DBIx::Class::Schema::Loader dbicdump 是否可以为 JSYNC 自动添加序列化程序

Question

I'm running

dbicdump -o dump_directory=./lib \
          -o components='["InflateColumn::DateTime", "InflateColumn::Serializer"]' \
          -o debug=1 \
          -o db_schema=foo \
          -o qualify_objects=1 \
          Foo::Schema \
          'dbi:Pg:dbname=foo' username password

against a Postgres database with the hopes that the generated schema will generate the code necessary to inflate timestamp and JSON column types.

Looking at the generated Schema files, I see the

__PACKAGE__->load_components("InflateColumn::Serializer", "InflateColumn::DateTime");

line, but now I have to add serializer_class => 'JSYNC' to each of the JSON columns I want to inflate and deflate.

Is there a way to get dbicdump to do this automatically or is it expected practice to manually set these accessors?

Thanks!

Answer 1

Not sure this is possible from the dbicdump program, but if you use the the following perl program, you can do exactly what I want:

#!/usr/bin/env perl
use strict;
use warnings;
use DBIx::Class::Schema::Loader qw/make_schema_at/;

make_schema_at(
    'Foo::Schema',
    {
        dump_directory => './lib',
        components     => [ "InflateColumn::Serializer", "InflateColumn::DateTime" ],
        db_schema      => 1,
        qualify_objects => 1,
        custom_column_info => sub {
            my ($table, $column_name, $column_info) = @_;
            if ( $column_info->{data_type} eq "json" ) {
                return { serializer_class => "JSYNC" };
            }
        },
    },
    [
         "dbi:Pg:dbname=foo',
         "username",
         "password"
     ]
);

Hope this saves someone some research time.