Catmandu(3pm) | User Contributed Perl Documentation | Catmandu(3pm) |
Catmandu - a data toolkit
# From the command line # Convert data from one format to another $ catmandu convert JSON to CSV < data.json $ catmandu convert CSV to YAML < data.csv $ catmandu convert MARC to YAML < data.mrc # Fix data, add, delete, change fields $ catmandu convert JSON --fix 'move_field(title,my_title)' < data.json $ catmandu convert JSON --fix all_my_fixes.txt < data.json # Use a moustache preprocessor on the fix script $ catmandu convert JSON --fix all_my_fixes.txt --var opt1=foo --var opt2=bar < data.json # Import data into a database # Requires: Catmandu::MongoDB and Catmandu::ElasticSearch $ catmandu import YAML to MongoDB --database_name bibliography < data.yml $ catmandu import CSV to ElasticSearch --index_name mystuff < data.csv # Export data from a database # Requires: Catmandu::MongoDB and Catmandu::ElasticSearch $ catmandu export MongoDB --database_name bibliography to YAML > data.yml $ catmandu export ElasticSearch --index_name mystuff to CSV > data.csv # Copy data from one store to another $ catmandu copy MongoDB --database_name mydb to ElasticSearch --index_name mydb # Show the contents of catmandu.yml $ catmandu config # Count items in a store $ catmandu count test1 # Delete items from store $ catmandu delete test1 --query 'title:"My Rabbit"' # run a fix script $ catmandu run myfixes.fix # or, create an executable fix script $ cat myfixes.fix #!/usr/local/bin/catmandu run do importer(OAI,url:"http://biblio.ugent.be/oai") retain(_id) end $ chmod 755 myfixes.fix $ ./myfixes.fix # From Perl use Catmandu; # If you have Catmandu::OAI and Catmandu::MongoDB installed my $importer = Catmandu->importer('OAI',url => 'https://biblio.ugent.be/oai') my $store = Catmandu->store('MongoDB',database_name => 'test'); # Import all the OAI records into MongoDB $store->add_many($importer); # Export all the MongoDB records to YAML and apply some fixes # myfixes.txt: # upcase(title.*) # remove_field(_metadata) # join_field(creator,'; ') # join_field(subject,'-- ') my $fixer = Catmandu->fixer('myfixes.txt'); my $exporter = Catmandu->exporter('YAML'); $exporter->add_many( $fixer->fix($store) ); $exporter->commit;
Catmandu provides a command line client and a Perl API to ease the export (E) transformation (T) and loading (L) of data into databases or data file, ETL in short.
Most of the daily work processing structured data can be done on the command line executing the "catmandu" command. With our catmandu command ETL processing is available in a Perl context. Catmandu is different from other ETL tools by its focus on command line processing with much support for dataformats available in (academic) libraries: MARC, MODS, OAI and SRU. But, also generic formats such as JSON, YAML, CVS, Excel, XML, RDF, Atom are supported.
Read :
The documentation below describes the methods available when including Catmandu as part of a Perl script. For an overview of the command line tool itself read the documentation on catmandu.
To include Catmandu in a Perl script it should be loaded with a "use" command:
use Catmandu;
By default no methods are imported into the Perl context. To import all or some Catmandu methods, provide them as a list to the "use" command:
use Catmandu -all; use Catmandu qw(config store exporter);
Catmandu can load configuration options for exports, importers, fixers via configuration files (see the CONFIG section below). When adding the --load option (optionally with a path) to the "use" command, these configuration files will be loaded at the start of your script.
use Catmandu -load; use Catmandu --load => ['/my/config/directory']; # or use all the options use Catmandu -all -load => [qw(/config/path' '/another/config/path)];
Return the current Log::Any logger.
