Hi,

I recently needed to generate the dataset's hashed-id from a reporting script outside galaxy (a script that reads Galaxy database directly).

Here's how to generate the hashed-id from perl and postgres, hope it would save you some time if you ever need it. (The python code to do the same is of course already in galaxy, see "encode_id" in galaxy/web/security/__init__.py).

=== Perl === #!/usr/bin/perl use strict; use warnings; use Crypt::Blowfish;

my $id = shift or die "Usage: $0 HISTORY-DATASET-ASSOCIATION-ID\n"; die "Invalid ID ($id)\n" unless $id =~ /^\d+$/;

## This is the same value of "id_secret" from your "universe_wsgi.ini" file my $glx_id_secret = "My_Secret_Code" ;

my $cipher = new Crypt::Blowfish ( $glx_id_secret ); my $padded_str_id = ( '!' x (8- length($id) % 8 ) ). $id ; my $ciphertext = $cipher->encrypt($padded_str_id); my $hash_id = unpack('H*', $ciphertext ) ;

print $hash_id, "\n"; ===============

=== PostgreqSQL === CREATE FUNCTION galaxy_encode_id(HISTORY_DATASET_ASSOCIATION_ID integer, GALAXY_ID_SECRET bytea) RETURNS text AS $$ SELECT substring( encode( encrypt( decode(lpad('', 8-length($1)%8,'!')||$1, 'escape'), $2, 'bf'), 'hex') from 1 for 16 ) ; $$ LANGUAGE SQL; =========== (For postgreSQL, you'll need the "postgres-contrib" package, and run the "pgcrypto.sql" script).

This allows you to run a SQL command that automatically creates the preview URL, e.g.:

Show all datasets belonging to me that are bigger than 1GB, and add a "preview" URL link, so I can quickly check them out: ============ select hda.name, history.name, dataset.file_size, 'http://rave.cshl.edu/galaxy/datasets/' || galaxy_encode_id(hda.id, 'my_secret_id') || '/display?preview=True' as PREVIEW_URL from history_dataset_association hda, galaxy_user, history, dataset where history.user_id = galaxy_user.id AND hda.history_id = history.id AND hda.dataset_id = dataset.id AND hda.deleted = false AND galaxy_user.email ='gordon@cshl.edu' AND dataset.file_size > 1000000000 ============

-gordon