platform-tests

urltestdata.txt

# FORMAT NOT DOCUMENTED YET (parser is urltestparser.js)

# Based on http://trac.webkit.org/browser/trunk/LayoutTests/fast/url/script-tests/segments.js
http://example\t.\norg http://example.org/foo/bar s:http h:example.org p:/
http://user:pass@foo:21/bar;par?b#c  s:http u:user pass:pass h:foo port:21 p:/bar;par q:?b f:#c
http:foo.com  s:http h:example.org p:/foo/foo.com
\t\s\s\s:foo.com\s\s\s\n  s:http h:example.org p:/foo/:foo.com
\sfoo.com\s\s  s:http h:example.org p:/foo/foo.com
a:\t\sfoo.com  s:a p:\sfoo.com
http://f:21/\sb\s?\sd\s#\se\s  s:http h:f port:21 p:/%20b%20 q:?%20d%20 f:#\se
http://f:/c  s:http h:f p:/c
http://f:0/c  s:http h:f port:0 p:/c
http://f:00000000000000/c  s:http h:f port:0 p:/c
http://f:00000000000000000000080/c  s:http h:f p:/c
http://f:b/c
http://f:\s/c
http://f:\n/c  s:http h:f p:/c
http://f:fifty-two/c
http://f:999999/c  s:http h:f port:999999 p:/c
http://f:\s21\s/\sb\s?\sd\s#\se\s
  s:http h:example.org p:/foo/bar
\s\s\t  s:http h:example.org p:/foo/bar
:foo.com/  s:http h:example.org p:/foo/:foo.com/
:foo.com\\  s:http h:example.org p:/foo/:foo.com/
:  s:http h:example.org p:/foo/:
:a  s:http h:example.org p:/foo/:a
:/  s:http h:example.org p:/foo/:/
:\\  s:http h:example.org p:/foo/:/
:#  s:http h:example.org p:/foo/: f:#
\#  s:http h:example.org p:/foo/bar f:#
\#/  s:http h:example.org p:/foo/bar f:#/
\#\\  s:http h:example.org p:/foo/bar f:#\\
\#;?  s:http h:example.org p:/foo/bar f:#;?
?  s:http h:example.org p:/foo/bar q:?
/  s:http h:example.org p:/
:23  s:http h:example.org p:/foo/:23
/:23  s:http h:example.org p:/:23
::  s:http h:example.org p:/foo/::
::23  s:http h:example.org p:/foo/::23
foo://  s:foo p://
http://a:b@c:29/d  s:http u:a pass:b h:c port:29 p:/d
http::@c:29  s:http h:example.org p:/foo/:@c:29
http://&a:foo(b]c@d:2/  s:http u:&a pass:foo(b]c h:d port:2 p:/
http://::@c@d:2  s:http pass::%40c h:d port:2 p:/
http://foo.com:b@d/  s:http u:foo.com pass:b h:d p:/
http://foo.com/\\@  s:http h:foo.com p://@
http:\\\\foo.com\\  s:http h:foo.com p:/
http:\\\\a\\b:c\\d@foo.com\\  s:http h:a p:/b:c/d@foo.com/
foo:/  s:foo p:/
foo:/bar.com/  s:foo p:/bar.com/
foo://///////  s:foo p://///////
foo://///////bar.com/  s:foo p://///////bar.com/
foo:////://///  s:foo p:////://///
c:/foo  s:c p:/foo
//foo/bar  s:http h:foo p:/bar
http://foo/path;a??e#f#g  s:http h:foo p:/path;a q:??e f:#f#g
http://foo/abcd?efgh?ijkl  s:http h:foo p:/abcd q:?efgh?ijkl
http://foo/abcd#foo?bar  s:http h:foo p:/abcd f:#foo?bar
[61:24:74]:98  s:http h:example.org p:/foo/[61:24:74]:98
http:[61:27]/:foo  s:http h:example.org p:/foo/[61:27]/:foo
http://[1::2]:3:4
http://2001::1
http://2001::1]
http://2001::1]:80
http://[2001::1]  s:http h:[2001::1] p:/
http://[2001::1]:80  s:http h:[2001::1] p:/
http:/example.com/  s:http h:example.org p:/example.com/
ftp:/example.com/  s:ftp h:example.com p:/
https:/example.com/  s:https h:example.com p:/
madeupscheme:/example.com/  s:madeupscheme p:/example.com/
file:/example.com/  s:file p:/example.com/
ftps:/example.com/  s:ftps p:/example.com/
gopher:/example.com/  s:gopher h:example.com p:/
ws:/example.com/  s:ws h:example.com p:/
wss:/example.com/  s:wss h:example.com p:/
data:/example.com/  s:data p:/example.com/
javascript:/example.com/  s:javascript p:/example.com/
mailto:/example.com/  s:mailto p:/example.com/
http:example.com/  s:http h:example.org p:/foo/example.com/
ftp:example.com/  s:ftp h:example.com p:/
https:example.com/  s:https h:example.com p:/
madeupscheme:example.com/  s:madeupscheme p:example.com/
ftps:example.com/  s:ftps p:example.com/
gopher:example.com/  s:gopher h:example.com p:/
ws:example.com/  s:ws h:example.com p:/
wss:example.com/  s:wss h:example.com p:/
data:example.com/  s:data p:example.com/
javascript:example.com/  s:javascript p:example.com/
mailto:example.com/  s:mailto p:example.com/
/a/b/c  s:http h:example.org p:/a/b/c
/a/\s/c  s:http h:example.org p:/a/%20/c
/a%2fc  s:http h:example.org p:/a%2fc
/a/%2f/c  s:http h:example.org p:/a/%2f/c
\#\u03B2  s:http h:example.org p:/foo/bar f:#\u03B2
data:text/html,test#test  s:data p:text/html,test f:#test

