Bug #93409

Updated by Sybille Peters 2 months ago

Parsing of URLs without a <a href tag in tt_content.bodytext gets inconsistent results. 

 Sometimes it is a link in the frontend, sometimes not, the rendering of frontend and RTE is not always consistent. 

 h2. Example 1 

 <p>Weitere Informationen sind auf der Webseite des Projekts zu finden https://www.klasse-klima.de/.<br /> 
 BUNDjugend, Klasse Klima (https://www.klasse-klima.de/sfsd)<br /> 

 * In the frontend, the first URL is rendered as link, the second is not 
 * In the RTE, both links are not rendered as links 

 h2. Example 2 

 URL without <a href and directly followed by a ")" gets different results: 

 1. is rendered as link in frontend 
 2. is not rendered as link in RTE 
 3. is incorrectly evaluated by linkvalidator as URL with the ), so e.g. URL "https://example.org" becomes "https://example.org)" 

 tt_content.bodytext has softref:  


 I have wondered about the "url" in there, if you remove that you could just parse explicit links with "<a>" as links which seems to me to be a "cleaner" solution.  

 h2. Example 

 The following (raw) example text in tt_content.bodytext: 

 <pre><code class="html"> 
 <p>(some url http://t3coredev/sdfdsf2) more text ...</p> 

 h2. system 

 * latest master 
 * with Introduction package 
 * points 2 and 3 can be reproduced without Introduction Package, did not test 1) as that requires rendering of content in frontend