Tangram::Type::Dump::Storable(3) map any Perl object as scalar dump via Storable


use Tangram::Core;
use Tangram::Type::Dump::Storable; # always
$schema = Tangram::Schema->new(
classes => { NaturalPerson => { fields => {
storable =>
diary => # diary is a perl hash
col => 'diarydata',
sql => 'BLOB',
indent => 0,
terse => 1,
purity => 0
lucky_numbers => 'int', # use defaults


Maps arbitrary Perl data structures by serializing to a string representation. The persistent fields are grouped in a hash under the "storable" key in the field hash.

Serialization is done by Storable::freeze, which traverses the Perl data structure and creates a binary representation of it. The resulting string will be mapped to the DBMS as a scalar value. During restore, the scalar value will be restored with Storable::thaw to reconstruct the original data structure.

The structure should be able to contain pretty much anything that may be safely dumped by Storable. However, be aware that Tangram::Type::Dump::flatten() must be able to find the persistent objects in the field. Unless you're using classes that are implemented in C and contain other objects, you should be fine (Set::Object has a special work-around).

The field names are passed in a hash that associates a field name with a field descriptor. The field descriptor may be either a hash or a string. The hash uses the following fields:

  • col
  • sql
  • indent
  • terse
  • purity

The optional fields "col" and "sql" specify the column name and the column type for the scalar value in the database. If not present, "col" defaults to the field name and "sql" defaults to VARCHAR(255). Values will be always quoted as they are passed to the database.

The remaining optional fields control the serialization process. They will be passed down to Data::Dumper as values to the corresponding Data::Dumper options. The default settings are: no indentation ("indent=0"), compact format ("terse=1"), and quick dump ("purity=0").


This mapping was contributed by Gabor Herr <[email protected]>