Geert,


<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1- transitional.dtd">


Ok, but that doesn't really matter, does it?

There is a little difference about &apos; which is defined in xhtml but not in html. Atm StringUtils.encodeHtml doesn't encode it.

Both engine types behave the same except for resolving the template name with a file name extension.

Yes

May be it's possible to do both of them:
- for any element or template, to give the choice between with or without html tags/entities capabilities - even in the case "without html tags/entities", to be able to use them with escaping, like
  summary-legend = XHTML Transitional 1.0\\<br /\\>Fragment
body-help = \\<div class=\\"form_help\\"\\>Vous pouvez saisir ici le texte complet de la "nouvelle". Seul \ du texte XHTML valide est acceptable.\ \</div\\>


I'm not so sure about this. You'll still have to educate people that they have to escape these character with a custom RIFE method.

Yes.

In my opinion, it's much better for them to know how to escape the characterd that encodeDefensive doesn't handle in a standard xhtml way. At least they already know it correctly.

My concern is to manage properties files with as few as possible code rules in them: - here, in addition to the specify rules of properties files (dynamic value, unicode, eol...), by default you must know at least the html codage of these characters - if you want to use these files in a not web context, you must then filter such html code.

Also, check out the frequency. How often do you add <>&"' to text. It's actually just really once in a while.

But it's something very nasty to deal with. And not so rare in language as french, with apostrophe and quotations

When you have xhtml enabled properties, you'll have a lot more characters for the tags. So imho this custom escaping might actually be more work.

Yes.

I think we should have the best of the two, as wanted:
- separation of concern ( :-) I succeeded to place it!), i.e. "pure" text: if xhtml tags are needed, then escape <>&" with something like "\\" - xhtml enabled: <>&"' as literals must be escaped with xhtml entity references

Regards

Pierre


_______________________________________________
Rife-users mailing list
[email protected]
http://www.uwyn.com/mailman/listinfo/rife-users

Reply via email to