commit 2933a8241a54af436f2df5eac73aa2acf5eabd40 (tree)
parent 9390e8b84883331757d3ce15cfca89279aceb090
Author: hryx <codroid@gmail.com>
Date: Sun, 5 Jan 2020 23:16:38 -0800
json: disallow overlong and out-of-range UTF-8
Fixes #2379
= Overlong (non-shortest) sequences
UTF-8's unique encoding scheme allows for some Unicode codepoints
to be represented in multiple ways. For any of these characters,
the spec forbids all but the shortest form. These disallowed longer
sequences are called "overlong". As an interesting side effect of
this rule, the bytes C0 and C1 never appear in valid UTF-8.
= Codepoint range
UTF-8 disallows representation of codepoints beyond U+10FFFF,
which is the highest character which can be encoded in UTF-16.
Because a 4-byte sequence is capable of resulting in such characters,
they must be explicitly rejected. This rule also has an interesting
side effect, which is that bytes F5 to FF never appear.
= References
Detecting an overlong version of a codepoint could get gnarly, but
luckily The Unicode Consortium did the hard work by creating this
handy table of valid byte sequences:
https://unicode.org/versions/corrigendum1.html
I thought this mapped nicely to the parser's state machine, so I
rearranged the relevant states to make use of it.
Diffstat:
| M | lib/std/json.zig | | | 76 | +++++++++++++++++++++++++++++++++++++++++++++++++++++++++------------------- |
| M | lib/std/json/test.zig | | | 66 | ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ |
2 files changed, 123 insertions(+), 19 deletions(-)
diff --git a/lib/std/json.zig b/lib/std/json.zig
@@ -87,6 +87,8 @@ pub const StreamingParser = struct {
string_last_was_high_surrogate: bool,
// Used inside of StringEscapeHexUnicode* states
string_unicode_codepoint: u21,
+ // The first byte needs to be stored to validate 3- and 4-byte sequences.
+ sequence_first_byte: u8 = undefined,
// When in .Number states, is the number a (still) valid integer?
number_is_integer: bool,
@@ -132,9 +134,12 @@ pub const StreamingParser = struct {
ValueBeginNoClosing,
String,
- StringUtf8Byte3,
- StringUtf8Byte2,
- StringUtf8Byte1,
+ StringUtf8Byte2Of2,
+ StringUtf8Byte2Of3,
+ StringUtf8Byte3Of3,
+ StringUtf8Byte2Of4,
+ StringUtf8Byte3Of4,
+ StringUtf8Byte4Of4,
StringEscapeCharacter,
StringEscapeHexUnicode4,
StringEscapeHexUnicode3,
@@ -581,35 +586,68 @@ pub const StreamingParser = struct {
// non-control ascii
p.string_last_was_high_surrogate = false;
},
- 0xC0...0xDF => {
- p.state = .StringUtf8Byte1;
+ 0xC2...0xDF => {
+ p.state = .StringUtf8Byte2Of2;
},
0xE0...0xEF => {
- p.state = .StringUtf8Byte2;
+ p.state = .StringUtf8Byte2Of3;
+ p.sequence_first_byte = c;
},
- 0xF0...0xFF => {
- p.state = .StringUtf8Byte3;
+ 0xF0...0xF4 => {
+ p.state = .StringUtf8Byte2Of4;
+ p.sequence_first_byte = c;
},
else => {
return error.InvalidUtf8Byte;
},
},
- .StringUtf8Byte3 => switch (c >> 6) {
- 0b10 => p.state = .StringUtf8Byte2,
+ .StringUtf8Byte2Of2 => switch (c >> 6) {
+ 0b10 => p.state = .String,
else => return error.InvalidUtf8Byte,
},
-
- .StringUtf8Byte2 => switch (c >> 6) {
- 0b10 => p.state = .StringUtf8Byte1,
+ .StringUtf8Byte2Of3 => {
+ switch (p.sequence_first_byte) {
+ 0xE0 => switch (c) {
+ 0xA0...0xBF => {},
+ else => return error.InvalidUtf8Byte,
