the sqlt test suite is good :)
Create table schema manually from description
and sqlt doesn't appear to miss anything either
output from sqlt -f PostgreSQL -t DBIx::Class::File --prefix=EDINA::ORI < in > out
Is this a bug in dbicdump ? I can't tell :)
| [reply]
[d/l]
[select] |
dbicdump: Hmmm - Ubuntu has packaged 0.07010, and CPAN has 0.07023... I'll have a look at the changes file!
pg_dump: I've removed a pile of PostGIS stuff from this (I think I got it all....)
| [reply]
[d/l] |
FWIW, sqlt SQL::Translator v0.11010
chokes on
COMMENT ON TABLE xml IS 'Full XML records, to test for changes.
We don''t need to know the repo or anything: if the XML we have to pro
+cess matches a record, then we can skip to the next record.';
and ALTER SEQUENCE "Sources_id_seq" OWNED BY networks.id;
but otherwise merrily generates a gigantic module with no missing fields, same as it did your original snippet at Re^2: dbicdump not importing the full table (Postgres)
I did not try the latest dev version RIBASUSHI/SQL-Translator-0.11010_06.tar.gz | [reply]
[d/l]
[select] |
Hmmmm... the plot thickens!
W did some testing with a colleague, who had got good results when he demonstrated the advantage of Catalyst to me.
- He's running Perl 5.12, I'm on 5.14
- We are running the same version of catalyst
- He's running DBIx::Class::Schema::Loader 0.07017, I went from 0.07020 to 0.07023
I have rolled back to ::::Loader 0.07017, and I'm no longer getting errors.
I would class this as a bug, and will report it back.
Thankee for your time, and attention.
-- Ian Stuart
A man depriving some poor village, somewhere, of a first-class idiot.
| [reply] |
