Avoid decoding XHR type="json" responses as UTF-16BE/LE

https://infra.spec.whatwg.org/#parse-json-from-bytes says to use
"UTF-8 decode" rather than "decode", so UTF-16BE/LE BOM should
not be honored.
This commit is contained in:
Henri Sivonen 2017-11-01 16:50:53 +02:00
parent edb2db55b7
commit 18a52ea0e9
4 changed files with 46 additions and 3 deletions

View file

@ -1164,8 +1164,8 @@ impl XMLHttpRequest {
return NullValue(); return NullValue();
} }
// Step 4 // Step 4
fn decode_to_utf16(bytes: &[u8], encoding: &'static Encoding) -> Vec<u16> { fn decode_to_utf16_with_bom_removal(bytes: &[u8], encoding: &'static Encoding) -> Vec<u16> {
let mut decoder = encoding.new_decoder(); let mut decoder = encoding.new_decoder_with_bom_removal();
let capacity = decoder.max_utf16_buffer_length(bytes.len()).expect("Overflow"); let capacity = decoder.max_utf16_buffer_length(bytes.len()).expect("Overflow");
let mut utf16 = Vec::with_capacity(capacity); let mut utf16 = Vec::with_capacity(capacity);
let extra = unsafe { let extra = unsafe {
@ -1179,7 +1179,12 @@ impl XMLHttpRequest {
} }
utf16 utf16
} }
let json_text = decode_to_utf16(&bytes, UTF_8); // https://xhr.spec.whatwg.org/#json-response refers to
// https://infra.spec.whatwg.org/#parse-json-from-bytes which refers to
// https://encoding.spec.whatwg.org/#utf-8-decode which means
// that the encoding is always UTF-8 and the UTF-8 BOM is removed,
// if present, but UTF-16BE/LE BOM must not be honored.
let json_text = decode_to_utf16_with_bom_removal(&bytes, UTF_8);
// Step 5 // Step 5
rooted!(in(cx) let mut rval = UndefinedValue()); rooted!(in(cx) let mut rval = UndefinedValue());
unsafe { unsafe {

View file

@ -180079,6 +180079,11 @@
{} {}
] ]
], ],
"XMLHttpRequest/resources/utf16-bom.json": [
[
{}
]
],
"XMLHttpRequest/resources/utf16.txt": [ "XMLHttpRequest/resources/utf16.txt": [
[ [
{} {}
@ -293760,6 +293765,12 @@
{} {}
] ]
], ],
"XMLHttpRequest/no-utf16-json.htm": [
[
"/XMLHttpRequest/no-utf16-json.htm",
{}
]
],
"XMLHttpRequest/open-after-abort.htm": [ "XMLHttpRequest/open-after-abort.htm": [
[ [
"/XMLHttpRequest/open-after-abort.htm", "/XMLHttpRequest/open-after-abort.htm",
@ -377781,6 +377792,10 @@
"6804845b3ba0e52ee407fc7e8036ce905283751c", "6804845b3ba0e52ee407fc7e8036ce905283751c",
"testharness" "testharness"
], ],
"XMLHttpRequest/no-utf16-json.htm": [
"4546215ad6f138676d3c473782e6e34b4c922589",
"testharness"
],
"XMLHttpRequest/open-after-abort.htm": [ "XMLHttpRequest/open-after-abort.htm": [
"082fa646606cf8f278d61960f02fafa264e57e9f", "082fa646606cf8f278d61960f02fafa264e57e9f",
"testharness" "testharness"
@ -378273,6 +378288,10 @@
"e1addc2a9f014c2546b5770dd328b1562dc4fdc3", "e1addc2a9f014c2546b5770dd328b1562dc4fdc3",
"support" "support"
], ],
"XMLHttpRequest/resources/utf16-bom.json": [
"3d344828308b32594a8ed51a3cf186ccfa83cfb1",
"support"
],
"XMLHttpRequest/resources/utf16.txt": [ "XMLHttpRequest/resources/utf16.txt": [
"47e95b463051a904934ec51df445a39301c5f671", "47e95b463051a904934ec51df445a39301c5f671",
"support" "support"

View file

@ -0,0 +1,19 @@
<!doctype html>
<meta charset=utf-8>
<title>XMLHttpRequest: responseType "json" should be decoded as UTF-8 even when there's a UTF-16 BOM</title>
<script src="/resources/testharness.js"></script>
<script src="/resources/testharnessreport.js"></script>
<link rel="help" href="https://infra.spec.whatwg.org/#parse-json-from-bytes" data-tested-assertations="following::ol/li[1]" />
<div id="log"></div>
<script>
async_test(function() {
var client = new XMLHttpRequest();
client.responseType = 'json';
client.onload = this.step_func_done(function(e) {
assert_equals(client.response, null);
});
client.open("GET", "resources/utf16-bom.json");
client.send(null);
}, 'JSON asa UTF-16 with BOM should decode as UTF-8 (and, therefore, fail)');
</script>