in reply to RFC: Data::Sync

Looks very useful, gOn++. Am I right in thinking that this only applies to DBI? If so, I'd really suggest a name in the DBIx::* hierarchy since Data:: applies, in my mind to all kinds of data, not just DBI. I suggest you drop a note to dbi-users@perl.org mentioning the url of this node so the folks over there can get in on the RFC. Also, you should make it explicit what happens in these cases: a) when you try to merge a column of one type in a db with a column of another type in the second db and b) when the length of the data in one db is different from that in the second e.g. you try to put a VARCHAR(50) value into a VARCHAR(10) column.

Also, is this meant to only work on two database in the same RDBMS or is it meant to be able to go, e.g. from a MySQL database to a Pg database?

Replies are listed 'Best First'.
Re^2: RFC: Data::Sync
by g0n (Priest) on Sep 22, 2005 at 14:28 UTC
    thanks :-)

    It doesn't only apply to DBI: it currently only works with DBI & Net::LDAP handles (and the two are handled differently by the code). I did think about a DBI::* namespace, but as well as also supporting LDAP it's intended to develop to cover as many data source types as possible. I'm open to rename suggestions, but I wanted to use as descriptive and general a name as possible.

    /me slaps forehead - behaviour isn't defined for length and type mismatches. Those are going on the TODO list now and will be addressed ASAP.

    In theory you should be able to go from one RDBMS type to another - I've run from SQLite->LDAP, LDAP->SQLite, and LDAP->DBD::CSV so far.

    --------------------------------------------------------------

    $perlquestion=~s/Can I/How do I/g;

[reply]

In Section Meditations