<URL: https://rt.cpan.org/Ticket/Display.html?id=86214 >
On Mon Jun 17 17:47:43 2013, MSCHWERN wrote: > While answering a question on Stack Overflow about how to make > SQL::Statement > understand an Oracle dialect, I found it very difficult to puzzle out > how to > extend SQL. While the option to issue SQL commands is straight > forward, it does > not scale. Because it's not designed to be. I always wanted to create such a thing for Perl6 - but ... you know ... missing tuit's, insufficient spare time, no business case - choose one ;) > Information about how to write a subclass is scant and does not answer > the > question of how to add to the SQL dialect. Indeed - even if the documentation at http://search.cpan.org/dist/SQL-Statement/lib/SQL/Parser.pm#dialect%28%29 is not as bad as you pretend. But don't discuss RT#86213 here, I'm annoyed enough about opening random tickets instead of asking at dbi-dev@ or dbi-users@ as documentation (you refer - because you read it entirely otherwise you can't decide what is missing) suggest for such cases. > There's little documentation on extending an existing dialect. I dug > into the > code and found get_config_from_hash could be extended (rather than the > documented get_config) but could not puzzle out how to add a function > like > LENGTH(). Well - the functiom get_config_from_hash is introduced by yourself (see #53128). Please blame the responsible one ;) > See my answer on SO for details. > http://stackoverflow.com/a/17155334/14660 Your answer there is absolutely correct. And the is no intention to hurry with a newly written SQL::Parser (you're invited to do - but please on at least the quality the new code in SQL::Statement, DBI::DBD::SqlEngine & Co. have - neither the get_config_from_hash nor 1.15 and below are a reference). BTW: I don't know what this ticket was good for. I assume it was a question and I hope the question is answered and the ticket can be closed. Otherwise, please feel free to clarify the goal of the ticket. Cheers Jens
