Re: [Tiki-devel] Making the old parser just work without changing how entities are stored

I am too familiar with this topic, but I think you are right that if the storage format (data) is changed a chain-reaction of problems may occur in other places.

So, keeping the change to the code/parser sounds right to me.

I am using the WYSIWYG editor, and I fear that changes in the storage may have very unfortunate effects there,

Arild

From: Robert Plummer [mailto:robertleeplummerjr@...]
Sent: 25. april 2012 14:56
To: Tikiwiki developers
Subject: [Tiki-devel] Making the old parser just work without changing how entities are stored

Hey guys,
After these issues keep popping up about the parser and how html entities are stored, is it really that big of a problem to store things encoded vs decoded?  After jumping onto the parser, and creating the framework for protecting entities till they are needed, I feel we could easily change the script to account for encoded entities vs decoded.  There seems to be one problem after another, and I don't feel that the storage is to blame, it comes back to the parser.  I never understood why we changed it to begin with.  I mean think about it, we are running into issues with storage of wiki pages, wiki history, blogs, trackers, the list goes on.  If we "fix" this in storage, it is going to haunt us for A LONG time to remain backward compatible.  The fix we need (I feel) is in the parser.  If we can simply ensure that things are encoded correctly, does anybody care how it is stored in the db?


So my vote it to remove the conversion script all together, and make the parser know better how to process entities, which it already does in part, we just need to change what it is looking for, rather then decoded, encoded entities.
--
Robert Plummer