inserted text take the wrong alignment

i try to insert the following string in the database

the red car (driver)

this string save like this

)the red car (driver

i have a problem when inserting string contains special character at the end of the string.

we have arabic and english string like this

???? ????? (R) radial ????

and it appear in reverse like this

???? (R) radial ???? ?????

You need to check the application that is inserting the data specifically the API commands being used. This is not a SQL Server problem per se. The database engine will store the values as passed from the client and doesn't manipulate it on the server. Also, where are you checking the display of the values? It is possible that the tool is doing something based on your language / regional settings. So this could just be a display issue also. Start with verifying the data in the back end tables directly, then your client code and then whatever UI you are using.|||

hello Umachandar,

me and Batool posted this one together

I do import the data into the database through a certain script, but I thought it was an sql problem, because the data were in the correct alignment before inserting, I see them reversed in the tables directly, actually to test this issue I tried to enter data directly into the database so in the cell I press ctrl + Alt + shift to reverse the alignment inside the cell in table, and when I start submitting my data it is reversed.

how could this be a display problem when it's correct in all other applications on my machine

thank you

|||

I believe I've seen funny behavior in Management Studio when you try to display mixed right-left and left-right scripts. (I doubt this is unique to MS.) Can you inspect the binary contents of the strings and see whether it contains what you expect?

Cheers,

|||

You should verify the data first without involving any UI elements into the picture. The reason I say that it could be a display issue is that the tool might be doing something different when reading and displaying the data. This happens for float data type values today. The accuracy of the digits are different from ISQLW and in some cases two values that differ in say the 17th decimal digit will look the same. But this doesn't mean that the values are the same.

So you could write a script or program that does the insert, reads the data back and verifies it using SQL only. This will eliminate the UI from the picture. Additionally, tracing the calls to the server from the UI / tool via Profiler will also help. You can find out if the provider/driver is translating the string based on code page settings. There are just too many variables involved in this. Is it possible to do the following?

1. Post a simple DDL, insert statements and SELECT which shows the behavior (note that you may have to use the appropriate collation and Unicode data type to avoid any character translation)

2. If #1 doesn't work for you, is it possible to post some steps using say a particular UI (like ISQLW or SSMS). Please be clear on how you are inputting the data (open table, script/open table combination) and so on. Schema and data type of the column(s) are important here also. You talk about entering something in a cell - where is this? What UI are you talking about?

Lastly, the configuration of the OS (language/regional settings) may also be a factor and version of SQL Server. So please post those also.