Data::Record::Serialize::Encode::dbi - store a record in a database
version 0.18
use Data::Record::Serialize; my $s = Data::Record::Serialize->new( encode => 'sqlite', ... ); $s->send( \%record );
Data::Record::Serialize::Encode::dbi writes a record to a database using DBI.
It performs both the Data::Record::Serialize::Role::Encode and Data::Record::Serialize::Role::Sink roles.
You cannot construct this directly. You must use "new" in Data::Record::Serialize.
Field types are recognized and converted to SQL types via the following map:
S => 'text' N => 'real' I => 'integer'
By default numeric fields are set to NULL if they are empty. This can be changed by setting the nullify attribute.
NULL
nullify
Records are by default written to the database in batches (see the batch attribute) to improve performance. Each batch is performed as a single transaction. If there is an error during the transaction, record insertions during the transaction are not rolled back.
batch
Transaction errors result in an exception in the Data::Record::Serialize::Error::Encode::dbi::insert class. See Data::Record::Serialize::Error for more information on exception objects.
Data::Record::Serialize::Error::Encode::dbi::insert
queue
$queue = $obj->queue;
The queue containing records not yet successfully transmitted to the database. This is only of interest if "batch" is not 0.
0
Each element is an array containing values to be inserted into the database, in the same order as the fields in "output_fields" in Data::Serialize.
$s->flush;
Flush the queue of records to the database. It returns true if all of the records have been successfully written.
If writing fails:
Writing of records ceases.
The failing record is left at the head of the queue. This ensures that it is possible to retry writing the record.
an exception object (in the Data::Record::Serialize::Error::Encode::dbi::insert class) will be thrown. The failing record (in its final form after formatting, etc) is available via the object's payload method.
payload
If a record fails to be written, it will still be queued for the next attempt at writing to the database. If this behavior is undesired, make sure to remove it from the queue:
use Data::Dumper; if ( ! eval { $output->flush } ) { warn "$@", Dumper( $@->payload ); shift $output->queue->@*; }
As an example of completely flushing the queue while notifying of errors:
use Data::Dumper; until ( eval { $output->flush } ) { warn "$@", Dumper( $@->payload ); shift $output->queue->@*; }
$s->send( \%record );
Send a record to the database. If there is an error, an exception object (with class Data::Record::Serialize::Error::Encode::dbi::insert) will be thrown, and the record which failed to be written will be available via the object's payload method.
If in "batch" mode, the record is queued for later transmission. When the number of records queued reaches that specified by the "batch" attribute, the flush method is called. See "flush" for more information on how errors are handled.
flush
$s->close;
Close the database handle. If writing is batched, records in the queue are written to the database via "flush". An exception will be thrown if a record cannot be written. See "flush" for more details.
As an example of draining the queue while notifying of errors:
use Data::Dumper; until ( eval { $output->close } ) { warn "$@", Dumper( $@->payload ); shift $output->queue->@*; }
This method is called when the object is destroyed. It closes the database handle but does not flush the record queue.
A warning is emitted if the record queue is not empty; turn off the Data::Record::Serialize::Encode::dbi::queue warning to silence it.
Data::Record::Serialize::Encode::dbi::queue
These attributes are available in addition to the standard attributes defined for Data::Record::Serialize->new|Data::Record::Serialize/new>.
dsn
Required The DBI Data Source Name (DSN) passed to DBI. It may either be a string or an arrayref containing strings or arrayrefs, which should contain key-value pairs. Elements in the sub-arrays are joined with =, elements in the top array are joined with :. For example,
=
:
[ 'SQLite', { dbname => $db } ]
is transformed to
SQLite:dbname=$db
The standard prefix of dbi: will be added if not present.
dbi:
table
Required The name of the table in the database which will contain the records. It will be created if it does not exist.
schema
The schema to which the table belongs. Optional.
drop_table
If true, the table is dropped and a new one is created.
create_table
If true, a table will be created if it does not exist.
primary
A single output column name or an array of output column names which should be the primary key(s). If not specified, no primary keys are defined.
db_user
The name of the database user
db_pass
The database password
The number of rows to write to the database at once. This defaults to 100.
If greater than 1, batch rows are cached and then sent out in a single transaction. See "Performance" for more information.
dbitrace
A trace setting passed to DBI.
You can make new bug reports, and view existing ones, through the web interface at https://rt.cpan.org/Public/Dist/Display.html?Name=Data-Record-Serialize.
Please see those modules/websites for more information related to this module.
Data::Record::Serialize
Diab Jerius <djerius@cpan.org>
This software is Copyright (c) 2017 by Smithsonian Astrophysical Observatory.
This is free software, licensed under:
The GNU General Public License, Version 3, June 2007
To install Data::Record::Serialize, copy and paste the appropriate command in to your terminal.
cpanm
cpanm Data::Record::Serialize
CPAN shell
perl -MCPAN -e shell install Data::Record::Serialize
For more information on module installation, please visit the detailed CPAN module installation guide.