use Catmandu; use Log::Any::Adapter; use Log::Log4perl; Log::Any::Adapter->set('Log4perl'); # requires Log::Any::Adapter::Log4perl Log::Log4perl::init('./log4perl.conf'); my $logger = Catmandu->log; $logger->info("Starting main program");
with log4perl.conf like:
# Send a copy of all logging messages to STDERR log4perl.rootLogger=DEBUG,STDERR # Logging specific for your main program log4perl.category.myprog=INFO,STDERR # Logging specific for on part of Catmandu log4perl.category.Catmandu::Fix=DEBUG,STDERR # Where to send the STDERR output log4perl.appender.STDERR=Log::Log4perl::Appender::Screen log4perl.appender.STDERR.stderr=1 log4perl.appender.STDERR.utf8=1 log4perl.appender.STDERR.layout=PatternLayout log4perl.appender.STDERR.layout.ConversionPattern=%d [%P] - %p %l time=%r : %m%n
Returns the default location where Catmandu looks for configuration and lib when called with no argument. Sets the default location if a path is given. The default load path is the script directory or it's parent if the script directory is "bin".
Load all the configuration options in the catmandu.yml configuration file. See CONFIG below for extended examples of configuration options.
Load all the configuration options stored at alternative paths.
A load path ':up' will search upwards from your program for configuration.
See CONFIG below for extended examples of configuration options.
Returns an ARRAYREF of paths where configuration was found. Note that this list is empty before "load".
Returns the first path where configuration was found. Note that this is "undef" before "load".
Returns the current configuration as a HASHREF.
Set a new configuration and reload the environment.
Return the name of the default store.
Return an instance of Catmandu::Store. The NAME is a name of a Catmandu::Store or the name of a store configured in a catmandu.yml configuration file. When no NAME is given, the 'default' store in the configuration file will be used.
E.g. if the configuration file 'catmandu.yml' contains:
store: default: package: ElasticSearch options: index_name: blog test: package: Mock
then in your program:
# This will use ElasticSearch my $store = Catmandu->store('ElasticSearch', index_name => 'blog'); # or because we have a 'default' set in the configuration file my $store = Catmandu->store('default'); # or because 'default' will be used when no name was provided my $store = Catmandu->store; # This will use Mock my $store = Catmandu->store('test');
Configuration settings can be overwritten by the store command:
my $store2 = Catmandu->store('default', index_name => 'test2');
Return the name of the default fixer.
Return an instance of Catmandu::Fix. NAME can be the name of a fixer section in a catmandu.yml file. Or, one or more Catmandu::Fix-es can be provided inline.
E.g. if the configuration file 'catmandu.yml' contains:
fixer: default: - do_this() - do_that()
then in your program al these lines below will create the same fixer:
my $fixer = Catmandu->fixer('do_this()', 'do_that()'); my $fixer = Catmandu->fixer(['do_this()', 'do_that()']); my $fixer = Catmandu->fixer('default'); my $fixer = Catmandu->fixer(); # The default name is 'default'
FIX-es can be also written to a Fix script. E.g. if myfixes.txt contains:
do_this() do_that()
then the above code will even be equivalent to:
my $fixer = Catmandu->fixer('myfixes.txt');
Return the name of the default importer.
Return the name of the default importer package if no package name is given in the config or as a param.
Return an instance of Catmandu::Importer. The NAME is a name of a Catmandu::Importer or the name of a importer configured in a catmandu.yml configuration file. When no NAME is given, the 'default' importer in the configuration file will be used.
E.g. if the configuration file 'catmandu.yml' contains:
importer: default: package: OAI options: url: http://www.instute.org/oai/
then in your program all these lines will be equivalent:
my $importer = Catmandu->importer('OAI', url => 'http://www.instute.org/oai/'); my $importer = Catmandu->importer('default'); my $importer = Catmandu->importer(); # The default name is 'default'
Configuration settings can be overwritten by the importer command:
my $importer2 = Catmandu->importer('default', url => 'http://other.institute.org');
Return the name of the default exporter.
Return the name of the default exporter package if no package name is given in the config or as a param.
Return an instance of Catmandu::Exporter with name NAME (or the default when no name is given). The NAME can be in a configuration file (see 'importer').
Return an instance of Catmandu::Validator with name NAME (or the default when no name is given). The NAME can be in a configuration file (see 'importer').
Export data using a default or named exporter or exporter instance.
