com.gargoylesoftware.htmlunit.html.DomNode.asText() returns nested elements with linebreaks

In the case of nested elements, the text of span element is output with linebreaks:

<span>
    <select name="day" size="1">...</select>
    <select name="month" size="1">...</select>
    <select name="year" size="1">...</select>    
</span>

output: 1 JAN 2021

Therefore, real line breaks, like < /br> cannot be distinguished in output text from blocks. This worked fine in the earlier versions like 2.18 It seems that the status is set incorrectly after a block, see: com.gargoylesoftware.htmlunit.html.HtmlSerializer.HtmlSerializerTextBuilder.append(String, Mode)

[rbri]

Hi @dennisduysak ,

this is work in progress, i have not done much on this. The plan is

• to introduce the new method and make the old deprecated (done)
• switch from system dependent line separator to \n (done)
• make unit tests for at least all the cases we have unit test for the old method
• make sure to remove all the special cases we have implemented in the old hierarchy - have test for all the cases and adapt the impl to something meaningful (at this step your case might be fixed)
• cleanup, add more tests and release

But please keep in mind i do all this in my spare time, having a real >40h per week job - means any help is welcome.

[dennisduysak]

Hi @rbri , I saw that the changes for the select tags are already commited. Thanks a lot for this! However, the tables are still incorrectly formatted (see unit tests of HtmlTableTest.java in coment before). Is it possible to format them in the same way like the select tags?