$test = $dbh->quote($test);
[download]

DBI and DBD::Sybase give you some nice methods for dealing with your quoting problem like placeholders and binding

But back to your original question, in the sybperl docs it explains you have both the 'dbsafestr' method and placeholders available with the 'ct_dyn_prepare' method. My personal preference is for placeholders since they make for efficient queries if reused.

Of course you can always go to the monastary's mpeppler's page and read about both.

grep

These are not the monks you are looking for, move along

I have to second use DBI instead. Also, you should get in the habit of using placeholders and binding, as they can make a huge difference on performance.

Where I'm working now, most of our old code didn't use placeholders and binding. Now that we are literally doing 1000X more transactions then when the program was written, we are feeling the pain. We've gone back and started to change all of our old code, and it has made a huge difference. Not only are we able to keep up now, but the program now responds noticeably faster then it did before. I see using placeholders and binding like using strict and warnings, you may not think you need to use them, but if you don't use them, it will come back to haunt you.

PS We use an Oracle database.

PPS We are doing this to both our perl and java code, and both languages are benifiting from using placeholders.

Note that while usually usage of placeholders is standart way to deal with such problems and it is highly recomended with most databases in case of Sybase implementation of placeholders have a number of different problems. It is covered in POD documentation for DBD::Sybase in chapter 'Using ? Placeholders & bind parameters to $sth->execute'.

--
Ilya Martynov (http://martynov.org/)

For a description of some of the problems and pitfalls when using placeholders and DBD::Sybase, see DBI prepare_cached and DBD::Sybase.

Michael

$test = q/RunSql "delete sometable where strike like '%[A-Za-z]%'"/;

$test =~ s/'/''/g;

$sql = "update job set cmd='$test' where jobname='somejob'";
[download]

There are other methods as well - placeholders for example when using DBI or Sybase::CTlib as mentioned by others in this thread, but there are drawbacks to those as well - nothing's perfect :-)

Michael

Thanks all for the advice on placeholders and the use of DBI, both I will further investigate and learn. :)

The delete does work as indicated, but that delete statement is from a field already in an existing database. I'm just trying to read from one table, and store the field in another table. When I try to store the data, I always get an syntax error due to the inside single quotes. Sometimes there are backticks as well. I guess the real question was how do I stop Perl from interperting the command line as just text and not code that should be executed.

I'll try Michael's suggestion to see if that works for me.

Thanks again.
budman

That was quick...

Just doubling the single quotes did the job!

Thanks.
budman

No. In Sybase Transact-SQL, the from keyword in a delete statement is optional.



---

Everything went worng, just as foreseen.

The reason the FROM is optional is that you can do the following in Transact-SQL:

delete mytable
  from mytable m, theirtable t
 where m.somevalule = t.somevalue
   and ...
[download]

The same syntax can be used for UPDATE. Very nice for doing set based operations.

Michael