I am curious about the same thing. We have a server-side integration where the &nbsp after the period throws off a server-side routine to count sentences.

ditto. I haven't been able to figure this out either. if the editor has "this is a test. This is a test" (double space after the period) getXHTML() will return "This is a test.& this is a test."

If you don't want 2 spaces then don't type them.
This behavior is done automatically by the browser as it's the way to show the 2 spaces that the user has typed. Or do you think that it would be better for the user to start pressing the space bar and don't see any change on the screen, or that he sees some spaces in the editor and when he sees the final page all those spaces are gone because they were plain spaces instead of  ?

alfonsoml wrote:If you don't want 2 spaces then don't type them.

This really isn't an option. The users of the application are usually copying and pasting from Word. Some of the time there is one space after a period, sometimes there is two.

alfonsoml wrote:This behavior is done automatically by the browser as it's the way to show the 2 spaces that the user has typed. Or do you think that it would be better for the user to start pressing the space bar and don't see any change on the screen, or that he sees some spaces in the editor and when he sees the final page all those spaces are gone because they were plain spaces instead of  ?

I don't have much of a preference either way. The problem I am seeing is that getXHTML() doesn't seem to be converting the second space into  . it is only converting it to & and leaving the 'nbsp;' part out. So when the data is saved, there is an & and a space in between sentences where there used to be two spaces.

alfonsoml wrote:If you don't want 2 spaces then don't type them.

With a forum application, it's difficult to tell the users to only put in one space - people are going to do whatever people are going to do.

It's probably easier to strip out   characters in the back-end application than it is to get the editor to do it for you. A simple regex should do the job...

s/(\s*( )+)+\s*/ /mg

in Perl or any PCRE regex parser (like PHP pcre_replace, or JavaScript string.replace...)

LesBarstow wrote:It's probably easier to strip out   characters in the back-end application than it is to get the editor to do it for you. A simple regex should do the job...

s/(\s*( )+)+\s*/ /mg

in Perl or any PCRE regex parser (like PHP pcre_replace, or JavaScript string.replace...)

Thanks, that's what we're doing now. I just thought maybe there was a cleaner way further down the food chain, but I guess not.

Well (F)CK me! There's really no way to turn this feature off?!

The problem I'm having is that somehow through convoluted editing - probably going back and changing things with multiple spaces - my users are getting code like this:

 This is supposed to be all one sentence!

Hi all, as alfonsoml write there, you won't have this question anymore once you figure out it's the way how browser rendering sequential white-spaces and what is  , I found this short and consise article:
http://www.xs4all.nl/~dimaroan/htl/faq1.html

I'm sorry, but I find both the devs' replies to be very short sighted.

You are both entirely correct, and we all acknowledge and agree that, in order for a browser to render two visible consecutive space characters, one of them must be encoded as a non-breaking space html entity.

We also agree that there are times and places where users will specifically want this behaviour; and where this preference will outweigh the niggles raised here. We have no complaint with this being the standard behaviour of (F)CKEditor.

However, what you need to understand in return, is that there are other situations (such as an HTML email building tool) where we, the techies implementing a WYSIWYG editor for completely non-technical luddite users, must prevent users entering two spaces, whether accidentally or intentionally.

There are three variations I can see for how a user could end up with an unwanted non-breaking space separating two words:

1) Type a sentence; delete a word in the middle. Once there are 2 space characters only, one is converted. 50:50 odds on then deleting the natural space.
2) Accidentally typing 2 spaces (e.g. not prssing the 'a' hard enough "in a sentence") and then deleting the second one
3) Type a sentence; add a word in the middle by first pressing space and then adding the word.


If we can't have disabling non-breaking space conversion as an option in the browser, would you at least please tell us where in the code it's implemented, so we can take it out in our own installations? (Yeah, it's been that hard to find...)

Is anyone able to point me to the relevant bit of code?

Thanks!

Yes I need this answered too!

I am allwing users to edit posts after they have been created, and if they decide to edit mid sentence.. the   is introduced even when only one space is present between the characters!

This is a major flaw in using this editor for future projects!

Please point us coders in the right direction on this one!

I tried to explain this 10 months ago but it's obvious that I failed:
There's no code in CKEditor to detect such double space and insert  . This is done by the browser itself and so the behavior depends on each browser, ones might be more zealous than others with regards to this issue and if you don't like it you'll have to write the code to detect the behavior and fix it.

alfonsoml wrote:I tried to explain this 10 months ago but it's obvious that I failed:
There's no code in CKEditor to detect such double space and insert  . This is done by the browser itself and so the behavior depends on each browser, ones might be more zealous than others with regards to this issue and if you don't like it you'll have to write the code to detect the behavior and fix it.

While this can be fixed on the server side, could it be fixed by CKEditor? There are output formatting option (that might introduce more trouble than convenience) to control the HTML source, won't it be able to replace (\S \S) by (\S \S)?

You can do the replacement anywhere that you like. At the server or at the client. It's up to you.

Can anyone give hints on how to do such a replacing within CKEditor?

I think, there could (should?) be some code in CKEditor, that 

• (while typing) will replace at least single nbsp's with spaces, maybe under the condition, they were not in the document initially.
• or, a plugin for a button to convert all nbsp's.

Does anything like that already exist?
Thoughts?

similar questions here
http://ckeditor.com/forums/CKEditor/Recommendations-for-how-to-deal-with-unwanted-nbsp-entities

and here
http://ckeditor.com/forums/FCKeditor-2.x/display-invisible-chars