TYPO3 Core

I've discovered some drawbacks to my proposed solution.

It has been observed that after any <hr> in the RTE, all the links gets rendered like "<link ....>", breaking the whole validation and the HTML output.

This has been observed with "config.xhtml_cleaning = all". I don't know if disabling the cleaning makes any difference.

I have the same problem, but found out, that if an other block-element is placed in the content, the parsing is ok after that block-element.

I added the bodytext-class to my p-tag so i can see, what happens:
< p class="bodytext">some text< /p>
< hr />
< p>after hr, the class is missing< /p>
< ul>
< li>listelement< /li>
< /ul>
< p class="bodytext">and here the class is back again< /p>

Instead of:
lib.parseFunc_RTE.externalBlocks = table, blockquote, ol,ul, hr
lib.parseFunc_RTE.externalBlocks.hr.stripNL = 1

I added the hr-tag to:
lib.parseFunc_RTE.nonTypoTagStdWrap.encapsLines.encapsTagList
Links are parsed correctly with this - even directly after the hr-Tag.

this TS works for correct inserting of the HR:
lib.parseFunc_RTE.externalBlocks := addToList(hr)
lib.parseFunc_RTE.externalBlocks.hr.stripNL = 1
lib.parseFunc_RTE.externalBlocks.hr.stdWrap.parseFunc < lib.parseFunc

Hey all,

@Holger: Yes, this TS works as long as there is no table in the content afterwards...

Anyway, this HR tag is really confusing. I cannot even do a

lib_parseFunc_RTE.tags.hr = TEXT
lib_parseFunc_RTE.tags.hr.value = Here comes the line

The HR tag really messes up the parseFunc. I think we need to fix this bug soon, anyone willing to dive into to the mighty t3lib_parsehtml ?

Alrighty,

I found the problem...

The RTE / DB sends the parseFunc this content for instance:

My bla<hr>bla2

So, if I consider <hr> an externalBlocks, the block starts there and never closes, so we need an exception for the empty tags to be considered only open.

For now I solved this by simply putting
$content = str_replace('<hr>', '<hr></hr>', $content);
in the first line of t3lib_class.t3lib_parsehtml.php:splitIntoBlock()

I know it will belong somewhere else. Basically, to get a clean solution, it should replace all "empty tags" (img, hr, br) <hr> => <hr/> and then we should modify the regular expression to detect them correctly.

This one is related to #10609 (in fact they are the same only with different angles).

Still an issue 4.3.3 and 4.4.0

Still exists in 4.4.2!

I managed to make it work in 4.4.1

These are the lines I used...
lib.parseFunc_RTE.externalBlocks = hr,ul,ol,table
lib.parseFunc_RTE.externalBlocks.hr.stripNL = 1

Seems to work fine, and tables are rendered after a hr tag. HR tag is no more surrounded by the paragraphs.

The only problem so far is that the default "contenttable" class is removed from the table. If you don't want to have classes on your tables, I don't think this should be a problem, otherwise, you'll need to find a way to fix that issue too.

Still exists in 4.4.4. Anchors are not parsed after <hr>

The attached patch fixes this issue:
1. Fix the RTE tranformation (t3lib_parsehtml_proc):
1.1 Remove hr from $blockElementList;
1.2 Add hr to allowTagsOutside;
1.3 Wrap hr tags with LF on way to database
1.4 Avoid wrapping hr in paragraph on way to RTE.
2. Fix hr rendering (tslib_content):
2.1 Fix parseFunc handling of empty tags;
2.2 Fix encaps_lineSplit handling of empty tags (if on a single line) (from patch by Bernhard Kraft)
3.0 Add hr to lib.parseFunc_RTE.nonTypoTagStdWrap.encapsLines.encapsTagList (css_styled_content static template).

Awesome work. Again and again!
Successfully testet your patch for 4.4

EDIT
Radu Mogos is using this configuration:
lib.parseFunc_RTE.externalBlocks = hr,ul,ol,table

After applying the patch I had to change this line (removed the hr-tag):
lib.parseFunc_RTE.externalBlocks = ul,ol,table

@Henrik: The patch is pending in core list. Please give your +1 there.

Committed to SVN TYPO3core trunk (revision 9883).

Committed to branch TYPO3_4-4 (revision 10016).