Question

I am trying to prove a table design flaw in a production db, that a table must not have a clustered primary key on a column that can have a random data, in this case a code keyed in by end user.

Though we know the solution is to make the PK as non-clustered, I still need to add rows to it for testing purpose on its replica. Therefore, I will need to know what would be the character I can use after ‘Z’ as a prefix.

More, the column is not a unicode, and it would be a mess to prefix my fake data with a series of Zs. The table is now having hundred-thousands rows, and each insertion is taking seconds.

Answer 1

Just run this and go down the list. I added the sandwiching dots for clarity, esp. when non-visible characters are involved.

select number, '.' + char(number) + '.' collate SQL_Latin1_General_CP1_CI_AS thechar
from master..spt_values
where type='p' and number between 28 and 255
order by thechar

There are only 4 characters coming after ‘Z’, since you say the column is not N(Var)Char.

121 .y.
89  .Y.
253 .ý.
221 .Ý.
255 .ÿ.
90  .Z.
122 .z.
208 .Ð.
240 .ð.
254 .þ.
222 .Þ.