HTML Tag Cleaner Fails
Dar Scott
dsc at swcp.com
Tue Aug 8 18:49:35 EDT 2006
On Aug 8, 2006, at 2:58 PM, Bill Marriott wrote:
>>
>> What do you base this assumption on?
>
> Oh, I dunno, the name of the function beginning with "HTML?" The
> documentation which states, "Specifies the contents of a field,
> with its
> text formatting represented as HTML tags and special characters
> represented
> as HTML entities."
Look at what that actually says.
It says "with its text formatting represented as HTML tags".
It never ever says the field displays HTML.
It is clear that htmlText was never intended to have to do with
HTML. What is the lowest most functionality of HTML? The handling
of white space. The htmlText property does not do that in output or
input. Metacard and RunRev were clear some time ago in that this is
not a bug.
BUT, I think it would be a good idea to expand htmlText (when
fetched) to be a string that would display the same content (or very
similar) content when dumped into an HTML display. I suspect that
anything that can be displayed in a field can be represented in
HTML. This enhancement will probably break scripts, though. (Links
have completely separate semantics, though, and will not transfer
completely.)
> Links are handled.
They are?
> There can be no doubt that the intention of the function is to a)
> export the
> styled content of a text field as faithfully as possible using HTML
A candidate for an enhancement request. I'd vote for this. I'm not
sure how it should handle dontWrap.
> and b)
> to render incoming HTML as properly as possible.
Nonsense. It can't render HTML.
> No, it's exactly the right thing to do if you're implementing a
> function
> named HTMLText() that is designed to process HTML tags at best-effort.
It is clear from its behavior that that was not what it was designed
for. No effort was made to make that happen. Would you have
released htmlText the way it is if you made it to render HTML?
So removing <title> is bizarre. Somebody complained or asked for it
and it was tacked on.
Dar Scott
More information about the use-livecode
mailing list