MySQL Query Browser & Greek - help



  • I hope someone can shed some light on how to deal with this issue, its really frustrating:

    1) I am copying text from an Excel file that is greek and paste it into a table row cell in MySQL Query Browser (v1.2.12) on WinXP.
    Its an old version of Excel 2000 (9.0.2720) if that makes a difference, I never normally need it

    2) Before I can even copy/paste, the pre-existing text in the database does not display correctly in the Query Browser, it looks right on the website.  I assume this means everything on the server (and in my web browser) is okay, but that there is an issue with the Query Browser.  The server is running MySQL 5.0.45/linux.  The table's Character Set is utf8.

    3) If I do copy/paste anything into a field, it shows as a series of the 'black square characters' and is useless.

    I really haven't had to delve into language support before, so I am pretty much at a loss here.  I didn't find anything helpful on google - anyone have any idea how to fix this sort of problem?  I imagine it must be simple because its got to be pretty common, I just don't know where to start looking.



  •  Obviously the query browser doesn't understand utf8 or is not configured to do so. I don't know how MySQL handles that, but in Oracle, there is an environment variable called NLS_LANG, which also specifies the character set, e.g. GERMAN_AUSTRIA.WE8ISO8859P1. If this environment variable is set to a 8bit character set like in the given example, it's not (easily) possible to enter unicode strings which contain characters not in the specified 8bit-character set.

    As far as google tells me, the same is done in MySQL using the "set character set" command



  • I had the same problem.

    Try downloading this app: 

    http://www.sqlmanager.net/en/products/mysql/manager

    and pasting there.

    If it works, then the problem is definitely the QueryBrowser. 

    HTH



  • Thanks both of you, I did DL the sqlmanager and it also didn't help, but did give a lot more options (I may try it in the future if I run into other problems.)

    I think I have a theory as to what the issue is:  In this multi-lang package, they do have an encoding attribute, which setting the encoding character set to "iso-8859-7" for the greek text upon output of the viewed page, and I think the database is just storing it as utf8, but the client tool doesn't have any clue.  It uses the same table for all languages with a column called "lang" to group the different languages, most with a different character set to use for the encoding - so any query browser really has no idea how to represent any given row and the text in there.  All it sees is 8 bit characters which just happen to work in the browser when you view the site and the right character set is applied.

    I think my only solution is to edit the language import file, and re-import it for every change I make to see it.  I like being able to verify something is working throughout updates, but I guess I'll have to do it in chunks and fly blind for good portions.  At least I backed up the database - this should be fun.

     

    Thanks again 


Log in to reply