Skip to content

Commit 6b9a6fb

Browse files
Allow more characters in element/attribute names and prefixes
This patch significantly changes the parsing of element names, attribute names, and namespace prefixes for DOM APIs to allow more flexibility and better parity with the HTML parser. I am planning on making an intent to ship for this behavior before enabling it by default. I am planning on making another WPT patch to change the existing tests to match the new parsing behavior once the spec is merged and the I2S is complete, and maybe also after the new behavior reaches stable with no issues. Spec PR: whatwg/dom#1079 Bug: 40122442, 40228234 Change-Id: Ifbb5ac47a08a8f14489c694649ab5be1f59647ac Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/4251683 Commit-Queue: Joey Arhar <[email protected]> Reviewed-by: Mason Freed <[email protected]> Cr-Commit-Position: refs/heads/main@{#1468337}
1 parent 1329b91 commit 6b9a6fb

File tree

1 file changed

+286
-0
lines changed

1 file changed

+286
-0
lines changed
Lines changed: 286 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,286 @@
1+
<!DOCTYPE html>
2+
<meta name=timeout content=long>
3+
<link rel=author href="mailto:[email protected]">
4+
<link rel=help href="https://github.com/whatwg/dom/pull/1079">
5+
<script src="/resources/testharness.js"></script>
6+
<script src="/resources/testharnessreport.js"></script>
7+
8+
<script>
9+
function isAsciiAlpha(codePoint) {
10+
return (codePoint >= 0x41 && codePoint <= 0x5A) || (codePoint >= 0x61 && codePoint <= 0x7A);
11+
}
12+
function isAsciiDigit(codePoint) {
13+
return codePoint >= 0x30 && codePoint <= 0x39;
14+
}
15+
function isAsciiWhitespace(codePoint) {
16+
return codePoint == 0x9 || codePoint == 0xA || codePoint == 0xC || codePoint == 0xD || codePoint == 0x20;
17+
}
18+
19+
function debugString(str) {
20+
const codePoints = [];
21+
for (let i = 0; i < str.length; i++) {
22+
codePoints.push(str.charCodeAt(i));
23+
}
24+
return `code points: ${JSON.stringify(codePoints)}, string: "${str}"`;
25+
}
26+
27+
const latin1CodePoint = 100;
28+
const latin1 = String.fromCodePoint(latin1CodePoint);
29+
const smallEmoji = 'smallEmoji🆖';
30+
const bigEmoji = 'bigEmoji🅱️';
31+
32+
// Testing every variation of a namespace prefix with every variation of a
33+
// local name would make the test take too long to run, so use these instead when
34+
// combining with a namespace prefix.
35+
const validElementLocalNamesShortened = [
36+
'div', `latin1${latin1}`, smallEmoji, bigEmoji
37+
];
38+
const invalidElementLocalNamesShortened = [
39+
'', 'space ', 'newline\n', 'null\0', `:soh${String.fromCodePoint(1)}`, '5'
40+
];
41+
const validAttributeLocalNamesShortened = [
42+
'attr', `latin1${latin1}`, smallEmoji, bigEmoji
43+
];
44+
const invalidAttributeLocalNamesShortened = [
45+
'', 'space ', 'newline\n', 'null\0'
46+
];
47+
48+
const validElementLocalNames = validElementLocalNamesShortened.slice();
49+
const invalidElementLocalNames = invalidElementLocalNamesShortened.slice();
50+
const validAttributeLocalNames = validAttributeLocalNamesShortened.slice();
51+
const invalidAttributeLocalNames = invalidAttributeLocalNamesShortened.slice();
52+
const validNamespacePrefixes = ['', smallEmoji, bigEmoji];
53+
const invalidNamespacePrefixes = [];
54+
const validDoctypes = [''];
55+
const invalidDoctypes = [];
56+
57+
const codePoints = [];
58+
for (let i = 0; i < 0x80; i++) {
59+
codePoints.push(i);
60+
}
61+
codePoints.push(latin1CodePoint);
62+
63+
// attributes and namespaces
64+
for (const codePoint of codePoints) {
65+
const str = String.fromCodePoint(codePoint);
66+
if (codePoint == 0 || isAsciiWhitespace(codePoint) || codePoint == 0x2F || codePoint == 0x3E) {
67+
invalidNamespacePrefixes.push(str);
68+
invalidAttributeLocalNames.push(str);
69+
} else if (codePoint == 0x3A) {
70+
// colons are not valid namespace prefixes, but due to parsing they can
71+
// never be considered as a namespace prefix, only as a separator between the
72+
// prefix and the local name.
73+
validAttributeLocalNames.push(str);
74+
} else if (codePoint == 0x3D) {
75+
validNamespacePrefixes.push(str);
76+
invalidAttributeLocalNames.push(str);
77+
} else {
78+
validNamespacePrefixes.push(str);
79+
validAttributeLocalNames.push(str);
80+
}
81+
}
82+
83+
// valid element local names
84+
for (const firstChar of codePoints) {
85+
for (const secondChar of codePoints) {
86+
const str = `${String.fromCodePoint(firstChar)}${String.fromCodePoint(secondChar)}`;
87+
if (isAsciiAlpha(firstChar)) {