# Based on http://trac.webkit.org/browser/trunk/LayoutTests/fast/url/file.html

# Basic canonicalization, uppercase should be converted to lowercase
file:c:\\foo\\bar.html file:///tmp/mock/path s:file p:/c:/foo/bar.html

# Spaces should fail
\s\sFile:c|////foo\\bar.html  s:file p:/c:////foo/bar.html

# This should fail
C|/foo/bar  s:file p:/C:/foo/bar

# This should fail
/C|\\foo\\bar  s:file p:/C:/foo/bar
//C|/foo/bar  s:file p:/C:/foo/bar
//server/file  s:file h:server p:/file
\\\\server\\file  s:file h:server p:/file
/\\server/file  s:file h:server p:/file
file:///foo/bar.txt  s:file p:/foo/bar.txt
file:///home/me  s:file p:/home/me
//  s:file p:/
///  s:file p:/
///test  s:file p:/test
file://test  s:file h:test p:/
file://localhost  s:file h:localhost p:/
file://localhost/  s:file h:localhost p:/
file://localhost/test  s:file h:localhost p:/test
test  s:file p:/tmp/mock/test
file:test  s:file p:/tmp/mock/test

# Based on http://trac.webkit.org/browser/trunk/LayoutTests/fast/url/script-tests/path.js
http://example.com/././foo about:blank s:http h:example.com p:/foo
http://example.com/./.foo  s:http h:example.com p:/.foo
http://example.com/foo/.  s:http h:example.com p:/foo/
http://example.com/foo/./  s:http h:example.com p:/foo/
http://example.com/foo/bar/..  s:http h:example.com p:/foo/
http://example.com/foo/bar/../  s:http h:example.com p:/foo/
http://example.com/foo/..bar  s:http h:example.com p:/foo/..bar
http://example.com/foo/bar/../ton  s:http h:example.com p:/foo/ton
http://example.com/foo/bar/../ton/../../a  s:http h:example.com p:/a
http://example.com/foo/../../..  s:http h:example.com p:/
http://example.com/foo/../../../ton  s:http h:example.com p:/ton
http://example.com/foo/%2e  s:http h:example.com p:/foo/
http://example.com/foo/%2e%2  s:http h:example.com p:/foo/%2e%2
http://example.com/foo/%2e./%2e%2e/.%2e/%2e.bar  s:http h:example.com p:/%2e.bar
http://example.com////../..  s:http h:example.com p://
http://example.com/foo/bar//../..  s:http h:example.com p:/foo/
http://example.com/foo/bar//..  s:http h:example.com p:/foo/bar/
http://example.com/foo  s:http h:example.com p:/foo
http://example.com/%20foo  s:http h:example.com p:/%20foo
http://example.com/foo%  s:http h:example.com p:/foo%
http://example.com/foo%2  s:http h:example.com p:/foo%2
http://example.com/foo%2zbar  s:http h:example.com p:/foo%2zbar
http://example.com/foo%2\u00C2\u00A9zbar  s:http h:example.com p:/foo%2%C3%82%C2%A9zbar
http://example.com/foo%41%7a  s:http h:example.com p:/foo%41%7a
http://example.com/foo\t\u0091%91  s:http h:example.com p:/foo%C2%91%91
http://example.com/foo%00%51  s:http h:example.com p:/foo%00%51
http://example.com/(%28:%3A%29)  s:http h:example.com p:/(%28:%3A%29)
http://example.com/%3A%3a%3C%3c  s:http h:example.com p:/%3A%3a%3C%3c
http://example.com/foo\tbar  s:http h:example.com p:/foobar
http://example.com\\\\foo\\\\bar  s:http h:example.com p://foo//bar
http://example.com/%7Ffp3%3Eju%3Dduvgw%3Dd  s:http h:example.com p:/%7Ffp3%3Eju%3Dduvgw%3Dd
http://example.com/@asdf%40  s:http h:example.com p:/@asdf%40
http://example.com/\u4F60\u597D\u4F60\u597D  s:http h:example.com p:/%E4%BD%A0%E5%A5%BD%E4%BD%A0%E5%A5%BD
http://example.com/\u2025/foo  s:http h:example.com p:/%E2%80%A5/foo
http://example.com/\uFEFF/foo  s:http h:example.com p:/%EF%BB%BF/foo
http://example.com/\u202E/foo/\u202D/bar  s:http h:example.com p:/%E2%80%AE/foo/%E2%80%AD/bar

