Have you ever put a piece of content through your system only to find that when you get it back some characters aren't quite right? If this sounds like something you've seen recently chances are you are seeing a character encoding problem.
Unfortunately, characters aren't characters aren't characters when it comes to putting text through an application. You need to make sure all the layers of your application from the database through to the text editing solution are all encoding using the same character set. If this is not the case, you're likely to see character corruption, particularly for special characters i.e. characters from the extended character set.
Spotting the Issue
So how do you spot a character encoding problem? It's usually pretty obvious but here are some key pointers:
-
The symbol ? appears on entering text. This means character encoding has been corrupted.
-
The curved quote characters “ and ” or apostrophe character ’ - typically present in content imported from Word - are corrupted
-
Special characters you insert using EditLive!'s symbol dialog insert correctly but are incorrect once the content has been submitted through the system.
This is not to be confused with situations where you get the □ character in the document. If you see the box character this signals that the character cannot be displayed because the system doesn't have the required font installed (more info). This is typically seen when trying to render Asian characters on an English system.
Fixing the Problem
The good news is that this problem is pretty easy to fix. All you need to do is change the character encoding used in EditLive! to match that of your web application. All you need to do is open up the configuration file for EditLive! and add a <meta> tag. Your configuration file(s) can be found in:
-
IBM/PortalServer/installedApps/WCM_Authoring_UI__PA_XXXXXX.ear/PA_XXXXXX.war/ephox/editlivejava for ILWCM 5.1.0.x
-
IBM/PortalServer/installedApps/WCM_Authoring_UI__PA_XXXXXX.ear/res/editlivejava for ILWCM 6.0 or above
We've found that setting the character encoding to UTF-8 usually does the trick:
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8" />
However, if that still doesn't work you'll have to investigate your ILWCM settings to determine the character set correctly. Note that fixing the character set won't restore characters that have previously been corrupted.
If this doesn't solve your character encoding issue then chances are the issue is something else masquerading as a character encoding issue and you should get in touch with Ephox support.