Bug 829816 - Treat \0 and U+0000 in CSS style sheets as U+FFFD. r=dbaron

This commit is contained in:
Corey Ford 2013-06-26 15:17:14 -07:00
parent 7a5a420e22
commit d0b5b5aabf
10 changed files with 109 additions and 39 deletions

View File

@ -1,22 +1,22 @@
<!doctype html>
<html><head>
<!-- This is a test for behavior made up by Mozilla in the absence of
specification. Future CSS specifications may define the behavior
differently. -->
<!-- This tests behavior specified in CSS Syntax Level 3,
as of the Editor's Draft 19 June 2013:
http://dev.w3.org/csswg/css-syntax/ -->
<title>\0 in CSS</title>
<style>
p#a:before { content: "0x" }
p#b:before { content: "00x" }
p#c:before { content: "000x" }
p#d:before { content: "0000x" }
p#e:before { content: "00000x" }
p#f:before { content: "000000x" }
p#g:before { content: "0 x" }
p#h:before { content: "00 x" }
p#i:before { content: "000 x" }
p#j:before { content: "0000 x" }
p#k:before { content: "00000 x" }
p#l:before { content: "000000 x" }
p#a:before { content: "\FFFDx" }
p#b:before { content: "\FFFDx" }
p#c:before { content: "\FFFDx" }
p#d:before { content: "\FFFDx" }
p#e:before { content: "\FFFDx" }
p#f:before { content: "\FFFDx" }
p#g:before { content: "\FFFD x" }
p#h:before { content: "\FFFD x" }
p#i:before { content: "\FFFD x" }
p#j:before { content: "\FFFD x" }
p#k:before { content: "\FFFD x" }
p#l:before { content: "\FFFD x" }
</style>
</head><body>
<p id="a">(a)</p>

View File

@ -1,8 +1,8 @@
<!doctype html>
<html><head>
<!-- This is a test for behavior made up by Mozilla in the absence of
specification. Future CSS specifications may define the behavior
differently. -->
<!-- This tests behavior specified in CSS Syntax Level 3,
as of the Editor's Draft 19 June 2013:
http://dev.w3.org/csswg/css-syntax/ -->
<title>\0 in CSS</title>
<style>
p#a:before { content: "\0x" }

Binary file not shown.

View File

@ -138,21 +138,24 @@ IsVertSpace(int32_t ch) {
}
/**
* True if 'ch' is a character that can appear in the middle of an
* identifier.
* True if 'ch' is a character that can appear in the middle of an identifier.
* This includes U+0000 since it is handled as U+FFFD, but for purposes of
* GatherText it should not be included in IsOpenCharClass.
*/
static inline bool
IsIdentChar(int32_t ch) {
return IsOpenCharClass(ch, IS_IDCHAR);
return IsOpenCharClass(ch, IS_IDCHAR) || ch == 0;
}
/**
* True if 'ch' is a character that by itself begins an identifier.
* This includes U+0000 since it is handled as U+FFFD, but for purposes of
* GatherText it should not be included in IsOpenCharClass.
* (This is a subset of IsIdentChar.)
*/
static inline bool
IsIdentStart(int32_t ch) {
return IsOpenCharClass(ch, IS_IDSTART);
return IsOpenCharClass(ch, IS_IDSTART) || ch == 0;
}
/**
@ -539,7 +542,7 @@ nsCSSScanner::GatherEscape(nsString& aOutput, bool aInString)
// character.
Advance();
if (!aInString) {
aOutput.Append(0xFFFD);
aOutput.Append(UCS2_REPLACEMENT_CHAR);
}
return true;
}
@ -561,7 +564,11 @@ nsCSSScanner::GatherEscape(nsString& aOutput, bool aInString)
// return, or form feed) can be escaped with a backslash to remove
// its special meaning." -- CSS2.1 section 4.1.3
Advance(2);
aOutput.Append(ch);
if (ch == 0) {
aOutput.Append(UCS2_REPLACEMENT_CHAR);
} else {
aOutput.Append(ch);
}
return true;
}
@ -584,22 +591,21 @@ nsCSSScanner::GatherEscape(nsString& aOutput, bool aInString)
ch = Peek();
} while (i < 6 && IsHexDigit(ch));
// Silently deleting \0 opens a content-filtration loophole (see
// bug 228856), so what we do instead is pretend the "cancels the
// meaning of special characters" rule applied.
// "Interpret the hex digits as a hexadecimal number. If this number is zero,
// or is greater than the maximum allowed codepoint, return U+FFFD
// REPLACEMENT CHARACTER" -- CSS Syntax Level 3
if (MOZ_UNLIKELY(val == 0)) {
do {
aOutput.Append('0');
} while (--i);
aOutput.Append(UCS2_REPLACEMENT_CHAR);
} else {
AppendUCS4ToUTF16(ENSURE_VALID_CHAR(val), aOutput);
// Consume exactly one whitespace character after a nonzero
// hexadecimal escape sequence.
if (IsVertSpace(ch)) {
AdvanceLine();
} else if (IsHorzSpace(ch)) {
Advance();
}
}
// Consume exactly one whitespace character after a
// hexadecimal escape sequence.
if (IsVertSpace(ch)) {
AdvanceLine();
} else if (IsHorzSpace(ch)) {
Advance();
}
return true;
}
@ -644,6 +650,11 @@ nsCSSScanner::GatherText(uint8_t aClass, nsString& aText)
int32_t ch = Peek();
MOZ_ASSERT(!IsOpenCharClass(ch, aClass),
"should not have exited the inner loop");
if (ch == 0) {
Advance();
aText.Append(UCS2_REPLACEMENT_CHAR);
continue;
}
if (ch != '\\') {
break;

