UnescapeURLComponent: Don't unescape UTF-8 space characters.

Bug: 824715
Change-Id: I71d7f38a2dbe9de6515b8e9d284ab622c2311276
Reviewed-on: https://chromium-review.googlesource.com/1014367
Commit-Queue: Matt Giuca <mgiuca@chromium.org>
Reviewed-by: Matt Giuca <mgiuca@chromium.org>
Cr-Commit-Position: refs/heads/master@{#551260}

net/base/escape.cc

@@ -228,7 +228,28 @@ bool ShouldUnescapeCodePoint(UnescapeRule::Type rules, uint32_t code_point) {
       code_point == 0x1F50F ||  // LOCK WITH INK PEN    (%F0%9F%94%8F)
       code_point == 0x1F510 ||  // CLOSED LOCK WITH KEY (%F0%9F%94%90)
       code_point == 0x1F512 ||  // LOCK                 (%F0%9F%94%92)
-      code_point == 0x1F513);   // OPEN LOCK            (%F0%9F%94%93)
+      code_point == 0x1F513 ||  // OPEN LOCK            (%F0%9F%94%93)
+
+      // Spaces are also banned, as they can be used to scroll text out of view.
+      code_point == 0x0085 ||  // NEXT LINE                  (%C2%85)
+      code_point == 0x00A0 ||  // NO-BREAK SPACE             (%C2%A0)
+      code_point == 0x1680 ||  // OGHAM SPACE MARK           (%E1%9A%80)
+      code_point == 0x2000 ||  // EN QUAD                    (%E2%80%80)
+      code_point == 0x2001 ||  // EM QUAD                    (%E2%80%81)
+      code_point == 0x2002 ||  // EN SPACE                   (%E2%80%82)
+      code_point == 0x2003 ||  // EM SPACE                   (%E2%80%83)
+      code_point == 0x2004 ||  // THREE-PER-EM SPACE         (%E2%80%84)
+      code_point == 0x2005 ||  // FOUR-PER-EM SPACE          (%E2%80%85)
+      code_point == 0x2006 ||  // SIX-PER-EM SPACE           (%E2%80%86)
+      code_point == 0x2007 ||  // FIGURE SPACE               (%E2%80%87)
+      code_point == 0x2008 ||  // PUNCTUATION SPACE          (%E2%80%88)
+      code_point == 0x2009 ||  // THIN SPACE                 (%E2%80%89)
+      code_point == 0x200A ||  // HAIR SPACE                 (%E2%80%8A)
+      code_point == 0x2028 ||  // LINE SEPARATOR             (%E2%80%A8)
+      code_point == 0x2029 ||  // PARAGRAPH SEPARATOR        (%E2%80%A9)
+      code_point == 0x202F ||  // NARROW NO-BREAK SPACE      (%E2%80%AF)
+      code_point == 0x205F ||  // MEDIUM MATHEMATICAL SPACE  (%E2%81%9F)
+      code_point == 0x3000);   // IDEOGRAPHIC SPACE          (%E3%80%80)
 }
 
 // Unescapes |escaped_text| according to |rules|, returning the resulting

net/base/escape_unittest.cc

@@ -153,12 +153,12 @@ TEST(EscapeTest, UnescapeURLComponent) {
        "Some%20random text %25%2dOK"},
       {"Some%20random text %25%2dOK", UnescapeRule::NORMAL,
        "Some%20random text %25-OK"},
-      {"Some%20random text %25%E2%80", UnescapeRule::NORMAL,
-       "Some%20random text %25\xE2\x80"},
-      {"Some%20random text %25%E2%80OK", UnescapeRule::NORMAL,
-       "Some%20random text %25\xE2\x80OK"},
-      {"Some%20random text %25%E2%80%84OK", UnescapeRule::NORMAL,
-       "Some%20random text %25\xE2\x80\x84OK"},
+      {"Some%20random text %25%E1%A6", UnescapeRule::NORMAL,
+       "Some%20random text %25\xE1\xA6"},
+      {"Some%20random text %25%E1%A6OK", UnescapeRule::NORMAL,
+       "Some%20random text %25\xE1\xA6OK"},
+      {"Some%20random text %25%E1%A6%99OK", UnescapeRule::NORMAL,
+       "Some%20random text %25\xE1\xA6\x99OK"},
 
       // BiDi Control characters should not be unescaped unless explicity told
       // to
@@ -236,6 +236,18 @@ TEST(EscapeTest, UnescapeURLComponent) {
        UnescapeRule::NORMAL | UnescapeRule::SPOOFING_AND_CONTROL_CHARS,
        "Some%20random text %25\xF0\x9F\x94\x93OK"},
 
+      // Spaces
+      {"(%C2%85)(%C2%A0)(%E1%9A%80)(%E2%80%80)", UnescapeRule::NORMAL,
+       "(%C2%85)(%C2%A0)(%E1%9A%80)(%E2%80%80)"},
+      {"(%E2%80%81)(%E2%80%82)(%E2%80%83)(%E2%80%84)", UnescapeRule::NORMAL,
+       "(%E2%80%81)(%E2%80%82)(%E2%80%83)(%E2%80%84)"},
+      {"(%E2%80%85)(%E2%80%86)(%E2%80%87)(%E2%80%88)", UnescapeRule::NORMAL,
+       "(%E2%80%85)(%E2%80%86)(%E2%80%87)(%E2%80%88)"},
+      {"(%E2%80%89)(%E2%80%8A)(%E2%80%A8)(%E2%80%A9)", UnescapeRule::NORMAL,
+       "(%E2%80%89)(%E2%80%8A)(%E2%80%A8)(%E2%80%A9)"},
+      {"(%E2%80%AF)(%E2%81%9F)(%E3%80%80)", UnescapeRule::NORMAL,
+       "(%E2%80%AF)(%E2%81%9F)(%E3%80%80)"},
+
       // Two spoofing characters in a row should not be unescaped.
       {"%D8%9C%D8%9C", UnescapeRule::NORMAL, "%D8%9C%D8%9C"},
       // Non-spoofing characters surrounded by spoofing characters should be