# Based on http://trac.webkit.org/browser/trunk/LayoutTests/fast/url/script-tests/relative.js
http://www.google.com/foo?bar=baz# about:blank s:http h:www.google.com p:/foo q:?bar=baz f:#
http://www.google.com/foo?bar=baz#\s\u00BB  s:http h:www.google.com p:/foo q:?bar=baz f:#\s\u00BB
data:test#\s\u00BB  s:data p:test f:#\s\u00BB
http://[www.google.com]/
http://www.google.com  s:http h:www.google.com p:/
http://192.0x00A80001  s:http h:192.168.0.1 p:/
http://www/foo%2Ehtml  s:http h:www p:/foo%2Ehtml
http://www/foo/%2E/html  s:http h:www p:/foo/html
http://user:pass@/
http://%25DOMAIN:foobar@foodomain.com/  s:http u:%25DOMAIN pass:foobar h:foodomain.com p:/
http:\\\\www.google.com\\foo  s:http h:www.google.com p:/foo
http://foo:80/  s:http h:foo p:/
http://foo:81/  s:http h:foo port:81 p:/
httpa://foo:80/  s:httpa p://foo:80/
http://foo:-80/
https://foo:443/  s:https h:foo p:/
https://foo:80/  s:https h:foo port:80 p:/
ftp://foo:21/  s:ftp h:foo p:/
ftp://foo:80/  s:ftp h:foo port:80 p:/
gopher://foo:70/  s:gopher h:foo p:/
gopher://foo:443/  s:gopher h:foo port:443 p:/
ws://foo:80/  s:ws h:foo p:/
ws://foo:81/  s:ws h:foo port:81 p:/
ws://foo:443/  s:ws h:foo port:443 p:/
ws://foo:815/  s:ws h:foo port:815 p:/
wss://foo:80/  s:wss h:foo port:80 p:/
wss://foo:81/  s:wss h:foo port:81 p:/
wss://foo:443/  s:wss h:foo p:/
wss://foo:815/  s:wss h:foo port:815 p:/
http:/example.com/  s:http h:example.com p:/
ftp:/example.com/  s:ftp h:example.com p:/
https:/example.com/  s:https h:example.com p:/
madeupscheme:/example.com/  s:madeupscheme p:/example.com/
file:/example.com/  s:file p:/example.com/
ftps:/example.com/  s:ftps p:/example.com/
gopher:/example.com/  s:gopher h:example.com p:/
ws:/example.com/  s:ws h:example.com p:/
wss:/example.com/  s:wss h:example.com p:/
data:/example.com/  s:data p:/example.com/
javascript:/example.com/  s:javascript p:/example.com/
mailto:/example.com/  s:mailto p:/example.com/
http:example.com/  s:http h:example.com p:/
ftp:example.com/  s:ftp h:example.com p:/
https:example.com/  s:https h:example.com p:/
madeupscheme:example.com/  s:madeupscheme p:example.com/
ftps:example.com/  s:ftps p:example.com/
gopher:example.com/  s:gopher h:example.com p:/
ws:example.com/  s:ws h:example.com p:/
wss:example.com/  s:wss h:example.com p:/
data:example.com/  s:data p:example.com/
javascript:example.com/  s:javascript p:example.com/
mailto:example.com/  s:mailto p:example.com/

