Yes.

Our Japanese had the same problem for searching as well.

c5 uses the search system called, Zend Lucene, which does not really have the capability to search those non-ASCII characters.

I think Remo came up with the solution. So search Remo's post regarding Lucene Search...

Since this forum is getting very big... I've been struggling to find one....

We came up to modify the Zend Lucene to be able to search Japanese... but the system is only valid for Japanese right now.



Also Andrew (CTO of C5) is now building c5's original search engine.

So this problem will be fixed in next version or later (we hope)

Thanks katz515

I have read all posts by Remo, and in general all of the section "Internationalization". Even I have read many forum messages concrete5.fr because the French share many characteristics with the Spanish, except the letter "ñ" and I've applied all patches and proposing changes in codes and have not achieved anything.

See if you got lucky and someone can help me.

Greetings from Spain

what exactly did you try?

Did you set entities to raw (tinymce)? Otherwise it encodes these "special characters". And make sure you build a 100% utf-8 site and also make sure you save them like it (entities:raw), otherwise it never worked for me...

Thanks Remo!

My English is very poor, but I will try to explain.

My problem is when you search for words with accents or special characters.

Only in the search, as in content, titles and navigation bar everything is perfect.

My php knowledge are limited, I am more designer than programmer.
Even so I tried to put "http-equiv="content-type" <meta content="text/html; charset=UTF-8" />", "define ( 'DB_CHARSET', 'utf8');" and retouch adodb_mysql.inc.php code and other solutions that I read in this forum.

I am very interested in doing work since c5 is the easiest CMS to maintain that I have found.

You can see the test that I am doing in c5 and test the system to search inhttp://www.xn--malagadiseo-beb.com/OFM/index.php...

I very much appreciate your interest and support

Greetings

Took one weeks studying the problem and there is no way that works. So much that I read it seems the problem is the framework used for searches, Zend Lucene, which provides fault with the character encoding.

I applied the patches released by Remo and katz515 and the problem persists.

In fact, if you do a search on this forum for the word "configuración" not find and the form field returns "configuración"
However if we look for "configuraci&oacute;n" you can see that if we get results.

Any suggestions or ideas?

Thanks and best regards to all

Well.
I did some reading on the zend stuff. I'm not a coder or anything but I really like to read. :)

I found something in the docs at zends website and tried changing htmlentities to htmlspecialchars. Now when I do a search the search field doesn't convert my character. Wich in this case was an "ä".

Before: välkommen became vã¤lkommen
Now: välkommen becomes välkommen

Here's what I did.
Go to:
concrete\blocks\search
Open up controller.php and go to line 164.
Look for
and change that to


But remember, i'm not a coder, so please check with someone that know PHP before you do this.

Now on to the RSS feed thing..

Copy the search block to the block folder outside the concrete folder. Then you don't have to worry when upgrading concrete.

I think this should work. Because tinymce saves all the content with specialchars in the database.

Thanks teekay.

The change in the code controller.php had already tried without success :-(

I also tried the second change that you are proposing moving the block searches, but neither got anything

I will continue studying the issue and asking everyone to see if I find the solution because I am very interested in this CMS for ease of editing frontend, which is simply great.

Thanks if you find something and do not forget to tell me.

Greetings from Spain

When I wrote these patches and talked to Katz, we were able to fix all the problems.

There are several sites working the way you want.http://www.iow.ch/de/searchresult/?query=beh%C3%B6rden...
The character "ö" isn't an ascii character and has (had) the same problem.

Just work on the stuff that has already been posted, it always worked for me and Katz too as far as I know.

I just checked your site. You still didn't fix the entities problem!

Set it to raw, "rewrite" some of the characters and search for them and you'll find them!

Your site contains all kind of encoded stuff like "crete &aacute;&eacute;&ia"

have the same problem in greek.

Whenever I type characters they are converted to their &alpha &beta et.c., but the acute characters are lost in the translation as there are no html codes for them.

Tried to put raw entities in the editor, but when I set it to custom and add the entities: "raw", and save, when I get back to edit block, I only get the very reduced version of the editor which shows html codes et.c. plus, when I enter άέήίόύώ I get ??????? instead.

This is REALLY a party stopper, I'm a mech.eng. not a computer guru, I have my partner fuming over cement5 for he want to write the greek version ofhttp://www.imeda.eu and I got all the flak for choosing it.

Needless to say that even if I enter HTML block and add each character by its UTF-8 numeric code, it just ignores it.

IHMO, this is major.

I found the solution to my "Greek" problem.

It was not Concrete5 or the editor, it was the database which the host set wrong as ... Swedish for some reason.

Basically, the problem was as follows:

Enter greek text, and all letters are converted to &alpha and &beta et.c.

Which is alright for all small and capital letters, BUT the acute letters.

The funny thing was that I checked the HTML that editor was producing and seemed alright, BUT when it was saved and then displayed it was wrong.

Apparently, the mysql could not understand these characters and was putting a "?" instead.

So, after mysql collation was set as to save utf-8 greek, all was nice and good.

Something in the lines:
ALTER DATABASE db_name
CHARACTER SET utf8
COLLATE utf8_general_ci

Hope it helps someone.

Bottom-line:
- Problem is not with concrete, it is with MySQL settings
- Change collation to your language

Suggestion:
Developers put something in Das Boat (joking!) that it sets the database collation from within the CMS in accordance to the language that one selects somewhere in Concrete5.

Regards from Athens
GPD

p.s. really love your work!