From d4802bcdb6593878b8a58973536b5e74536428d4 Mon Sep 17 00:00:00 2001
From: "J. King" <jking@jkingweb.ca>
Date: Tue, 27 Aug 2019 15:18:02 -0400
Subject: [PATCH] Handle IDNs

While IPv6 address normalization was originally planned, this was deemed
too much effort to bother with such a niche feature; IPv6 addresses are
instead passed through unmodified
---
 lib/Misc/URL.php             | 22 +++++++---------------
 tests/cases/Misc/TestURL.php |  3 +++
 2 files changed, 10 insertions(+), 15 deletions(-)

diff --git a/lib/Misc/URL.php b/lib/Misc/URL.php
index c8589b0..da47006 100644
--- a/lib/Misc/URL.php
+++ b/lib/Misc/URL.php
@@ -24,15 +24,16 @@ class URL {
      * Normalizations performed are:
      * 
      * - Lowercasing scheme
-     * - Lowercasing host names
-     * - IDN normalization (IDN rather than punycode is returned)
-     * - IPv6 address normalization
+     * - Lowercasing ASCII host names
+     * - IDN normalization
      * - Resolution of relative path segments
      * - Discarding empty path segments
      * - Discarding empty queries
      * - %-encoding normalization
      * - Fragment discarding
      * 
+     * It does NOT perform IPv6 address normalization, nor does it drop trailing slashes from paths
+     * 
      * @param string $url The URL to normalize. Relative URLs are returned unchanged
      * @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
@@ -56,11 +57,7 @@ class URL {
             }
             $out .= "@";
         }
-        if ($host[0] === "[") {
-            $out .= self::normalizeIPv6($host);
-        } else {
-            $out .= self::normalizeHost($host);
-        }
+        $out .= self::normalizeHost($host);
         $out .= isset($port) ? ":$port" : "";
         $out .= self::normalizePath($path ?? "");
         if (isset($query) && strlen($query)) {
@@ -75,13 +72,8 @@ class URL {
     }
 
     protected static function normalizeHost(string $host): string {
-        // stub
-        return $host;
-    }
-
-    protected static function normalizeIPv6(string $addr): string {
-        // stub
-        return $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 */
diff --git a/tests/cases/Misc/TestURL.php b/tests/cases/Misc/TestURL.php
index b6c95b2..2a8a773 100644
--- a/tests/cases/Misc/TestURL.php
+++ b/tests/cases/Misc/TestURL.php
@@ -58,6 +58,9 @@ class TestURL extends \JKingWeb\Arsse\Test\AbstractTest {
             ["http://example.com/a/../b/",    "http://example.com/b/"],
             ["http://example.com/.a/",        "http://example.com/.a/"],
             ["http://example.com/..a/",       "http://example.com/..a/"],
+            ["http://日本.example.com/",      "http://日本.example.com/"],
+            ["http://EXAMPLE.COM/",           "http://example.com/"],
+            ["http://É.example.com/",         "http://é.example.com/"],
         ];
     }
 }