# Based on http://trac.webkit.org/browser/trunk/LayoutTests/fast/url/segments-userinfo-vs-host.html
http:@www.example.com about:blank s:http h:www.example.com p:/
http:/@www.example.com  s:http h:www.example.com p:/
http://@www.example.com  s:http h:www.example.com p:/
http:a:b@www.example.com  s:http u:a pass:b h:www.example.com p:/
http:/a:b@www.example.com  s:http u:a pass:b h:www.example.com p:/
http://a:b@www.example.com  s:http u:a pass:b h:www.example.com p:/
http://@pple.com  s:http h:pple.com p:/
http::b@www.example.com  s:http pass:b h:www.example.com p:/
http:/:b@www.example.com  s:http pass:b h:www.example.com p:/
http://:b@www.example.com  s:http pass:b h:www.example.com p:/
http:/:@/www.example.com
http://user@/www.example.com
http:@/www.example.com
http:/@/www.example.com
http://@/www.example.com
https:@/www.example.com
http:a:b@/www.example.com
http:/a:b@/www.example.com
http://a:b@/www.example.com
http::@/www.example.com
http:a:@www.example.com  s:http u:a pass: h:www.example.com p:/
http:/a:@www.example.com  s:http u:a pass: h:www.example.com p:/
http://a:@www.example.com  s:http u:a pass: h:www.example.com p:/
http://www.@pple.com  s:http u:www. h:pple.com p:/
http:@:www.example.com
http:/@:www.example.com
http://@:www.example.com
http://:@www.example.com  s:http pass: h:www.example.com p:/

#Others
/ http://www.example.com/test s:http h:www.example.com p:/
/test.txt  s:http h:www.example.com p:/test.txt
.  s:http h:www.example.com p:/
..  s:http h:www.example.com p:/
test.txt  s:http h:www.example.com p:/test.txt
./test.txt  s:http h:www.example.com p:/test.txt
../test.txt  s:http h:www.example.com p:/test.txt
../aaa/test.txt  s:http h:www.example.com p:/aaa/test.txt
../../test.txt  s:http h:www.example.com p:/test.txt
\u4E2D/test.txt  s:http h:www.example.com p:/%E4%B8%AD/test.txt
http://www.example2.com  s:http h:www.example2.com p:/
//www.example2.com  s:http h:www.example2.com p:/

# Based on http://trac.webkit.org/browser/trunk/LayoutTests/fast/url/host.html

# Basic canonicalization, uppercase should be converted to lowercase
http://ExAmPlE.CoM http://other.com/ s:http p:/ h:example.com

# Spaces should fail
http://example\sexample.com

# This should fail
http://Goo%20\sgoo%7C|.com

# U+3000 is mapped to U+0020 (space) which is disallowed
http://GOO\u00a0\u3000goo.com

# Other types of space (no-break, zero-width, zero-width-no-break) are
# name-prepped away to nothing.
# U+200B, U+2060, and U+FEFF, are ignored
http://GOO\u200b\u2060\ufeffgoo.com  s:http p:/ h:googoo.com

# Ideographic full stop (full-width period for Chinese, etc.) should be
# treated as a dot.
# U+3002 is mapped to U+002E (dot)
http://www.foo\u3002bar.com  s:http p:/ h:www.foo.bar.com

# Invalid unicode characters should fail...
# U+FDD0 is disallowed; %ef%b7%90 is U+FDD0
http://\ufdd0zyx.com

# ...This is the same as previous but escaped.
http://%ef%b7%90zyx.com

# Test name prepping, fullwidth input should be converted to ASCII and NOT
# IDN-ized. This is "Go" in fullwidth UTF-8/UTF-16.
http://\uff27\uff4f.com  s:http p:/ h:go.com

# URL spec forbids the following.
# https://www.w3.org/Bugs/Public/show_bug.cgi?id=24257
http://\uff05\uff14\uff11.com
http://%ef%bc%85%ef%bc%94%ef%bc%91.com

# ...%00 in fullwidth should fail (also as escaped UTF-8 input)
http://\uff05\uff10\uff10.com
http://%ef%bc%85%ef%bc%90%ef%bc%90.com

