summaryrefslogtreecommitdiffstats
path: root/testing/web-platform/tests/fetch/corb/script-resource-with-json-parser-breaker.tentative.sub.html
blob: f0eb1f0ab1f30be45abda439ad987de7c78015ad (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
<!DOCTYPE html>
<meta charset="utf-8">
<!-- Test verifies CORB will block responses beginning with a JSON parser
  breaker regardless of their MIME type (excluding text/css - see below).

  A JSON parser breaker is a prefix added to resources with sensitive data to
  prevent cross-site script inclusion (XSSI) and similar attacks.  For example,
  it may be included in JSON files to prevent them from leaking data via a
  <script> tag, making the response only useful to a fetch or XmlHttpRequest.
  See also https://chromium.googlesource.com/chromium/src/+/main/services/network/cross_origin_read_blocking_explainer.md#Protecting-JSON

  The assumption is that all images, other media, scripts, fonts and other
  resources that may be embedded cross-origin will never begin with a JSON
  parser breaker.  For example an JPEG image should always being with FF D8 FF,
  a PNG image with 89 50 4E 47 0D 0A 1A 0A bytes and an SVG image with "<?xml"
  substring.

  The assumption above excludes text/css which (as shown by
  style-css-with-json-parser-breaker.sub.html) can parse as valid stylesheet
  even in presence of a JSON parser breaker.
-->
<script src="/resources/testharness.js"></script>
<script src="/resources/testharnessreport.js"></script>
<div id=log></div>
<script>
setup({allow_uncaught_exception : true});

// A subset of JSON security prefixes (only ones that are parser breakers).
json_parser_breakers = [
  ")]}'",
  "{}&&",
  "{} &&",
]

// JSON parser breaker should trigger CORB blocking for any Content-Type - even
// for resources that claim to be of a MIME type that is normally allowed to be
// embedded in cross-origin documents (like images and/or scripts).
mime_types = [
  // CORB-protected MIME types
  "text/html",
  "text/xml",
  "text/json",
  "text/plain",

  // MIME types that normally are allowed by CORB.
  "application/javascript",
  "image/png",
  "image/svg+xml",

  // Other types.
  "application/pdf",
  "application/zip",
]

function test(mime_type, body) {
  // The test below depends on a global/shared event handler - we need to ensure
  // that no tests run in parallel - this is achieved by using `promise_test`
  // instead of `async_test`.  See also
  // https://web-platform-tests.org/writing-tests/testharness-api.html#promise-tests
  promise_test(t => new Promise(function(resolve, reject) {
    var script = document.createElement("script")

    // Without CORB, the JSON parser breaker would cause a syntax error when
    // parsed as JavaScript, but with CORB there should be no errors (because
    // CORB will replace the response body with an empty body). With ORB,
    // the script loading itself should error out.
    script.onload = resolve;
    script.onerror = resolve;
    addEventListener("error", t.unreached_func(
        "Empty body of a CORS-blocked response shouldn't trigger syntax errors."))

    // www1 is cross-origin, so the HTTP response is CORB-eligible.
    var src_prefix = "http://{{domains[www1]}}:{{ports[http][0]}}/fetch/corb/resources/sniffable-resource.py";
    script.src = src_prefix + "?type=" + mime_type + "&body=" + encodeURIComponent(body);
    document.body.appendChild(script)
  }), "CORB-blocks '" + mime_type + "' that starts with the following JSON parser breaker: " + body);
}

mime_types.forEach(function(type) {
    json_parser_breakers.forEach(function(body) {
        test(type, body);
    });
});

</script>