You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When parsing selectors, we don't really handle EOF like the spec says. Does this really matter? 🤷♂️
When dealing with CSS parsing, the parser is supposed to translate an escaped EOF to (U+FFFD). If the escaped EOF is in a string, it is supposed to be ignored.
So, according to spec, we are supposed to translate something like this div\ (when the escape is containing the EOF) to something like this div�. When dealing with a string, something like this "value\ would translate to "value".
As far as strings go, none of our selectors can ever be valid if the string contains an escaped EOF. This is because strings are either in [attr="value"] or something like :lang("en-*"). If the string contained an escape EOF, the selector would be incomplete and throw an error anyways, so we really don't care about this edge case.
As far as identifiers are concerned, we could run into the case of div\. If this occurs, we should handle it in a sane manner. And according to the spec, this would translate to (U+FFFD).
The text was updated successfully, but these errors were encountered:
When parsing selectors, we don't really handle EOF like the spec says. Does this really matter? 🤷♂️
When dealing with CSS parsing, the parser is supposed to translate an escaped EOF to (U+FFFD). If the escaped EOF is in a string, it is supposed to be ignored.
So, according to spec, we are supposed to translate something like this
div\
(when the escape is containing the EOF) to something like thisdiv�
. When dealing with a string, something like this"value\
would translate to"value"
.As far as strings go, none of our selectors can ever be valid if the string contains an escaped EOF. This is because strings are either in
[attr="value"]
or something like:lang("en-*")
. If the string contained an escape EOF, the selector would be incomplete and throw an error anyways, so we really don't care about this edge case.As far as identifiers are concerned, we could run into the case of
div\
. If this occurs, we should handle it in a sane manner. And according to the spec, this would translate to (U+FFFD).The text was updated successfully, but these errors were encountered: