Hello,
I have a German translation. On my computer the german special characters look OK, but when I deploy “?” appears. Also in the remote database they are sored with “?”, exactly as they appear on screen.
[attachment=0]german.JPG[/attachment]
The database is UTF-8. In case I wrote directly in the database table where the i18n messages are keep it works. But I would like to avoid writing again all message directly in the remote database. This should work directly when import solution.
Thanks,
Bogdan.
So when exporting messages through a solution file and importing it to your remote server the messages are inserted with ?
Then somewhere it goes wrong alone the line, does the driver need to know that your database is in UTF8? Is that configured correctly?
Can you export a sample solution with those messages and attach that to a case ? (so that we can see if it is not the export but the import that goes wrong)
I think the database is OK, at least it works if I manually wrote the texts in it.
I created the database on PostgreSQL using command createdb -E UTF8 -U DBA databasename This is how it looks using DBVisualizer.
[attachment=1]database.JPG[/attachment]
And also please see that if I manually change the texts it works, and also it works in the application.
[attachment=2]german2.JPG[/attachment]
[attachment=0]german3.JPG[/attachment]
For the German translation I’ve added the properties texts file manually in my Servoy development, then export the solution and then import it on the online application server.
I’ll create a solution and post a case.
Thanks.
udrescu_bogdan:
For the German translation I’ve added the properties texts file manually in my Servoy development, then export the solution and then import it on the online application server.
Thanks.
This was the problem, the file was not saved as UTF-8 encoding so export solution was not correct (Servoy reads the files as UTF-8).
Hello Laurian,
I have again a problem with the charset. I’ve added also a French, I took care to have the files in UTF8, although I noticed that if I add another language through Servoy i18n, it saves it in ANSI, not in UTF8 and the chars looks like this “\u0395\u03BB\u03BB\u03AC\u03B4\u03B1”.
Anyway, in my file the text looks like this “lbl.general_preferences=Préférences générales”, the file is UTF8, on server looks fine, but locally they look like this:
[attachment=0]app.png[/attachment]
If I wrote the special chars using \u format, in UTF8 files, they look OK locally, but on the server they are nasty. What’s stranger, is that first time when I deploy with French, the chars were in \u format, in UTF8 files, but they look OK on server and wrong on my computer, then after another deploy, when I changed something in the i18n file, they switched. ![:roll:]( "Rolling Eyes")
. So right now I’m really confused about the correct way to make this works.
Any idea?
Thanks,
Bogdan.
udrescu_bogdan:
Hello Laurian,
I have again a problem with the charset. I’ve added also a French, I took care to have the files in UTF8, although I noticed that if I add another language through Servoy i18n, it saves it in ANSI, not in UTF8 and the chars looks like this “\u0395\u03BB\u03BB\u03AC\u03B4\u03B1”.
Anyway, in my file the text looks like this “lbl.general_preferences=Préférences générales”, the file is UTF8, on server looks fine, but locally they look like this:
[attachment=0]app.png[/attachment]
If I wrote the special chars using \u format, in UTF8 files, they look OK locally, but on the server they are nasty. What’s stranger, is that first time when I deploy with French, the chars were in \u format, in UTF8 files, but they look OK on server and wrong on my computer, then after another deploy, when I changed something in the i18n file, they switched. ![:roll:]( "Rolling Eyes")
. So right now I’m really confused about the correct way to make this works.
Any idea?
Thanks,
Bogdan.
In developer we always read/write as UTF8 . Do you have a scenario that goes wrong (starting from 0) ?
Hello Laurian,
I created a messages file with Servoy i18n, by setting a text for the french translation. The file initially was ANSI with \u for special chars.
Then I copied the translation into this file. And then to make it works I had to set the encoding to UTF8 which transforms all that chars into ugly ones, then I copied again the content in it, so the chars in the file looked OK (or something like this, because I changed them a lot due to their ugliness when deploy ![:)]( "Smile")
It was really annoying).
On the server looks OK, but in developer looks as I showed in the image.
Thanks,
Bogdan.
udrescu_bogdan:
Hello Laurian,
I created a messages file with Servoy i18n, by setting a text for the french translation. The file initially was ANSI with \u for special chars.
The file created from Servoy should be UTF8.
udrescu_bogdan:
Hello Laurian,
Then I copied the translation into this file. And then to make it works I had to set the encoding to UTF8 which transforms all that chars into ugly ones, then I copied again the content in it, so the chars in the file looked OK (or something like this, because I changed them a lot due to their ugliness when deploy ![:)]( "Smile")
It was really annoying).
On the server looks OK, but in developer looks as I showed in the image.
Thanks,
Bogdan.
Maybe you manipulated the file with some external text editor and then saved the file as ANSI?
Hi,
I didn’t write here for long time, but I still have this problem since we discussed last time. But now it’s really bothering me so please try to do something about it.
When I add a text in the servoy developer, all the special characters will become \uXXXX. If I start the solution locally, they look OK, but if I deploy on the server they look ugly.
