NAME
MPX::RIF - build cheap mpx from filenames etc.
VERSION
version 0.027
SYNOPSIS
Read a yaml config file, parse a directory, extract information from
filepath write it in human-readable format for debugging and write MPX
Mulitmediaobjekt information (XML).
use MPX::RIF;
my $faker = MPX::RIF->new($config);
my $faker->run; #run calls all the steps in the normal order
#alternatively you can call the steps yourself
#1st step
$faker->scandir();
#2nd step
$faker->parsedir();
#3rd step
$faker->lookupObjId();
#4th step
$faker->filter();
#5th step
$faker->writeXML();
#6th step
$faker->validate();
#Everything else is considered to be the private parts of this module.
#For more info see the method run (overview) and the individual methods.
METHODS
my $faker=MPX::RIF::new ($CONFIG);
REQUIRED config is the path to a yaml configuration file.
CONIG=>'/path/to/config.yml'
Config file parameters are described inside the example config.
OPTIONAL BEGIN => 1 #makes MPX::RIF start with yml file 0 for off 1 read
scandir from yml file 2 read parsedir from yml file 3 read lookup from
yml file DEBUG=>1, # turns debug messages on/off; 1 for on - 0 for off
STOP=> 1, # stop after step 1, 0 don't stop 1 stop after step 1, 2 stop
after step 2, 3 stop after step 3, 4 and higher - ignored (same as 0)
$faker->lookupObjId ('path/to/big-file.mpx');
3rd step. Associate each resource with an objId. Also filter stuff out
that does not have required information.
$faker->$filter;
4th step. Drops resources from store if they don't have specified keys.
a new step
my $newpath=$self->filemover ($oldpath);
primitive filemover to eliminate non-unique mulIds. Use at own risk!
$faker->parsedir
Second step.
Will call extension to extract information from file name and path. Also
adds constants.
Naming convention: Every file that is recognized by the find rules
specified in the config.yml is treated as an object with one or several
features.
$faker->run();
Executes all steps one after according to configuration. See mpx-rif.pl
for high-level description.
$faker->scandir
First step. Just scans the directory according to info from
configuration file. It saves info into a yml file (1-scandir.yml) for
manual proof reading. Use the STOP option during initialization to abort
after this step, e.g. MPX::RIF->new (STOP=>1);
$self->validate();
Validate resulting mpx and check for duplicate mulIds. Log errors.
$self->writeXML();
TODO: maybe I should check if an resource is complete before I xml-ify it
$self->stop ($location);
Location is the number of the step where stop is called from. If
location matches the $self->{STOP}, MPX::RIF stops gracefully and
outputs an a log and or debug message (if debug and log are on).
my $ret=$faker->_dirparser ($path);
Calls the dirparser callback specified in config.yml. Expects a single
path. Returns a hashref representing one object. If something is
returned the result is saved in $self->{data};
my $object={ key=>value, feature=>blue, }
$self->_loadStore('path/to/store.yml');
$self->_dumpStore('path/to/store.yml');
$self->_storeResource ($resource);
Stores the resource in the data store. Requires resource to have an id.
If I don't want redundancy, i.e. the id twice, I need to extract it when
I save it and reinstate when I get it back. Sofar I, only access from
_resources.
@arr=$self->_resourceIds();
OLD: Don't know how to do this:
Returns one resource at a time from the resource store. Use in while
(preferred) or foreach, e.g.:
foreach my $resource ($self->_resources()) {
#bla
}
my $resource=$self->_getResource ($id);
FUNCTIONS
my $xpc=registerNS ($doc);
Expects DOM or doc, never sure about it. Returns xpc.
Register prefix mpx with http://www.mpx.org/mpx
RATIONALE
There are many images. It can take a long to enter them manually in the
database. For each item, there are many repetitive information items,
e.g. 1000 fotos were made by the same fotographer.
This little perl tool parses a directory and writes XML/MPX with the
metadata. It is good with repetative metadata. Of course, this is not a
silver bullett. It is a just a cheap solution that works only if you
know your photos well.
The whole process is broken down in several consecutive steps. State
information is dumped a couple of times during executing as yaml, to
facilitate proof-reading and error checking. There are debug messages
and log messages which should help you finding quirks in your data.
The tool is configurable. It's written in a haste i.e. no great code,
but it should at least be readable.
my $objId=$self->_lookupObjId;
return () on failure.
INTERNAL INTERFACE
AUTHOR
Maurice Mengel <mauricemengel@gmail.com>
COPYRIGHT AND LICENSE
This software is copyright (c) 2011 by Maurice Mengel.
This is free software; you can redistribute it and/or modify it under
the same terms as the Perl 5 programming language system itself.