+ },
+ 0xE1...0xEF => switch (c) {
+ 0x80...0xBF => {},
+ else => return error.InvalidUtf8Byte,
+ },
+ else => return error.InvalidUtf8Byte,
+ }
+ p.state = .StringUtf8Byte3Of3;
+ },
+ .StringUtf8Byte3Of3 => switch (c) {
+ 0x80...0xBF => p.state = .String,
else => return error.InvalidUtf8Byte,
},
-
- .StringUtf8Byte1 => switch (c >> 6) {
- 0b10 => {
- p.state = .String;
- p.string_last_was_high_surrogate = false;
- },
+ .StringUtf8Byte2Of4 => {
+ switch (p.sequence_first_byte) {
+ 0xF0 => switch (c) {
+ 0x90...0xBF => {},
+ else => return error.InvalidUtf8Byte,
+ },
+ 0xF1...0xF3 => switch (c) {
+ 0x80...0xBF => {},
+ else => return error.InvalidUtf8Byte,
+ },
+ 0xF4 => switch (c) {
+ 0x80...0x8F => {},
+ else => return error.InvalidUtf8Byte,
+ },
+ else => return error.InvalidUtf8Byte,
+ }
+ p.state = .StringUtf8Byte3Of4;
+ },
+ .StringUtf8Byte3Of4 => switch (c) {
+ 0x80...0xBF => p.state = .StringUtf8Byte4Of4,
+ else => return error.InvalidUtf8Byte,
+ },
+ .StringUtf8Byte4Of4 => switch (c) {
+ 0x80...0xBF => p.state = .String,
else => return error.InvalidUtf8Byte,
},
diff --git a/lib/std/json/test.zig b/lib/std/json/test.zig
@@ -27,6 +27,20 @@ fn err(comptime s: []const u8) void {
} else |_| {}
}
+fn utf8Error(comptime s: []const u8) void {
+ std.testing.expect(!std.json.validate(s));
+
+ var mem_buffer: [1024 * 20]u8 = undefined;
+ const allocator = &std.heap.FixedBufferAllocator.init(&mem_buffer).allocator;
+ var p = std.json.Parser.init(allocator, false);
+
+ if (p.parse(s)) |_| {
+ unreachable;
+ } else |e| {
+ std.testing.expect(e == error.InvalidUtf8Byte);
+ }
+}
+
fn any(comptime s: []const u8) void {
_ = std.json.validate(s);
@@ -1936,3 +1950,55 @@ test "i_structure_UTF-8_BOM_empty_object" {
\\{}
);
}
+
+test "truncated UTF-8 sequence" {
+ utf8Error("\"\xc2\"");
+ utf8Error("\"\xdf\"");
+ utf8Error("\"\xed\xa0\"");
+ utf8Error("\"\xf0\x80\"");
+ utf8Error("\"\xf0\x80\x80\"");
+}
+
+test "invalid continuation byte" {
+ utf8Error("\"\xc2\x00\"");
+ utf8Error("\"\xc2\x7f\"");
+ utf8Error("\"\xc2\xc0\"");
+ utf8Error("\"\xc3\xc1\"");
+ utf8Error("\"\xc4\xf5\"");
+ utf8Error("\"\xc5\xff\"");
+ utf8Error("\"\xe4\x80\x00\"");
+ utf8Error("\"\xe5\x80\x10\"");
+ utf8Error("\"\xe6\x80\xc0\"");
+ utf8Error("\"\xe7\x80\xf5\"");
+ utf8Error("\"\xe8\x00\x80\"");
+ utf8Error("\"\xf2\x00\x80\x80\"");
+ utf8Error("\"\xf0\x80\x00\x80\"");
+ utf8Error("\"\xf1\x80\xc0\x80\"");
+ utf8Error("\"\xf2\x80\x80\x00\"");
+ utf8Error("\"\xf3\x80\x80\xc0\"");
+ utf8Error("\"\xf4\x80\x80\xf5\"");
+}
+
+test "disallowed overlong form" {
+ utf8Error("\"\xc0\x80\"");
+ utf8Error("\"\xc0\x90\"");
+ utf8Error("\"\xc1\x80\"");
+ utf8Error("\"\xc1\x90\"");
+ utf8Error("\"\xe0\x80\x80\"");
+ utf8Error("\"\xf0\x80\x80\x80\"");
+}
+
+test "out of UTF-16 range" {
+ utf8Error("\"\xf4\x90\x80\x80\"");
+ utf8Error("\"\xf5\x80\x80\x80\"");
+ utf8Error("\"\xf6\x80\x80\x80\"");
+ utf8Error("\"\xf7\x80\x80\x80\"");
+ utf8Error("\"\xf8\x80\x80\x80\"");
+ utf8Error("\"\xf9\x80\x80\x80\"");
+ utf8Error("\"\xfa\x80\x80\x80\"");
+ utf8Error("\"\xfb\x80\x80\x80\"");
+ utf8Error("\"\xfc\x80\x80\x80\"");
+ utf8Error("\"\xfd\x80\x80\x80\"");
+ utf8Error("\"\xfe\x80\x80\x80\"");
+ utf8Error("\"\xff\x80\x80\x80\"");
+}