So, if in the file I have character ‘ä’, it looks OK on the server but ugly local. It corresponding UTF8, ‘\u00C3\u00A4’ looks as ‘ä’ local, but ugly on the server.
And by the way, Servoy developer creates a new i18n file as ANSI, not UFT8, I tested and retested and it just does so! Notepad++ states that the charset is ANSI. But even if you writes the special chars using UTF8 coding, as ‘\uXXXX’, the server does not displays them OK.
Please have a look at the attached images to see what happen if I just change a text using the i18n editor. I set the german transation for the word ‘Lesson’ and look how the german i18n file become. But I just can’t upload it like servoy creates it on the server because everything will be a mess!
Thanks,
Bogdan.
The i18n files are saved and loaded with ISO 8859-1 character encoding. For characters that cannot be directly represented in this encoding,
Unicode escapes are used; however, only a single ‘u’ character is allowed in an escape sequence
In your synch view, I see that in the local file, you have for each special char two unicode escapes, with the first one always 00C3, was this
saved by the servoy i18n editor ? does this chars display right in the editor ? also, how can it be that in the svn, you don’t have escapes for
special chars ? isn’t that a file commited from developer ?
Hello,
The file in the SVN is made by hand. Because only in that way it can be deployed correctly on the server, and the special chars will look OK in the Smart or Web Client.
The file in the left, the local one is generated by i18n editor after I’ve changed some text in i18n editor. As it is there with all the \u…\u… it look OK in the Smart or Web Client when open from the developer, but when deploy on the server they look ugly.
You can see in the image how they look in the editor with the file managed by hand. But when deploy this file they look OK on the server. It’s the only way it works OK on the server. But as soon as I change some text in the editor, the file will be saved with the \u…\u… and when deploy it it looks bad on the server.
So, it’s really a mess. And actually, when I deployed from the Mac with the file managed by hand, all special chars were missing. I had to send everything back to Windows and deploy from there.
Thanks,
Bogdan.
I just saved this text with servoy i18n editor : “können” and in the i18n file it is encoded like : “k\u00F6nnen”,
that is ok, but from you screenshout, in your file this is encoded like : “k\u00C3\u00B6nnen” ?
can you just create a new solution with i18n, and create a key with the value “können”, using the servoy i18n editor, and check the i18n file ?
do you have that double \u encoding ? on both mac & win ?
Yes, for a new file it looks as you said!
But if I change those that I already have it behaves like as explained. It might be because we put the text in there by hand. But we have more then 1000 messages. So we can’t just write them all one by one in the i18n editor. It’s a nightmare. What we do is we send the english file to a translator, receive the translation with the special chars not written in \u… format, because he can’t do that, then Servoy should be able to recognize those chars and transform them in \u… if it needs them this way, but instead it does it using 2 codes \u…\u… Or the best would be to provide some importing tool in servoy, for large amount of messages, with several formatting interpreters.
i am afraid that there is no other option that really use a properties file editor for that
something like:
http://sourceforge.net/projects/eclipse-rbe/
properties file just has to be in ISO 8859-1
or we could do UTF8 but the big question is… in what format do you get them back??? is that utf8 or is that the system default charset of the user that just edited that file???
Thats just the problem, they properties file itself doesn’t say that… so normal system text editors will very likely save it in the system charset. Then we still don’t know how to read it back in.
I was always able to paste the text received from the translator in an UTF8 text file and it was OK. The special chars looked OK in the text file, and OK when deploying on the server.
And actually this is true, the charset of the file that you saw in the SVN screen, with the special chars looking OK, not using \uXXXX, and that when deploy on the server it look OK, is an UTF8 file 100%.
So, maybe it’s a good idea to support also UTF8 files, which from what I saw until now, it’s much more easily to work with, compared with the ISO 8859-1 files. You don’t have all the special chars codes and the file is easy to read, especially when you want to translate from Japanese to Hebrew ![:wink:]( "Wink")
Even our database it has UTF8 charset because of this reason. You can see exactly what’s there!
i did find another editor:
http://propedit.sourceforge.jp/propertieseditor.jnlp
properties files are just defined in ISO 8859-1 java expects that
yes there are workarounds for it but then you are not standard anymore
Then it would be better to use xml for storage…
you may use the native2ascii tool from jdk to translate from utf-8 to the encoding used
by the i18n editor and back :
native2ascii -encoding UTF-8 utf8file.properties servoy.properties
native2ascii -encoding UTF-8 -reverse servoy.properties utf8file.properties
What is the schema for the XML files? And can servoy manage them? I’d rather use XML files with UTF8 format the property files.
no servoy doesn’t support that currently,
Then, do you plan to add UTF8 support for those files, xml or properties doesn’t matter? I can struggle some more time with them as I do now.
Can I make a request for this? Will it be implemented? If yes, can you estimate when?
But anyway, I think the saving of the file with 2 codes \u…\u… it’s a bug, isn’t it?