# Basic IDN support, UTF-8 and UTF-16 input should be converted to IDN
http://\u4f60\u597d\u4f60\u597d  s:http p:/ h:xn--6qqa088eba

# Invalid escaped characters should fail and the percents should be
# escaped. https://www.w3.org/Bugs/Public/show_bug.cgi?id=24191
http://%zz%66%a.com

# If we get an invalid character that has been escaped.
http://%25
http://hello%00

# Escaped numbers should be treated like IP addresses if they are.
# No special handling for IPv4 or IPv4-like URLs
http://%30%78%63%30%2e%30%32%35%30.01  s:http p:/ h:192.168.0.1
http://%30%78%63%30%2e%30%32%35%30.01%2e  s:http p:/ h:0xc0.0250.01.
http://192.168.0.257

# Invalid escaping should trigger the regular host error handling.
http://%3g%78%63%30%2e%30%32%35%30%2E.01

# Something that isn't exactly an IP should get treated as a host and
# spaces escaped.
http://192.168.0.1\shello

# Fullwidth and escaped UTF-8 fullwidth should still be treated as IP.
# These are "0Xc0.0250.01" in fullwidth.
http://\uff10\uff38\uff43\uff10\uff0e\uff10\uff12\uff15\uff10\uff0e\uff10\uff11  s:http p:/ h:192.168.0.1

# Broken IPv6
http://[google.com]

# Misc Unicode
http://foo:\uD83D\uDCA9@example.com/bar  s:http h:example.com p:/bar u:foo pass:%F0%9F%92%A9

# resolving a relative reference against an unknown scheme results in an error
x test:test

urltestparser.js

function URLTestParser(input) {
  var relativeSchemes = ["ftp", "file", "gopher", "http", "https", "ws", "wss"],
      tokenMap = { "\\": "\\", "#": "#", n: "\n", r: "\r", s: " ", t: "\t", f: "\f" }
      resultMap = { s: "scheme", u: "username", pass: "password", h: "host", port: "port", p: "path", q: "query", f: "fragment" },
      results = []
  function Test() {
    this.input = ""
    this.base = ""
    this.scheme = ""
    this.username = ""
    this.password = null
    this.host = ""
    this.port = ""
    this.path = ""
    this.query = ""
    this.fragment = ""
    Object.defineProperties(this, {
      "href": { get: function() {
        return !this.scheme ? this.input : this.protocol + (
          relativeSchemes.indexOf(this.scheme) != -1 ? "//" + (
            ("" != this.username || null != this.password) ? this.username + (
              null != this.password ? ":" + this.password : ""
            ) + "@" : ""
          ) + this.host : ""
        ) + (this.port ? ":" + this.port : "") + this.path + this.query + this.fragment
      } },
      "protocol": { get: function() { return this.scheme + ":" } },
      "search": { get: function() { return "?" == this.query ? "" : this.query } },
      "hash": { get: function() { return "#" == this.fragment ? "" : this.fragment } }
    })
  }
  function normalize(input) {
    var output = ""
    for(var i = 0, l = input.length; i < l; i++) {
      var c = input[i]
      if(c == "\\") {
        var nextC = input[++i]
        if(tokenMap.hasOwnProperty(nextC)) {
          output += tokenMap[nextC]
        } else if(nextC == "u") {
          output += String.fromCharCode(parseInt(input[++i] + input[++i] + input[++i] + input[++i], 16))
        } else {
          throw new Error("Input is invalid.")
        }
      } else {
        output += c
      }
    }
    return output
  }
  var lines = input.split("\n")
  for(var i = 0, l = lines.length; i < l; i++) {
    var line = lines[i]
    if(line === "" || line.indexOf("#", 0) === 0) {
      continue
    }
    var pieces = line.split(" "),
        result = new Test()
    result.input = normalize(pieces.shift())
    var base = pieces.shift()
    if(base === "" || base === undefined) {
      result.base = results[results.length - 1].base
    } else {
      result.base = normalize(base)
    }
    for(var ii = 0, ll = pieces.length; ii < ll; ii++) {
      var piece = pieces[ii]
      if(piece.indexOf("#", 0) === 0) {
        continue
      }
      var subpieces = piece.split(":"),
          token = subpieces.shift()
          value = subpieces.join(":")
      result[resultMap[token]] = normalize(value)
    }
    results.push(result)
  }
  return results
}
module.exports = URLTestParser