PHP bug #77423

0001-Fix-77423-parse_url-will-deliver-a-wrong-host-to-use.patch

From 8a44b8b739921cb56e8bf539784318b57ed1f366 Mon Sep 17 00:00:00 2001
From: "Christoph M. Becker" <cmbecker69@gmx.de>
Date: Thu, 13 Feb 2020 16:58:01 +0100
Subject: [PATCH] Fix #77423: parse_url() will deliver a wrong host to user

To avoid that `parse_url()` returns an erroneous host, which would be
valid for `FILTER_VALIDATE_URL`, we make sure that only userinfo which
is valid according to RFC 3986 is treated as such.

For consistency with the existing url parsing code, we use ctype
functions, although that is not necessarily correct.
---
 ext/standard/tests/strings/url_t.phpt         |  6 ++--
 ext/standard/tests/url/bug77423.phpt          | 30 +++++++++++++++++++
 .../tests/url/parse_url_basic_001.phpt        |  6 ++--
 .../tests/url/parse_url_basic_003.phpt        |  2 +-
 .../tests/url/parse_url_basic_005.phpt        |  2 +-
 .../tests/url/parse_url_unterminated.phpt     |  6 ++--
 ext/standard/url.c                            | 20 +++++++++++++
 7 files changed, 58 insertions(+), 14 deletions(-)
 create mode 100644 ext/standard/tests/url/bug77423.phpt

diff --git a/ext/standard/tests/strings/url_t.phpt b/ext/standard/tests/strings/url_t.phpt
index 79ff3bc4a8..f564f59f06 100644
--- a/ext/standard/tests/strings/url_t.phpt
+++ b/ext/standard/tests/strings/url_t.phpt
@@ -575,15 +575,13 @@ $sample_urls = array (
   string(16) "some_page_ref123"
 }
 
---> http://secret@hideout@www.php.net:80/index.php?test=1&test2=char&test3=mixesCI#some_page_ref123: array(7) {
+--> http://secret@hideout@www.php.net:80/index.php?test=1&test2=char&test3=mixesCI#some_page_ref123: array(6) {
   ["scheme"]=>
   string(4) "http"
   ["host"]=>
-  string(11) "www.php.net"
+  string(26) "secret@hideout@www.php.net"
   ["port"]=>
   int(80)
-  ["user"]=>
-  string(14) "secret@hideout"
   ["path"]=>
   string(10) "/index.php"
   ["query"]=>
diff --git a/ext/standard/tests/url/bug77423.phpt b/ext/standard/tests/url/bug77423.phpt
new file mode 100644
index 0000000000..be03fe95e2
--- /dev/null
+++ b/ext/standard/tests/url/bug77423.phpt
@@ -0,0 +1,30 @@
+--TEST--
+Bug #77423 (parse_url() will deliver a wrong host to user)
+--FILE--
+<?php
+$urls = array(
+    "http://php.net\@aliyun.com/aaa.do",
+    "https://example.com\uFF03@bing.com",
+);
+foreach ($urls as $url) {
+    var_dump(filter_var($url, FILTER_VALIDATE_URL));
+    var_dump(parse_url($url));
+}
+?>
+--EXPECT--
+bool(false)
+array(3) {
+  ["scheme"]=>
+  string(4) "http"
+  ["host"]=>
+  string(19) "php.net\@aliyun.com"
+  ["path"]=>
+  string(7) "/aaa.do"
+}
+bool(false)
+array(2) {
+  ["scheme"]=>
+  string(5) "https"
+  ["host"]=>
+  string(26) "example.com\uFF03@bing.com"
+}
diff --git a/ext/standard/tests/url/parse_url_basic_001.phpt b/ext/standard/tests/url/parse_url_basic_001.phpt
index d59f080fbf..9b555a3167 100644
--- a/ext/standard/tests/url/parse_url_basic_001.phpt
+++ b/ext/standard/tests/url/parse_url_basic_001.phpt
@@ -506,15 +506,13 @@ echo "Done";
   string(16) "some_page_ref123"
 }
 
