Hi Sean,
On 2/14/2015 11:22 AM, Sean Kanaley wrote:
Nevermind, that seems to be what prepared statements do.
On Sat, Feb 14, 2015 at 9:34 AM, Sean Kanaley <skana...@gmail.com
<mailto:skana...@gmail.com>> wrote:
Hello All,
I would like to query based on an incoming hash (with help of json
library), but I don't see a built-in way to generate query needed.
There is a rows->dict but I don't see either a dict->where-clause
or at least whatever query uses in that dollar sign magic to
convert Racket-val->SQL-string.
Currently I have a function which takes a hash and makes
"<col1>=$1 <boolean> <col2>=$2..." so I can then do something like
(apply query <the above string> <the vals returned by that same
function as it built that string>). It would be nice and probably
more efficient if the value could just be spliced in while I build
the string to begin with, but it seems like only query has such
dollar sign conversion power available.
Am I missing something?
Prepared statements will be the most performant approach, but be aware
that prepared statements are bound to the connection - if you close the
connection you will lose them. Nor can you release the connection to a
pool because you can't be certain to get the same connection back when
you need a particular prepared statement. Additionally the DBMS will
have limits on the number of prepared statements per connection -
prepare will fail if you exceed the limit. Prepared statements may cost
considerable amounts of memory on the server side if that is a
consideration.
The client side alternative is to to hash both the SQL and the variables
and use apply (as you noted already). It is slightly less performant,
but it has no impact on connection pooling or on the server so it may be
the better way if you have many threads or many clients.
Hope this helps,
George
____________________
Racket Users list:
http://lists.racket-lang.org/users
