Question

I am working on a C# project which reads values from a MS SQL Server 2008 or 2008 R2 Server.

The query looks something like this:

string charge="1234567";
string command = string.Format("SELECT * FROM tableX WHERE chargeField = '{0}' ", charge);

chargeField is from type nvarchar(50) – do not ask me why – I cannot change the database.
Also the entries are written to the database by another tool which I also cannot change.
Sometimes the chargeField value will have leading or ending space characters in it, so my
query will return nothing because “1234567” != ” 1234567 “

However, when I write the query like this (without ‘ ‘):

string command = string.Format("SELECT * FROM tableX WHERE chargeField = {0} ", charge);

it works. Why does this even work – don´t i need the ” signs to indicate that the database field is a string? Can I safely change all my queries? Is this a “feature” from MS SQL Server and what exactly happens here.

EDIT

I solved the problem by using:

string command = string.Format("SELECT * FROM tableX WHERE RTRIM(LTRIM(chargeField)) = {0} ", charge);

Thanks.

Answer 1

The SQL Server tries to convert the fields accordingly. Without quotation marks, you’re passing in a number, so SQL Server tries to convert the value in chargeField to a number, which works in spite of leading spaces.

This won’t work if chargeField contains the value 12345a. This will lead to a conversion error. It is enough for one record in the table to contain such a value to get the error.

To be on the safe side, change your query to

... WHERE LTRIM(chargeField) = '{0}'

but be aware that this leads to a possible index on the field no longer being used. What you could still do is create an additional column that is filled using a trigger with the value of LTRIM(chargeField), put an index on that column and adapt your WHERE clause accordingly.