---> http://secret@hideout@www.php.net:80/index.php?test=1&test2=char&test3=mixesCI#some_page_ref123: array(7) {
+--> http://secret@hideout@www.php.net:80/index.php?test=1&test2=char&test3=mixesCI#some_page_ref123: array(6) {
   ["scheme"]=>
   string(4) "http"
   ["host"]=>
-  string(11) "www.php.net"
+  string(26) "secret@hideout@www.php.net"
   ["port"]=>
   int(80)
-  ["user"]=>
-  string(14) "secret@hideout"
   ["path"]=>
   string(10) "/index.php"
   ["query"]=>
diff --git a/ext/standard/tests/url/parse_url_basic_003.phpt b/ext/standard/tests/url/parse_url_basic_003.phpt
index f917b77b97..8bff9c6299 100644
--- a/ext/standard/tests/url/parse_url_basic_003.phpt
+++ b/ext/standard/tests/url/parse_url_basic_003.phpt
@@ -68,7 +68,7 @@ echo "Done";
 --> http://secret:@www.php.net/index.php?test=1&test2=char&test3=mixesCI#some_page_ref123   : string(11) "www.php.net"
 --> http://:hideout@www.php.net:80/index.php?test=1&test2=char&test3=mixesCI#some_page_ref123   : string(11) "www.php.net"
 --> http://secret:hideout@www.php.net/index.php?test=1&test2=char&test3=mixesCI#some_page_ref123   : string(11) "www.php.net"
---> http://secret@hideout@www.php.net:80/index.php?test=1&test2=char&test3=mixesCI#some_page_ref123   : string(11) "www.php.net"
+--> http://secret@hideout@www.php.net:80/index.php?test=1&test2=char&test3=mixesCI#some_page_ref123   : string(26) "secret@hideout@www.php.net"
 --> http://secret:hid:out@www.php.net:80/index.php?test=1&test2=char&test3=mixesCI#some_page_ref123   : string(11) "www.php.net"
 --> nntp://news.php.net   : string(12) "news.php.net"
 --> ftp://ftp.gnu.org/gnu/glic/glibc.tar.gz   : string(11) "ftp.gnu.org"
diff --git a/ext/standard/tests/url/parse_url_basic_005.phpt b/ext/standard/tests/url/parse_url_basic_005.phpt
index 716d8d1dfa..d7883eca8b 100644
--- a/ext/standard/tests/url/parse_url_basic_005.phpt
+++ b/ext/standard/tests/url/parse_url_basic_005.phpt
@@ -68,7 +68,7 @@ echo "Done";
 --> http://secret:@www.php.net/index.php?test=1&test2=char&test3=mixesCI#some_page_ref123   : string(6) "secret"
 --> http://:hideout@www.php.net:80/index.php?test=1&test2=char&test3=mixesCI#some_page_ref123   : string(0) ""
 --> http://secret:hideout@www.php.net/index.php?test=1&test2=char&test3=mixesCI#some_page_ref123   : string(6) "secret"
---> http://secret@hideout@www.php.net:80/index.php?test=1&test2=char&test3=mixesCI#some_page_ref123   : string(14) "secret@hideout"
+--> http://secret@hideout@www.php.net:80/index.php?test=1&test2=char&test3=mixesCI#some_page_ref123   : NULL
 --> http://secret:hid:out@www.php.net:80/index.php?test=1&test2=char&test3=mixesCI#some_page_ref123   : string(6) "secret"
 --> nntp://news.php.net   : NULL
 --> ftp://ftp.gnu.org/gnu/glic/glibc.tar.gz   : NULL
diff --git a/ext/standard/tests/url/parse_url_unterminated.phpt b/ext/standard/tests/url/parse_url_unterminated.phpt
index 94975ee889..e6a8b50566 100644
--- a/ext/standard/tests/url/parse_url_unterminated.phpt
+++ b/ext/standard/tests/url/parse_url_unterminated.phpt
@@ -508,15 +508,13 @@ echo "Done";
   string(16) "some_page_ref123"
 }
 
---> http://secret@hideout@www.php.net:80/index.php?test=1&test2=char&test3=mixesCI#some_page_ref123: array(7) {
+--> http://secret@hideout@www.php.net:80/index.php?test=1&test2=char&test3=mixesCI#some_page_ref123: array(6) {
   ["scheme"]=>
   string(4) "http"
   ["host"]=>
-  string(11) "www.php.net"
+  string(26) "secret@hideout@www.php.net"
   ["port"]=>
   int(80)
-  ["user"]=>
-  string(14) "secret@hideout"
   ["path"]=>
   string(10) "/index.php"
   ["query"]=>
diff --git a/ext/standard/url.c b/ext/standard/url.c
index a04d941498..319bcf219c 100644
--- a/ext/standard/url.c
+++ b/ext/standard/url.c
@@ -87,6 +87,22 @@ PHPAPI php_url *php_url_parse(char const *str)
 	return php_url_parse_ex(str, strlen(str));
 }
 
+static int is_userinfo_valid(const char *str, size_t len)
+{
+	char *valid = "-._~!$&'()*+,;=:";
+	char *p = str;
+	while (p - str < len) {
+		if (isalpha(*p) || isdigit(*p) || strchr(valid, *p)) {
+			p++;
+		} else if (*p == '%' && p - str <= len - 3 && isdigit(*(p+1)) && isxdigit(*(p+2))) {
+			p += 3;
+		} else {
+			return 0;
+		}
+	}
+	return 1;
+}
+
 /* {{{ php_url_parse
  */
 PHPAPI php_url *php_url_parse_ex(char const *str, size_t length)
@@ -228,6 +244,9 @@ PHPAPI php_url *php_url_parse_ex(char const *str, size_t length)
 			ret->pass = zend_string_init(pp, (p-pp), 0);
 			php_replace_controlchars_ex(ZSTR_VAL(ret->pass), ZSTR_LEN(ret->pass));
 		} else {
+			if (!is_userinfo_valid(s, p-s)) {
+				goto check_port;
+			}
 			ret->user = zend_string_init(s, (p-s), 0);
 			php_replace_controlchars_ex(ZSTR_VAL(ret->user), ZSTR_LEN(ret->user));
 		}
@@ -235,6 +254,7 @@ PHPAPI php_url *php_url_parse_ex(char const *str, size_t length)
 		s = p + 1;
 	}
 
+check_port:
 	/* check for port */
 	if (s < ue && *s == '[' && *(e-1) == ']') {
 		/* Short circuit portscan,
-- 
2.25.0.windows.1