88+
if (!secondChar || secondChar == 0x2F || secondChar == 0x3E || isAsciiWhitespace(secondChar)) {
89+
invalidElementLocalNames.push(str);
90+
} else {
91+
validElementLocalNames.push(str);
92+
}
93+
} else {
94+
if (firstChar == 0x3A || firstChar == 0x5F || firstChar >= 0x80) {
95+
if (isAsciiAlpha(secondChar) ||
96+
isAsciiDigit(secondChar) ||
97+
secondChar == 0x2D ||
98+
secondChar == 0x2E ||
99+
secondChar == 0x3A ||
100+
secondChar == 0x5F ||
101+
secondChar >= 0x80) {
102+
validElementLocalNames.push(str);
103+
} else {
104+
invalidElementLocalNames.push(str);
105+
}
106+
} else {
107+
invalidElementLocalNames.push(str);
108+
}
109+
}
110+
}
111+
}
112+
113+
// doctypes
114+
for (const codePoint of codePoints) {
115+
const str = String.fromCodePoint(codePoint);
116+
if (codePoint == 0 || isAsciiWhitespace(codePoint) || codePoint == 0x3E) {
117+
invalidDoctypes.push(str);
118+
} else {
119+
validDoctypes.push(str);
120+
}
121+
}
122+
123+
test(() => {
124+
// This regex is provided in the spec and is used here to double check our
125+
// test input.
126+
const validNameRegex = /^(?:[A-Za-z][^\0\t\n\f\r\u0020/>]*|[:_\u0080-\u{10FFFF}][A-Za-z0-9-.:_\u0080-\u{10FFFF}]*)$/u;
127+
for (const validName of validElementLocalNames) {
128+
assert_true(
129+
validNameRegex.test(validName),
130+
`Regex should match: ${debugString(validName)}`);
131+
try {
132+
document.createElement(validName);
133+
} catch (error) {
134+
assert_unreached(
135+
`document.createElement should not have thrown an error for: ${debugString(validName)} ${error.toString()}`);
136+
}
137+
}
138+
for (const invalidName of invalidElementLocalNames) {
139+
assert_false(
140+
validNameRegex.test(invalidName),
141+
`Regex should not match: ${debugString(invalidName)}`);
142+
assert_throws_dom(
143+
'InvalidCharacterError',
144+
() => document.createElement(invalidName),
145+
`document.createElement should throw an error for: ${debugString(invalidName)}`);
146+
}
147+
}, 'Valid and invalid characters in createElement.');
148+
149+
test(() => {
150+
for (const validNamespace of validNamespacePrefixes) {
151+
for (const validName of validElementLocalNamesShortened) {
152+
try {
153+
document.createElementNS('namespaceuri', `${validNamespace}:${validName}`);
154+
} catch (error) {
155+
assert_unreached(
156+
`document.createElementNS should not have thrown an error for: ${debugString(validNamespace)} ${debugString(validName)} ${error.toString()}`);
157+
}
158+
try {
159+
document.implementation.createDocument('namespaceuri', `${validNamespace}:${validName}`);
160+
} catch (error) {
161+
assert_unreached(
162+
`createDocument should not have thrown an error for: ${debugString(validNamespace)} ${debugString(validName)} ${error.toString()}`);
163+
}
164+
}
165+
for (const invalidName of invalidElementLocalNamesShortened) {
166+
assert_throws_dom(
167+
'InvalidCharacterError',
168+
() => document.createElementNS('namespaceuri', `${validNamespace}:${invalidName}`),
169+
`document.createElementNS should throw an error for: ${debugString(validNamespace)} ${debugString(invalidName)}`);
170+
assert_throws_dom(
171+
'InvalidCharacterError',
172+
() => document.implementation.createDocument('namespaceuri', `${validNamespace}:${invalidName}`),
173+
`createDocument should throw an error for: ${debugString(validNamespace)} ${debugString(invalidName)}`);
174+
}
175+
}
176+
for (const invalidNamespace of invalidNamespacePrefixes) {
177+
for (const localName of validElementLocalNamesShortened.concat(invalidElementLocalNamesShortened)) {
178+
assert_throws_dom(
179+
'InvalidCharacterError',
180+
() => document.createElementNS('namespaceuri', `${invalidNamespace}:${localName}`),
181+
`document.createElementNS should throw an error for: ${debugString(invalidNamespace)} ${debugString(localName)}`);
182+
assert_throws_dom(
183+
'InvalidCharacterError',
184+
() => document.implementation.createDocument('namespaceuri', `${invalidNamespace}:${localName}`),
185+
`createDocument should throw an error for: ${debugString(invalidNamespace)} ${debugString(localName)}`);
186+
}
187+
}
188+
}, 'Valid and invalid characters in createElementNS and createDocument.');
189+
190+
test(() => {
191+
for (const validAttributeName of validAttributeLocalNames) {
192+
const element = document.createElement('div');
193+
try {
194+
element.setAttribute(validAttributeName, 'value');
195+
} catch (error) {
196+
assert_unreached(
197+
`element.setAttribute should not have thrown an error for: ${debugString(validAttributeName)} ${error.toString()}`);
198+
}
199+
try {
200+
element.toggleAttribute(validAttributeName);
201+
} catch (error) {
202+
assert_unreached(
203+
`element.toggleAttribute should not have thrown an error for: ${debugString(validAttributeName)} ${error.toString()}`);
204+
}
205+
try {
206+
document.createAttribute(validAttributeName);
207+
} catch (error) {
208+
assert_unreached(
209+
`document.createAttribute should not have thrown an error for: ${debugString(validAttributeName)} ${error.toString()}`);
210+
}
211+
}
212+
for (const invalidAttributeName of invalidAttributeLocalNames) {
213+
const element = document.createElement('div');
214+
assert_throws_dom(
215+
'InvalidCharacterError',
216+
() => element.setAttribute(invalidAttributeName, 'value'),
217+
`element.setAttribute should throw an error for: ${debugString(invalidAttributeName)}`);
218+
assert_throws_dom(
219+
'InvalidCharacterError',
220+
() => element.toggleAttribute(invalidAttributeName),
221+
`element.toggleAttribute should throw an error for: ${debugString(invalidAttributeName)}`);
222+
assert_throws_dom(
223+
'InvalidCharacterError',
224+
() => document.createAttribute(invalidAttributeName),
225+
`document.createAttribute should throw an error for: ${debugString(invalidAttributeName)}`);
226+
}
227+
}, 'Valid and invalid characters in setAttribute, toggleAttribute, and createAttribute.');
228+
229+
test(() => {
230+
for (const validNamespace of validNamespacePrefixes) {
231+
for (const validLocalName of validAttributeLocalNamesShortened) {
232+
const element = document.createElement('div');
233+
try {
234+
element.setAttributeNS('namespaceuri', `${validNamespace}:${validLocalName}`, 'value');
235+
} catch (error) {
236+
assert_unreached(`element.setAttributeNS should not have thrown an error for: ${debugString(validNamespace)} ${debugString(validLocalName)} ${error.toString()}`);
237+
}
238+
try {
239+
document.createAttributeNS('namespaceuri', `${validNamespace}:${validLocalName}`);
240+
} catch (error) {
241+
assert_unreached(`document.createAttributeNS should not have thrown an error for: ${debugString(validNamespace)} ${debugString(validLocalName)} ${error.toString()}`);
242+
}
243+
}
244+
for (const invalidLocalName of invalidAttributeLocalNamesShortened) {
245+
const element = document.createElement('div');
246+
assert_throws_dom(
247+
'InvalidCharacterError',
248+
() => element.setAttributeNS('namespaceuri', `${validNamespace}:${invalidLocalName}`, 'value'),
249+
`element.setAttributeNS should have thrown an error for: ${debugString(validNamespace)} ${debugString(invalidLocalName)}`);
250+
assert_throws_dom(
251+
'InvalidCharacterError',
252+
() => document.createAttributeNS('namespaceuri', `${validNamespace}:${invalidLocalName}`),
253+
`document.createAttributeNS should have thrown an error for: ${debugString(validNamespace)} ${debugString(invalidLocalName)}`);
254+
}
255+
}
256+
for (const invalidNamespace of invalidNamespacePrefixes) {
257+
for (const localName of validAttributeLocalNamesShortened.concat(invalidAttributeLocalNamesShortened)) {
258+
const element = document.createElement('div');
259+
assert_throws_dom(
260+
'InvalidCharacterError',
261+
() => element.setAttributeNS('namespaceuri', `${invalidNamespace}:${localName}`, ''),
262+
`element.setAttributeNS should have thrown an error for: ${debugString(invalidNamespace)} ${debugString(localName)}`);
263+
assert_throws_dom(
264+
'InvalidCharacterError',
265+
() => document.createAttributeNS('namespaceuri', `${invalidNamespace}:${localName}`),
266+
`document.createAttributeNS should have thrown an error for: ${debugString(invalidNamespace)} ${debugString(localName)}`);
267+
}
268+
}
269+
}, 'Valid and invalid characters in setAttributeNS and createAttributeNS.');
270+
271+
test(() => {
272+
for (const validDoctype of validDoctypes) {
273+
try {
274+
document.implementation.createDocumentType(validDoctype, 'publicid', 'systemid');
275+
} catch (error) {
276+
assert_unreached(`createDocumentType should not have thrown an error for ${debugString(validDoctype)} ${error.toString()}`);
277+
}
278+
}
279+
for (const invalidDoctype of invalidDoctypes) {
280+
assert_throws_dom(
281+
'InvalidCharacterError',
282+
() => document.implementation.createDocumentType(invalidDoctype, 'publicid', 'systemid'),
283+
`createDocumentType should have thrown an error for: ${debugString(invalidDoctype)}`);
284+
}
285+
}, 'Valid and invalid characters in createDocumentType.');
286+
</script>

0 commit comments

Comments
 (0)