Catmandu->export({ foo=>'bar'}); my $importer = Catmandu::Importer::Mock->new; Catmandu->export($importer, 'YAML', file => '/my/file'); Catmandu->export($importer, 'my_exporter'); Catmandu->export($importer, 'my_exporter', exporter_option => '...' , ...); Catmantu->export($importer, Catmandu::Exporter::YAML->new);
Export data using a default or named exporter to a string.
my $importer = Catmandu::Importer::Mock->new; my $yaml = Catmandu->export_to_string($importer, 'YAML'); # is the same as my $yaml = ""; Catmandu->export($importer, 'YAML', file => \$yaml);
Import data from a string using a default or named importer. Return value should be an array of hashes.
my $json = qq([{"name":"Nicolas"}]); { my $record = Catmandu->import_from_string( $json, "JSON" ); } # is the same as { my $record = Catmandu->importer('JSON', file => \$json)->to_array() }
Configure a new named importer.
Catmandu->define_importer(books => CSV => (fields => 'title,author,publisher')); Catmandu->importer(books => (file => 'mybooks.csv'))->each(sub { my $book = shift; say $book->{title}; }); # this is equivalent to Catmandu->config->{importer}{books} = { package => 'CSV', options => { fields => 'title,author,publisher', }, }
Configure a new named exporter.
Catmandu->define_exporter('books', 'CSV', fix => 'capitalize(title)'); my $csv = Catmandu->export_to_string({title => 'nexus'}, 'books'); # this is equivalent to Catmandu->config->{exporter}{books} = { package => 'CSV', options => { fix => 'capitalize(title)', }, }
Configure a new named store.
Catmandu->define_store(mydb => MongoDB => (database_name => 'mydb')); Catmandu->store->bag('books')->get(1234); # this is equivalent to Catmandu->config->{store}{mydb} = { package => 'MongoDB', options => { database_name => 'mydb', }, }
Configure a new named fixer.
Catmandu->define_fixer('cleanup', [ 'trim(title)', 'capitalize(title)', 'remove_field(junk)', # ... ]); Catmandu->fixer('cleanup')->fix($record);
use Catmandu -load; use Catmandu -load => []; # is the same as Catmandu->load; use Catmandu -load => ['/config/path']; # is the same as Catmandu->load('/config/path');
Catmandu configuration options can be stored in files in the root directory of your programming project. The file can be YAML, JSON or Perl and is called "catmandu.yml", "catmandu.json" or "catmandu.pl". In this file you can set the default Catmandu stores and exporters to be used. Here is an example of a "catmandu.yml" file:
store: default: package: ElasticSearch options: index_name: myrepository exporter: default: package: YAML
For large configs it's more convenient to split the config into several files. You can do so by having multiple config files starting with catmandu*.
catmandu.general.yml catmandu.db.yml ...
Split config files are processed and merged by Config::Onion.
Config files can indicate a path under which their keys will be nested. This makes your configuration more readable by keeping indentation to a minimum.
A config file containing
_prefix: foo: bar: baz: 1
will be loaded as
foo: bar: baz: 1
See Config::Onion for more information on how this works.
Nicolas Steenlant, "<nicolas.steenlant at ugent.be>"
Magnus Enger, "magnus at enger.priv.no"
Nicolas Franck, "nicolas.franck at ugent.be"
Patrick Hochstenbach, "patrick.hochstenbach at ugent.be"
Vitali Peil, "vitali.peil at uni-bielefeld.de"
Christian Pietsch, "christian.pietsch at uni-bielefeld.de"
Dave Sherohman, "dave.sherohman at ub.lu.se"
Jakob Voss, "nichtich at cpan.org"
Snorri Briem, "snorri.briem at ub.lu.se"
Johann Rolschewski, "jorol at cpan.org"
Pieter De Praetere, "pieter.de.praetere at helptux.be"
Doug Bell
Upsana, "me at upasana.me"
Stefan Weil
Tom Hukins
For any questions on the use of our modules please join our mailing list at:
librecat-dev@lists.uni-bielefeld.de
or send in your bug reports or feature requests to our issue tracker at:
https://github.com/LibreCat/Catmandu/issues
This program is free software; you can redistribute it and/or modify it under the terms of either: the GNU General Public License as published by the Free Software Foundation; or the Artistic License.
See <http://dev.perl.org/licenses/> for more information.
2020-06-06 | perl v5.30.3 |