FiveTech Software tech support forums

Hi Birol,

Do you still use charset "latin5_turkish_ci" in mariadb/mysql using fivewin new rowset?

If yes, do you have a problem like that.
http://forums.fivetechsupport.com/viewtopic.php?p=231305#p231305

Thanks.

Hi,

I can convert my dbf file to MariaDB using importfromdbf with "latin5_turkish_ci". I see the right characters in HeidiSQL. There is not problem.

But all turkish characters are shown "?" in gets and xbrowse.

I just sent revised lib Please use this and let us have your feedback Yes, It is ok now. I have also tested to codeblock. Thank you. Do you have a time to check next problem. charset "latin5_turkish_ci" in http://forums.fivetechsupport.com/viewtopic.php?p=214507#p214507

Even here the problem is with Turkish language.
Problem is coming because of the two letters "i" in the word "in_transaction"

... with this test: ::IsMariaDB := "MARIADB" $ Upper( ::cServerInfo )  We did not foresee that this would fail on a Turkish installation and may be it may fail on some other language installtions. We have now modified the test as         ::IsMariaDB ...

Code:
? "MARIADB" $ Upper( oCn:cServerInfo )
.F.
Code:
? Upper( oCn:cServerInfo )
10.4.12-MARİADB (lower 'i' ==> Upper 'İ' codepage Windows 1254 Turkish charset)
Code:
? oCn:IsMariaDB
.F.

This is the cause of all problems.
Now let us think about it.

Hi Mr. Rao,

Code: Select all  Expand view

? oCn:cServerInfo

10.4.12-MariaDB

Code: Select all  Expand view

? "MARIADB" $ Upper( oCn:cServerInfo )

.F.

Code: Select all  Expand view

? Upper( oCn:cServerInfo )

10.4.12-MARİADB (lower 'i' ==> Upper 'İ' codepage Windows 1254 Turkish charset)

Code: Select all  Expand view

? oCn:IsMariaDB

.F.

Horizon wrote:My clients were asking why there were foreign languages ​in some places.

I could not understand. Can you explain in greater detail?

Horizon wrote:FWMissingStrings() function always gives same result like below.

Code: Select all  Expand view

0

"0" means there are no missing strings.
Are you trying to add Turkish Language?

This works if we set FW_SetUnicode( .t. ). Without that, I could not find a way. Probably we need to set the right codepage for the Turkish language. I could not figure out what it is. Maybe other programmers from Turkey can help. Note: ? command and all Msg* functions like MsgInfo() display ...

... But you prefer to use latin5_turkish_ci collation. I need to test this at my place. Can you please send to my personal email one dbf file with Turkish characters and also tell me what is the correct Turkish codepage to be set in Harbour application? HB_SETCODEPAGE( ??? ) HB_LANGSELECT( ??? ...

Hi, Changing this line solve in HeidiSQL app. but The Turkish characters are same ("?") in my application.     FWCONNECT oCn HOST "localhost" USER "root" PASSWORD "a" CHARSET "latin5"   I use FW_SetUnicode( ...

...     endif  NEXT   ocn:close() return nil  All data is converted to table in MariaDB with no error. My problem is that All Turkish characters are converted as "?". Even in HeidiSQL app. If I use "utf8_turkish_ci", there is no problem. I want to use "latin5_turkish_ci". ...

Keep aside mysql or dbf. We need to address the Harbour collation issue with Turkish codepage. Try this example (normal English) function Main()   ? "Z" > "W" // --> .T.return nil  Next let us set codepage TRWIN ...

Let us see if other Turkish programmers can help.
Also please give me time till I return from Seminar

... TRWIN, Ü comes at the end and all other characters are sorted normally like English. First we need to resolve the issues with proper setting up of Turkish codepage in Harbour. I do not know Turkish. So I request you to find out how to configure the Turkish codepage in such a way that Harbour Sorting ...