Bonjour Pierre,
Thank you for your response. Unfortunately, alas, it appears that Oracle
doesn't work the same way as PostgreSQL in this respect. Although,
surprisingly (to me) it accepts an instruction such as
SET client_encoding TO 'LATIN9'"
with the result
ORA00000: normal, succesful completion
it does not appear to have any effect (nor does any variation of the parameter
to the client encoding).
With your suggestion as a clue, further googling suggests that with Oracle,
there is no way to change the encoding once the connection is opened. (Indeed
one remark suggests that [Oracle] "only allocates one OCI environment at the
time of the first connect, and so OCI only reads NLS_LANG once. This
environment is then shared across all future connections".) So the question
then is how the LiveCode driver for Oracle is implemented, and how it sets the
"NLS_LANG" or "NLS_CHARACTERSET"?
Which leads me back to asking - has anyone successfully read UTF8 data from
Oracle?
Many thanks,
Ben
On 15/02/2014 10:44, Pierre Sahores wrote:
Hi Ben,
Did't had to deal with Oracle backends for a while now but it should work in
the same way PostgreSQL UTF8 configured dbs does. Here is what works under
PostgreSQL 8.4 to 9.2 :
The worflow configuration context : LC Editor client app / Web Browser <- internet -> LC
server socket lisrener .lc script <--> LC lib stack.livecode <- localhost ->
PostgreSQL ;
Each LC server to PostgreSQL connection begins with the following line of code :
revExecuteSQL myDatabaseID,"SET client_encoding TO 'LATIN9';"
before any inteaction (INSERT /UPDATE / SELECT) follows as below described :
1.- Sending "INSERT / UPDATE" urlencoded data via "POST" to the server :
- Web browsers : UTF8 sets a as the charset (Head metatag of the displayed
HTML) ;
1.2.- Insert/Update with LC Editor client app running under windows (ISO
charset) : nothing to code on the LC Editor client app side : PostgreSQL
automatically converts the ISO received datas to UTF8 before inserting them in
the database ;
1.3.- Insert/Update with LC Editor client app running under macos x (Mac Roman
charset) : i just apply a MacToISO conversion to the datas to be posted to the
server before sending them : PostgreSQL automatically converts the ISO received
datas to UTF8 before inserting them in the database ;
1.4.- Insert/Update with LC Editor client app running under Linux (untested) :
should work in the same way Windows does ;
2.- Retrieving "SELECT" data from the server and displaying them on the client
app :
2.1.- Web browsers : UTF8 sets as the charset (Head metatag of the displayed
HTML) ;
2.2.- LC client app running under windows or android : nothing to do ;
2.3.- LC client app running under mac os x or iOS : IsotoMac conversion of the
received data before using them ;
2.4.- LC client app running under Linux (untested) : should work in the same
way Windows does ;
In your own context, and as long as no special coding would be needed to have
your Windows LC client app against PostgreSQL, i suspect that using :
revExecuteSQL myDatabaseID,"SET client_encoding TO 'LATIN9';"
should do the job. If not testing other client_encoding options would be what
to test first.
Best Regards,
Pierre
Le 14 févr. 2014 à 20:35, Ben Rubinstein a écrit :
I'm having trouble getting some data from an Oracle database using a LiveCode
app.
The application is running on a Windows 2008r2 machine, addressing Oracle
Database 11g Release 11.2.0.2.0 - 64bit Production.
The particular column in the database is apparently in UTF8, according to the Oracle SQL
Developer application. And making the SQL query in that app returns the text as
expected. But making the query through the LiveCode, characters such as the degree
symbol, or a right single quote, come through as "?". I don't think this is
happening in my script - I think that the data has been pre-converted before LiveCode
gets it. (Unfortunately I can only access the database by moving a compiled standalone
onto the Windows 2008 machine and running it there - I can't operate interactively.)
Does anyone have any experience of this, including that might confirm that it
should be possible? If this was through ODBC I'd be looking at the DSN
properties for character set conversion, but for Oracle it's a direct driver,
so I don't know how this would work.
Is it possible that the Oracle driver is just so old (I can't see any evidence
that it's been updated for years) that it somehow connects to the database in a
way that causes the database not to offer it UTF8?
Any suggestions, tips, or personal experience gratefully received.
Ben
_______________________________________________
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode
--
Pierre Sahores
mobile : 06 03 95 77 70
www.sahores-conseil.com
_______________________________________________
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode
_______________________________________________
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode
