[TYPO3-RTE] Avoiding stolen entities when switching between html code view and wysiwyg view?

[Uschi Renziehausen]

Hello list,

I have been struggling with all those transformations for weeks now and 
have been quite successful so far. I managed to convince t3 to treat DIV 
and P elements separately (at some cost, though), and I can write ABOUT 
html and xml elements even outside <pre>, but one problem remains 
unsolved: No way to talk about entities.

It seems, that this time my enemy is not the central html parser, but 
some magic done, when I am toggling between html code view and wysiwyg 
mode and (as I suppose) also on save. This is what happens:

- In wysiwyg I write &lt; (some of us do have to talk ABOUT entities!).
- I switch to code view and I see &lt; where it should be &amp;lt;
- Now, switching back to wysiwyg, I of course have <

I have read the documentation at least ten times now, but I still might 
have missed something. I know that I can control, which elements shall 
be removed on toggle or save and so on, but I did not find anything 
about controlling what shall happen to entities.

Is there a configuration option or not, and if not, could it be added?

My rtehtmlarea (1.4.1) is still running in Demo mode (will take some 
more time until I can switch to Advanced :)) and here is the config as 
far as transformations are concerned (comments left in):

	# Do not clean code pasted from word processors currently, because we 
first have
	# to find out what to clean! It is no good to just remove all 
formatting, because
	# authors will get frustrated when they have to format it all over again.
	# Better try and remap what makes sense. On the TODO list!
RTE.default.enableWordClean = 0


	# Still not really figured out, but with these settings we will be able
	# to put html example code in inline elements.
	# NOTE: This only counts for elements, but not for entities!
	# Even on JS level there is some unwanted magic when switching between
	# html code view and wysiwyg mode
RTE.default.proc.entryHTMLparser_db.htmlSpecialChars = 0
RTE.default.proc {
    dontHSC_rte = 1
    dontUndoHSC_db = 1
}

	# Leave my DIVs and do NOT treat them as paragraphs!
	# This can be done by hiding div and p from the central
	# parser.
	# Unfortunately this means, that <p> will also be stored
	# in the database and there are no linebreaks left! No way out of this
	# unless one writes a customized central parser!

RTE.default.proc.entryHTMLparser_db.tags.div.remap = box
RTE.default.proc.exitHTMLparser_db.tags.box.remap = div

RTE.default.proc.entryHTMLparser_db.tags.p.remap = para
RTE.default.proc.exitHTMLparser_db.tags.para.remap = p


RTE.default.proc.entryHTMLparser_rte = 1
RTE.default.proc.exitHTMLparser_rte = 1

	# Elements generally allowed
	# If you decide to keep your p and div elements on the way to the 
database by
	# mapping them to some fancy tag (here box and para), make sure that 
those fancy
	# tags are members of this list!
RTE.default.proc.allowTags = box, para, a, abbr, acronym, address, 
blockquote, b, br,caption, cite, code,	dfn, dl, dt, dd, div, em, h1, h2, 
h3, h4, h5, h6, hr, i, img, kbd, li, link, ol, p, pre, q, sdfield, span, 
strong, sub, sup, table, thead, tbody, tfoot, td, th, tr, tt, ul, var


RTE.default.proc.entryHTMLparser_db.allowTags < RTE.default.proc.allowTags
RTE.default.proc.entryHTMLparser_rte.allowTags < RTE.default.proc.allowTags
RTE.default.proc.exitHTMLparser_db.allowTags < RTE.default.proc.allowTags
RTE.default.proc.exitHTMLparser_rte.allowTags < RTE.default.proc.allowTags


	# Make sure that the fancy elements box and para are in the list
	# of allowedTagsOutside, because if not so, they will be wrapped
	# in <p>-elements
RTE.default.proc.allowTagsOutside = hr, address, box, para

	# I do not know why, but without the allowTagsOutside defined for 
entryHTMLparser_rte
	# and exitHTMLparser_rte the div-elements are preceded with <p></p> 
after loading the
	# content from the database for the first time, and if you then store 
the content again,
	# you end up with <p>&nbsp;</p>.
RTE.default.proc.exitHTMLparser_rte.allowTagsOutside < 
RTE.default.proc.allowTagsOutside
RTE.default.proc.entryHTMLparser_rte.allowTagsOutside < 
RTE.default.proc.allowTagsOutside
	
	# Not sure yet. We will see
RTE.default.proc.entryHTMLparser_db.keepNonMatchedTags = 0

RTE.default.proc.HTMLparser_db.xhtml_cleaning = 1


Ciao, Uschi