Hello.
I'm new to PostgreSQL and I'm using it for some project. I
am currently creating an entity engine that provides a web
interface, which is oriented towards end users, to the database
based on some enriched description of the tables (from which it
also generates the SQL code to create them).
My application connects to the database using the socket-level
frontend/backend protocol rather than providing a wrapper around
the libpq library.
When an error takes places executing a query, all I get is string
describing the error, such as ``ERROR: Cannot insert a duplicate
key into unique index locations_name_key''. However, I can't
pass this error description back to the end-users: I have to
translate it into something meaningful for them (for instance,
something that refers to the fields they are seeing in their HTML
forms and something in the language they selected).
Currently, when an error takes place, I send multiple queries
back to the database trying to detect what could be the cause(s)
of the error. This seems wrong as I can't know in advance, in a
maintainable way, all the possible causes for errors. It is also
slow as multiple (ideally redundant) queries have to be sent to
the database.
I could also parse the error string and try to detect what went
wrong based on it, but I'd rather stay away from that option,
which doesn't look very maintainable. Or should I do that?
Do those strings get translated to different languages, for
instance?
What would experienced PostgreSQL users suggest I do?
Alejo.
http://bachue.com/alejo
--
The mere formulation of a problem is far more essential than its solution.
-- Albert Einstein.
$0='!/sfldbi!yjoV0msfQ!sfiupob!utvK'x44;print map{("\e[7m \e[0m",chr ord
(chop$0)-1)[$_].("\n")[++$i%77]}split//,unpack'B*',pack'H*',($F='F'x19).
"F0F3E0607879CC1E0F0F339F3FF399C666733333CCF87F99E6133999999E67CFFCCF3".
"219CC1CCC033E7E660198CCE4E66798303873CCE60F3387$F"#Don't you love Perl?