Bug: clean() discards text which contains a single '<' character

[whchdorf]

Hello,

I discovered that Jsoup.clean() apparently discards all text after a lower than< is followed by another alpha (non-numeric) character.
Here is a short test to verify the behaviour:

@Test
public void test() {
    var result = Jsoup.clean("this is <some harmless input text", Safelist.basic());
    assertEquals("this is", result);
}

Expected is the assert to work like this:

assertEquals("this is <some harmless input text", result);

This is potentially severe, since a harmless input text can be discarded after cleaned with JSoup.
Might be caused by a logic that tries to unsuccessfully find the end of an opening tag. However, I would expect JSoup to not touch this text at all and the test should return the same input.

Is there a way to fix this?
Thanks

[akshaya57148]

I guess for reserved chars like < , > etc , you will need to use HTML encoding:
< for <
> for > and so on.
For details refer :https://www.w3schools.com/HTML/html_entities.asp

[whchdorf]

I guess for reserved chars like < , > etc , you will need to use HTML encoding: < for < > for > and so on. For details refer :https://www.w3schools.com/HTML/html_entities.asp

When I escape the input string, JSoup just does not do anything because the input is already clean.

In case it was not stated clear enough: I want JSoups cleaning of tags still to work, but an opening tag should not stop it from working.
here is another test that should be green if JSoup works correctly:

@Test
public void test() {
    var result = Jsoup.clean("<a>RemoveThisTag</a><b>AndThisTag</b>but <not this harmless string", Safelist.none());
    assertEquals("RemoveThisTagAndThisTagbut <not this harmless string", result);

[jhy]

This is effectively the same as the feature idea in #1230 - rewrite illegal tags as escaped entities.

If implemented, it should be as an option to Cleaner.