Recently we got aware of problems with the (spanish) language setting. Search results on accented vowels like "á" (&aacute) etc., " " (&uuml) or "ñ" (&ntilde) didn't show correctly. In the index, we were able to find "SEÑOR" but not "Señor" etc.

Therefor we changed fields 7 and 8 of the record SYSENV LND_SPANISH to match all required options and updated the corresponding indexes.

The search results now showed the proper values including accented characters, including lower case / upper case conversions.

But now, several programs won't work. We get errors in some programs and symbolics, saying that certain fields are not available in the dictionary, or the column headings show strange results, converting vowels to accented vowels.

(We don't use field names with accented characters, only standard A-Z plus "_".)

What are we missing?

Regards

Wilhelm

In order to understand the language configuration better:

Where can I find more information on the proper settings for field 7 and 8 in SYSENV LND_SPANISH? Apparently field 7/8 are @upper.case and @lower.case.

Do we also have to care for field 11? (lots of "0" and a few strange characters in between).

The language-critical searches occur mostly on XREF's. In this case we could also change the XREF_IDX call in the FIELD_XREF symbolic.

Wilhelm

Oops…

…at the end of the previous message it should read: In this case we could also change the XREF call in the FIELD_XREF symbolic….

Wilhelm

Wilhelm:

Your post called my attention, because I thought we were the only  users in the community programming in spanish. I can't give you a technical answer on those SYSENV fields, but we are working on a OI application that is completely in Spanish, and we are not using the LND_SPANISH because the spanish translation of the messages and windows seems more stranger to us than english. Are you developing the application in ANSI or UTF8 mode?. In ANSI, I understand there may be problems with the 

and the é, because they are delimiters. We are programming the entire application in UTF8, and we do not have the problems you mention. Unfortunately, we do have problems with multiple keys XLATE and the CheckLegalUser function, unsolved to date in OI 7.0.1. We are waiting to see if the next OI release solves them.

Regards,

Enrique

Enrique,

we are working with 7.1 beta using core functions of the OEngine. The GUI part in our system is handled by the Java/JOI client.

We have not switched to UTF because we couldn't get the CGI to behave decently with and ú.

Recently we stumbled over the accents in our XREF search engine and, since we didn't find a better solution, we rather went to intercept the XREF call in the symbolics and redirect to our function, where we switch accents back and forth. This seems to work by now …

Oh yes, the translations… we just ignore them. (There is still much oportunity for improvement here

Saludos

Wilhelm

Enrique-

The XLATE issue has definitely been resolved in 7.1; I'll check the BCheckLegalUser.

Mike

View this thread on the Works forum...