<Swing Dev> javax.swing.plaf.basic.BasicHTML.isHTMLString(String)

Wed Aug 7 15:31:39 UTC 2013

On 8/5/2013 6:44 PM, Kevin Connor Arpe wrote:
> Tom,
>
> [ Apologies for my late reply.  I forgot to confirm original request 
> for mailing list subscribe.  Duh. ]
>
> I agree with your reasoning.  It makes sense.  Can I submit a 
> Javadoc-only patch for widgets with text labels?

    Make sure that you signed the Oracle Contributor Agreement before 
contributing to the Open JDK:
http://openjdk.java.net/contribute/

> Examples: JLabel & AbstractButton subclasses: ctors + .setText(String)
> Maybe I could point people here: 
> http://docs.oracle.com/javase/tutorial/uiswing/components/html.html
> And make a reference to BasicHTML.isHTMLString(String) ... and update 
> its docs to explain the match algo -- very precise, intentionally.
>
> Re: "client property", are you referencing BasicHTML.htmlDisable?
> Ref: 
> http://grepcode.com/file/repository.grepcode.com/java/root/jdk/openjdk/7-b147/javax/swing/plaf/basic/BasicHTML.java#BasicHTML.0htmlDisable
> Woah, if so, that is pretty deep in the woods.  It would be nice if 
> there was a method to toggle that feature on relevant widgets.  
> Perhaps few others have found this rather esoteric feature reference & 
> docs.  Making the connection from a ctor or method setText(String) on 
> a widget to BasicHTML was a leap for me... and then I missed this 
> interesting and important feature via "htmlDisable".  Maybe some docs 
> to help point the way?  I only found this stuff by digging through the 
> OpenJDK source code.

     I agree that the better idea is to have a client property like 
"treat text as html"  that allows to not break the backward compatibility.

    Thanks,
    Alexandr.


>
> Thanks,
> Arpe
>
> >I wouldn't want the html detection to be any more lenient. It's already
> >problematic in that programmers need to ensure that they don't allow
> >user text to switch modes. Whilst it's best to do this with the client
> >property, that isn't always done in practice. Currently it just requires
> >that whoever edits the text uses the correct form, with any errors being
> >obvious.
>
> >Tom
>
> >On 26/07/2013 12:27, Kevin Connor Arpe wrote:
> >> Hello,
> >>
> >> This is my first post to this mailing list.
> >>
> >> I was playing around with HTML + text labels on Swing widgets today.  I
> >> noticed that the sniff test for HTML is defined by this static method:
> >> javax.swing.plaf.basic.BasicHTML.isHTMLString(String)
> >>
> >> The test is quite strict.  Basically, the string must start with
> >> "<html>" -- no whitespace allowed, but it is case insensitive.
> >>
> >> If there is a good reason for this strictness, please let me know.
> >> Otherwise, I would like to submit a small patch (+ test) to improve 
> this
> >> static method to be more forgiving.  If OK, please also advise about
> >> JDK7u vs JDK8.  (I don't know if a [simple] patch can be applied to 
> both
> >> releases easily.)
> >>
> >> In Perl regex parlance, something like: m/\s*<html\s*>/i
> >> Of course, my solution probably wouldn't use a regex, nor Perl. =)
> >>
> >> Please share you thoughts.
> >>
> >> Thanks,
> >> Kevin Connor ARPE
> >> Hongkong