(string $path, bool $hasHost): string {
$parts = explode("/", self::normalizeEncoding($path));
$absolute = ($hasHost || $path[0] === "/");
$index = (substr($path, -1) === "/");
$out = [];
foreach ($parts as $p) {
switch ($p) {
case "":
case ".":
break;
case "..":
array_pop($out);
break;
default:
$out[] = $p;
}
}
$out = implode("/", $out);
$out = ($absolute ? "/" : "").$out.($index ? "/" : "");
return str_replace("//", "/", $out);
}
/** Appends data to a URL's query component
*
* @param string $url The input URL
* @param string $data The data to append. This should already be escaped where necessary and not start with any delimiter
* @param string $glue The query subcomponent delimiter, usually "&". If the URL has no query, "?" will be prepended instead
*/
public static function queryAppend(string $url, string $data, string $glue = "&"): string {
if (!strlen($data)) {
return $url;
}
$insPos = strpos($url, "#");
$insPos = $insPos === false ? strlen($url) : $insPos;
$qPos = strpos($url, "?");
$hasQuery = $qPos !== false;
$glue = $hasQuery ? $glue : "?";
if ($hasQuery && $insPos > 0) {
if ($url[$insPos - 1] === $glue || ($insPos - 1) == $qPos) {
// if the URL already has excess glue, use it
$glue = "";
}
}
return substr($url, 0, $insPos).$glue.$data.substr($url, $insPos);
}
}