- Take account of surrogate pairs.

diff --git a/ext/standard/html.c b/ext/standard/html.c
index 14b4f2cabad614d6bbb0113091673c7a1637cf96..2cc1cc287b73ac0102f5d030eff6368958aaceb7 100644 (file)
--- a/ext/standard/html.c
+++ b/ext/standard/html.c
@@ -566,6 +566,8 @@ inline static unsigned int get_next_char(enum entity_charset charset,
                                        this_char = ((c & 0x0f) << 12) | ((str[pos + 1] & 0x3f) << 6) | (str[pos + 2] & 0x3f);
                                        if (this_char < 0x800) {
                                                MB_FAILURE(pos);
+                                       } else if (this_char >= 0xd800 && this_char <= 0xdfff) {
+                                               MB_FAILURE(pos);
                                        }
                                        MB_WRITE((unsigned char)c);
                                        MB_WRITE((unsigned char)str[pos + 1]);

diff --git a/ext/standard/tests/strings/bug49785.phpt b/ext/standard/tests/strings/bug49785.phpt
index f344855931cb672074d4957e8d0dde57813829fd..3f60e9fd372ccc79e67f22ae1fd2a29ce97a4ecd 100644 (file)
--- a/ext/standard/tests/strings/bug49785.phpt
+++ b/ext/standard/tests/strings/bug49785.phpt
@@ -36,10 +36,14 @@ var_dump(_bin2hex(htmlentities("\xf7\xbf\xbf\xff", ENT_QUOTES, "UTF-8")));
 var_dump(_bin2hex(htmlentities("\xf8\x88\x80\x80\x80", ENT_QUOTES, "UTF-8")));
 
 echo "--\n";
-// UTF-8: alternative (invalid) UTF-8 sequence
+// UTF-8: alternative (invalid) UTF-8 sequence / surrogate pairs
 var_dump(_bin2hex(htmlspecialchars("\xc0\xa6", ENT_QUOTES, 'UTF-8')));
 var_dump(_bin2hex(htmlspecialchars("\xe0\x80\xa6", ENT_QUOTES, 'UTF-8')));
 var_dump(_bin2hex(htmlspecialchars("\xf0\x80\x80\xa6", ENT_QUOTES, 'UTF-8')));
+var_dump(_bin2hex(htmlspecialchars("\xec\xbf\xbf", ENT_QUOTES, 'UTF-8')));
+var_dump(_bin2hex(htmlspecialchars("\xed\xa0\x80", ENT_QUOTES, 'UTF-8')));
+var_dump(_bin2hex(htmlspecialchars("\xed\xbf\xbf", ENT_QUOTES, 'UTF-8')));
+var_dump(_bin2hex(htmlspecialchars("\xee\x80\x80", ENT_QUOTES, 'UTF-8')));
 
 // Shift_JIS: non-lead byte >= 0x80
 var_dump(_bin2hex(htmlspecialchars("\x80", ENT_QUOTES, 'Shift_JIS')));
@@ -158,6 +162,10 @@ string(0) ""
 string(0) ""
 string(0) ""
 string(0) ""
+string(6) "ecbfbf"
+string(0) ""
+string(0) ""
+string(6) "ee8080"
 string(2) "80"
 string(2) "a0"
 string(2) "a1"