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.

(cherry picked from commit b132da7f9df3)

8 files changed, 62 insertions, 16 deletions

diff --git a/NEWS b/NEWS
index 4bee6b6484..524c295e46 100644
--- a/NEWS
+++ b/NEWS
@@ -85,6 +85,8 @@ PHP                                                                        NEWS
 
 - Standard:
   . Fixed bug #80366 (Return Value of zend_fstat() not Checked). (sagpant, cmb)
+  . Fixed bug #77423 (FILTER_VALIDATE_URL accepts URLs with invalid userinfo).
+    (CVE-2020-7071) (cmb)
 
 - Tidy:
   . Fixed bug #77594 (ob_tidyhandler is never reset). (cmb)
diff --git a/ext/standard/tests/strings/url_t.phpt b/ext/standard/tests/strings/url_t.phpt
index caa93cb9cf..dc13018b14 100644
--- a/ext/standard/tests/strings/url_t.phpt
+++ b/ext/standard/tests/strings/url_t.phpt
@@ -589,15 +589,13 @@ $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 f3abd703b2..89b1b7b3d6 100644
--- a/ext/standard/tests/url/parse_url_basic_001.phpt
+++ b/ext/standard/tests/url/parse_url_basic_001.phpt
@@ -514,15 +514,13 @@ 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 1eb64d6a1b..da35b92782 100644
--- a/ext/standard/tests/url/parse_url_basic_003.phpt
+++ b/ext/standard/tests/url/parse_url_basic_003.phpt
@@ -62,7 +62,7 @@ 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 dfbe7e7971..731a4adb31 100644
--- a/ext/standard/tests/url/parse_url_basic_005.phpt
+++ b/ext/standard/tests/url/parse_url_basic_005.phpt
@@ -62,7 +62,7 @@ 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 7c9150a513..f859fc7d40 100644
--- a/ext/standard/tests/url/parse_url_unterminated.phpt
+++ b/ext/standard/tests/url/parse_url_unterminated.phpt
@@ -522,15 +522,13 @@ 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 a33091a86b..113e010024 100644
--- a/ext/standard/url.c
+++ b/ext/standard/url.c
@@ -92,6 +92,22 @@ static const char *binary_strcspn(const char *s, const char *e, const char *char
 	return e;
 }
 
+static int is_userinfo_valid(const char *str, size_t len)
+{
+	const char *valid = "-._~!$&'()*+,;=:";
+	const 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)
 {
@@ -233,13 +249,17 @@ parse_host:
 			ret->pass = zend_string_init(pp, (p-pp), 0);
 			php_replace_controlchars_ex(ZSTR_VAL(ret->pass), ZSTR_LEN(ret->pass));
 		} else {
-			ret->user = zend_string_init(s, (p-s), 0);
-			php_replace_controlchars_ex(ZSTR_VAL(ret->user), ZSTR_LEN(ret->user));
+			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));
 		}
 
 		s = p + 1;
 	}
 
+check_port:
 	/* check for port */
 	if (s < ue && *s == '[' && *(e-1) == ']') {
 		/* Short circuit portscan,