View File

@ -84,6 +84,8 @@ MOCHITEST_FILES = test_acid3_test46.html \
test_bug716226.html \
test_bug765590.html \
test_bug798567.html \
test_bug829816.html \
file_bug829816.css \
test_cascade.html \
test_ch_ex_no_infloops.html \
test_compute_data_with_start_struct.html \

Binary file not shown.

View File

@ -0,0 +1,56 @@
<!DOCTYPE HTML>
<html>
<!--
https://bugzilla.mozilla.org/show_bug.cgi?id=829816
-->
<head>
<meta charset="utf-8">
<title>Test for Bug 829816</title>
<script type="application/javascript" src="/tests/SimpleTest/SimpleTest.js"></script>
<link rel="stylesheet" type="text/css" href="/tests/SimpleTest/test.css"/>
<style type="text/css">
b { content: "\0"; counter-reset: \0 }
b { content: "\00"; counter-reset: \00 }
b { content: "\000"; counter-reset: \000 }
b { content: "\0000"; counter-reset: \0000 }
b { content: "\00000"; counter-reset: \00000 }
b { content: "\000000"; counter-reset: \000000 }
</style>
<!-- U+0000 characters in <style> would be replaced by the HTML parser -->
<link rel="stylesheet" type="text/css" href="file_bug829816.css"/>
<script type="application/javascript">
/** Test for Bug 829816 **/
var ss = document.styleSheets[1];
for (var i = 0; i < 6; i++) {
is(ss.cssRules[i].style.content, "\"\uFFFD\"",
"\\0 in strings should be converted to U+FFFD");
is(ss.cssRules[i].style.counterReset, "\uFFFD",
"\\0 in identifiers should be converted to U+FFFD");
}
is(document.styleSheets[2].cssRules[0].style.content, "\"\uFFFD\"",
"U+0000 in strings should be converted to U+FFFD");
is(document.styleSheets[2].cssRules[0].style.counterReset, "\uFFFD",
"U+0000 in identifiers should be converted to U+FFFD");
is(document.styleSheets[2].cssRules[1].style.content, "\"\uFFFD\"",
"U+0000 in strings should be converted to U+FFFD");
is(document.styleSheets[2].cssRules[1].style.counterReset, "\uFFFD",
"U+0000 in identifiers should be converted to U+FFFD");
</script>
</head>
<body>
<a target="_blank" href="https://bugzilla.mozilla.org/show_bug.cgi?id=829816">Mozilla Bug 829816</a>
<p id="display"></p>
<div id="content" style="display: none">
</div>
<pre id="test">
</pre>
</body>
</html>

View File

@ -100,8 +100,9 @@ const substitutions = [
// U+000A LINE FEED, U+000C FORM FEED, and U+000D CARRIAGE RETURN
// cannot be put into a CSS token as escaped literal characters, so
// we do them with hex escapes instead.
{ t: "\\\x00\\\x01\\\x02\\\x03", i: "\\0 \\1 \\2 \\3 ",
s: "\\0 \\1 \\2 \\3 " },
// The parser replaces U+0000 with U+FFFD.
{ t: "\\\x00\\\x01\\\x02\\\x03", i: "<22>\\1 \\2 \\3 ",
s: "<22>\\1 \\2 \\3 " },
{ t: "\\\x04\\\x05\\\x06\\\x07", i: "\\4 \\5 \\6 \\7 ",
s: "\\4 \\5 \\6 \\7 " },
{ t: "\\\x08\\\x09\\000A\\\x0B", i: "\\8 \\9 \\A \\B ",