Re: DBI::CSV usage

[andyford]

In general, yes. But, given that different applications and people may have varying levels of strictness about their CSV-ish-ness, I would do some testing before I commit to automating in production or something like that. Even after that, you should try to think of some (all?) of the ways that the creator of the CSV file can screw up the format and try to detect them.

[jZed]

Usually DBD::CSV will die if it can't parse something, so using standard DBI error-trapping procedures in your DBI code should warn you about the kinds of problems AndyFord mentions.

and wrt non-Perl AndyFord - nice!