Needs to be replaced with stringByRemovingPercentEncoding, which always uses the recommended UTF-8 encoding.
Created attachment 250523 [details] Patch
Comment on attachment 250523 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=250523&action=review > Source/WebKit/mac/ChangeLog:9 > + We now use stringByRemovingPercentEncoding, which always > + uses the recommended UTF-8 encoding. This looks like an undesirable change in behavior. When UTF-8 fails, we'll just get a null string, which is unacceptable. We should fall back to Latin-1 or maybe to keep the string percent encoded, I'm not sure what's actually best.