Bug: Vertical Bar Character as class name throws an Exception #1998
Labels
Comments
|
Yes, Inspired by mathiasbynens/CSS.escape, I wrote the following code, /*
Given a CSS identifier (such as a tag, ID, or class), escape any CSS special characters that would otherwise not be
valid in a selector.
*/
public static String escapeCssIdentifier(String in) {
StringBuilder result = StringUtil.borrowBuilder();
int[] codePoints = in.codePoints().toArray();
int length = codePoints.length;
int firstCodePoint = codePoints[0];
// If the character is the first character and is a `-` (U+002D), and
// there is no second character, […]
if (length == 1 && firstCodePoint == 0x002D) {
return ESC + in;
}
int index = -1;
while (++index < length) {
int codePoint = codePoints[index];
// Note: there’s no need to special-case astral symbols, surrogate
// pairs, or lone surrogates.
// If the character is NULL (U+0000), then the REPLACEMENT CHARACTER
// (U+FFFD).
if (codePoint == 0x0000) {
result.append('\uFFFD');
continue;
}
if (
// If the character is in the range [\1-\1F] (U+0001 to U+001F) or is
// U+007F, […]
(codePoint >= 0x0001 && codePoint <= 0x001F) || codePoint == 0x007F ||
// If the character is the first character and is in the range [0-9]
// (U+0030 to U+0039), […]
(index == 0 && codePoint >= 0x0030 && codePoint <= 0x0039) ||
// If the character is the second character and is in the range [0-9]
// (U+0030 to U+0039) and the first character is a `-` (U+002D), […]
(
index == 1 &&
codePoint >= 0x0030 && codePoint <= 0x0039 &&
firstCodePoint == 0x002D
)
) {
// https://drafts.csswg.org/cssom/#escape-a-character-as-code-point
result.append(ESC).append(Integer.toHexString(codePoint)).append(' ');
continue;
}
// If the character is not handled by one of the above rules and is
// greater than or equal to U+0080, is `-` (U+002D) or `_` (U+005F), or
// is in one of the ranges [0-9] (U+0030 to U+0039), [A-Z] (U+0041 to
// U+005A), or [a-z] (U+0061 to U+007A), […]
if (
codePoint >= 0x0080 ||
codePoint == 0x002D ||
codePoint == 0x005F ||
codePoint >= 0x0030 && codePoint <= 0x0039 ||
codePoint >= 0x0041 && codePoint <= 0x005A ||
codePoint >= 0x0061 && codePoint <= 0x007A
) {
// the character itself
result.append(new String(codePoints, index, 1));
continue;
}
// Otherwise, the escaped character.
// https://drafts.csswg.org/cssom/#escape-a-character
result.append(ESC).append(new String(codePoints, index, 1));
}
return StringUtil.releaseBuilder(result);
}then It seems to be a bit troublesome to make it work with Android. And I've found out a few more related issues.
See, |
|
Was fixed in #2146 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Consider the following HTML:
This HTML is valid according to https://validator.w3.org/nu/#textarea
If I try to get the CSS selector of the button, a SelectorParseException is being thrown:
Stacktrace (Jsoup latest stable version 1.16.1):
A real-world example where a class name is being used that consists only of a vertical bar character can be found here: https://www.mueller.at/ (Search for
.\|in the browser DevTools to find an example node)If the vertical bar character appears inside a class name (e.g.
a|b) a different error is thrown:The text was updated successfully, but these errors were encountered: