Registry.pm

#
# Ensembl module for Registry
#
# Copyright EMBL/EBI
##
# You may distribute this module under the same terms as perl itself

# POD documentation - main docs before the code

=head1 NAME

Bio::EnsEMBL::Registry

=head1 SYNOPSIS

Bio::EnsEMBL::Registry->load_all("configuration_file");

$gene_adaptor =
  Bio::EnsEMBL::Registry->get_adaptor( "human", "core", "gene" );


=head1 DESCRIPTION

All Adaptors are stored/registered using this module. This module should
then be used to get the adaptors needed.

The registry can be loaded from a configuration file using the load_all
method.

If a filename is passed to load_all then this is used.  Else if the
enviroment variable ENSEMBL_REGISTRY is set to the name on an existing
configuration file, then this is used.  Else if the file .ensembl_init
in your home directory exist, it is used.

For the Web server ENSEMBL_REGISTRY should be set in SiteDefs.pm.  This
will then be passed on to load_all.


The registry can also be loaded via the method load_registry_from_db
which given a database host will load the latest versions of the Ensembl
databases from it.

The four types of registries are for db adaptors, dba adaptors, dna adaptors
and the standard type.

=head2 db

These are registries for backwards compatibility and enable the subroutines
to add other adaptors to connections. 

e.g. get_all_db_adaptors, get_db_adaptor, add_db_adaptor, remove_db_adaptor
are the old DBAdaptor subroutines which are now redirected to the Registry.

So if before we had

    my $sfa = $self->adaptor()->db()->get_db_adaptor('blast');

We now want to change this to

    my $sfa =
      Bio::EnsEMBL::Registry->get_adaptor( "human", "core", "blast" );


=head2 DBA

These are the stores for the DBAdaptors

The Registry will create all the DBConnections needed now if you set up the
configuration correctly. So instead of the old commands like

    my $db           = Bio::EnsEMBL::DBSQL::DBAdaptor->new(...);
    my $exon_adaptor = $db->get_ExonAdaptor;

we should now have just

    my $exon_adaptor =
      Bio::EnsEMBL::Registry->get_adaptor( "human", "core", "exon" );


=head2 DNA

This is an internal Registry and allows the configuration of a dnadb. 
An example here is to set the est database to get its dna data from the core database.

    ## set the est db to use the core for getting dna data.
    # Bio::EnsEMBL::Utils::ConfigRegistry->dnadb_add(
    #         "Homo Sapiens", "core", "Homo Sapiens", "est" );


=head2 adaptors

This is the registry for all the general types of adaptors like GeneAdaptor, ExonAdaptor, 
Slice Adaptor etc.

These are accessed by the get_adaptor subroutine i.e.

    my $exon_adaptor =
      Bio::EnsEMBL::Registry->get_adaptor( "human", "core", "exon" );

=head1 CONTACT

Post questions to the Ensembl developer list: <ensembl-dev@ebi.ac.uk>


=head1 METHODS

=cut


package Bio::EnsEMBL::Registry;

use strict;

use Bio::EnsEMBL::DBSQL::DBAdaptor;
use Bio::EnsEMBL::Utils::Exception qw( deprecate throw warning );
use Bio::EnsEMBL::Utils::Argument qw(rearrange);
use Bio::EnsEMBL::Utils::ConfigRegistry;
use DBI;

use vars qw(%registry_register);

my $API_VERSION = 51;

=head2 load_all

 Will load the registry with the configuration file which is obtained
 from the first in the following and in that order.

  1) If an argument is passed to this method, this is used as the name
     of the configuration file to read.

  2) If the enviroment variable ENSEMBL_REGISTRY is set, this is used as
     the name of the configuration file to read.

  3) If the file .ensembl_init exist in the home directory, it is used
     as the configuration file.

  Arg [1]    : (optional) string
               Name of file to load the registry from.
  Arg [2]    : (optional) integer
               If not 0, will print out all information.
  Arg [3]    : (optional) integer
               If not 0, the db connection will not be cleared, if 0 or
               if not set the db connections will be cleared (this is
               the default).
  Arg [4]:     (optional) int 1
               This option will turn off caching for slice features, so, 
               every time a set of features is retrieved, they will come from
               the database instead of the cache. This option is only recommended
               for advanced users, specially if you need to store and retrieve
               features. It might reduce performance when querying the database if 
               not used properly. If in doubt, do not use it or ask in ensembl-dev            
  Example    : Bio::EnsEMBL::Registry->load_all();
  Returntype : none
  Exceptions : none
  Status     : Stable

=cut

sub load_all {
    my $class = shift;
    my ( $config_file, $verbose, $no_clear, $no_cache ) = @_;

    $config_file ||= $ENV{ENSEMBL_REGISTRY}
      || $ENV{HOME} . "/.ensembl_init";

    $verbose  ||= 0;
    $no_clear ||= 0;
    $no_cache ||= 0;

    if ( !defined($config_file) ) {
        if ($verbose) {
            print( STDERR
                   "No default registry configuration to load.\n" );
        }
    } elsif ( !-e $config_file ) {
        if ($verbose) {
            printf( STDERR "Configuration file '%s' does not exist. "
                      . "Registry configuration not loaded.\n",
                    $config_file );
        }
    } else {
        if ( defined( $registry_register{'seen'} ) ) {
            if ( !$no_clear ) {
                if ($verbose) {
                    print( STDERR "Clearing previously loaded "
                           . "registry configuration\n" );
                }
                $class->clear();
            }
        }
        $registry_register{'seen'} = 1;

        if ($verbose) {
            printf( STDERR
                      "Loading registry configuration from '%s'.\n",
                    $config_file );
        }

        my $cfg;

        eval { require Config::IniFiles };
        if ($@) {
          # The user does not have the 'Config::IniFiles' module.
          if ($verbose) {
            print( STDERR "No Config::IniFiles module found, "
                   . "assuming this is not an ini-file\n" );
          }
          # If the configuration file *is* an ini-file, we can expect a
          # load of compilation errors from the next eval...
        } else {
          # The user has the 'Config::IniFiles' module installed.  See
          # if this is an ini-file or not...
          $cfg = Config::IniFiles->new( -file => $config_file );
        }

        if ( defined $cfg ) {
            # This is a map from group names to Ensembl DB adaptors.
            my %group2adaptor = (
                 'blast'   => 'Bio::EnsEMBL::External::BlastAdaptor',
                 'compara' => 'Bio::EnsEMBL::Compara::DBSQL::DBAdaptor',
                 'core'    => 'Bio::EnsEMBL::DBSQL::DBAdaptor',
                 'estgene' => 'Bio::EnsEMBL::DBSQL::DBAdaptor',
                 'funcgen' => 'Bio::EnsEMBL::Funcgen::DBSQL::DBAdaptor',
                 'haplotype' =>
                   'Bio::EnsEMBL::ExternalData::Haplotype::DBAdaptor',
                 'hive' => 'Bio::EnsEMBL::Hive::DBSQL::DBAdaptor',
                 'lite' => 'Bio::EnsEMBL::Lite::DBAdaptor',
                 'otherfeatures' => 'Bio::EnsEMBL::DBSQL::DBAdaptor',
                 'pipeline' =>
                   'Bio::EnsEMBL::Pipeline::DBSQL::DBAdaptor',
                 'snp' =>
                   'Bio::EnsEMBL::ExternalData::SNPSQL::DBAdaptor',
                 'variation' =>
                   'Bio::EnsEMBL::Variation::DBSQL::DBAdaptor',
                 'vega' => 'Bio::EnsEMBL::DBSQL::DBAdaptor' );

            my %default_adaptor_args = ();

            if ( $cfg->SectionExists('default') ) {
                # The 'default' section is special.  It contain default
                # values that should be implicit to all other section in
                # this configuration file.  Aliases are added if there
                # is also a 'species' setting.

                my $alias = $cfg->val( 'default', 'alias' );
                $cfg->delval( 'default', 'alias' );

                my $species = $cfg->val( 'default', 'species' );

                if ( defined($alias) && defined($species) ) {
                    Bio::EnsEMBL::Utils::ConfigRegistry->add_alias(
                                     -species => $species,
                                     -alias => [ split( /\n/, $alias ) ]
                    );
                }

                %default_adaptor_args =
                  map { '-' . $_ => $cfg->val( 'default', $_ ) }
                  $cfg->Parameters('default');
            }

            foreach my $section ( $cfg->Sections() ) {
                if ( $section eq 'default' )
                {    # We have already done the 'default' section.
                    next;
                }

                my $group = $cfg->val( $section, 'group' )
                  || $cfg->val( 'default', 'group' );

                if ( !defined($group) ) {
                    printf( STDERR "Key 'group' is undefined "
                              . "for configuration section '%s', "
                              . "skipping this section.\n",
                            $section );
                    next;
                }

                my $adaptor = $group2adaptor{ lc($group) };
                if ( !defined($adaptor) ) {
                    printf( STDERR "Unknown group '%s' "
                              . "for configuration section '%s', "
                              . "skipping this section.\n",
                            $group, $section );
                    next;
                }

                # Handle aliases.  A section must have both an 'alias'
                # setting and a 'species' setting for aliases to be
                # added.  The 'species' setting might be inherited from
                # the 'default' section.

                my $alias = $cfg->val( $section, 'alias' );
                $cfg->delval( $section, 'alias' );

                my $species = $cfg->val( $section, 'species' )
                  || $cfg->val( 'default', 'species' );

                if ( defined($alias) && defined($species) ) {
                    Bio::EnsEMBL::Utils::ConfigRegistry->add_alias(
                                     -species => $species,
                                     -alias => [ split( /\n/, $alias ) ]
                    );
                }

                # Fill in the adaptor initialization arguments.
                # We trust the user to provide sensible key-value pairs.
                my %adaptor_args = %default_adaptor_args;
                foreach my $parameter ( $cfg->Parameters($section) ) {
                    $adaptor_args{ '-' . $parameter } =
                      $cfg->val( $section, $parameter );
		    #when set, do not use the feature cache in the different adaptors
		    if ($no_cache){
			$adaptor_args{'-no_cache'} = 1;
		    }
                }

                if ($verbose) {
                    printf( "Configuring adaptor '%s' "
                              . "for configuration section '%s'...\n",
                            $adaptor, $section );
                }

                eval "require $adaptor";
                if ($@) { die($@) }

                $adaptor->new(%adaptor_args);

            } ## end foreach my $section ( $cfg->Sections...
        } else {
            # This is probably no ini-file but an old style piece
            # of configuration written in Perl.  We need to try to
            # require() it.

            eval { require($config_file) };
            if ($@) { die($@) }

            # To make the web code avoid doing this again:
            delete $INC{$config_file};
        }
    } ## end else [ if ( !defined($config_file...
} ## end sub load_all

=head2 clear

 Will clear the registry and disconnect from all databases.

  Example    : Bio::EnsEMBL::Registry->clear();
  Returntype : none
  Exceptions : none
  Status     : Stable

=cut

sub clear{
  my ($self);
  
  foreach my $dba (@{$registry_register{'_DBA'}}){
    if($dba->dbc->connected){
      $dba->dbc->db_handle->disconnect();
    }
  }
  %registry_register = ();
}

#
# db adaptors. (for backwards compatibility)
#

=head2 add_db

  Arg [1]    : db (DBAdaptor) to add adaptor to.
  Arg [2]    : name of the name to add the adaptor to in the registry.
  Arg [3]    : The adaptor to be added to the registry.
  Example    : Bio::EnsEMBL::Registry->add_db($db, "lite", $dba);
  Returntype : none
  Exceptions : none
  Status     : At Risk.
             : This is here for backwards compatibility only and may be removed 
             : eventually. Solution is to make sure the db and the adaptor have
             : the same species and the call is then no longer needed.
             
=cut

sub add_db{
  my ($class, $db, $name, $adap) = @_;


  if(lc($db->species()) ne lc($adap->species)){
    $registry_register{lc($db->species())}{lc($db->group())}{'_special'}{lc($name)} = $adap;
  }
}

=head2 remove_db

  Arg [1]    : db (DBAdaptor) to remove adaptor from.
  Arg [2]    : name to remove the adaptor from in the registry.
  Example    : my $db = Bio::EnsEMBL::Registry->remove_db($db, "lite");
  Returntype : adaptor
  Exceptions : none
  Status     : At Risk.
             : This is here for backwards compatibility only and may be removed 
             : eventually. Solution is to make sure the db and the adaptor have
             : the same species and the call is then no longer needed.

=cut

sub remove_db{
  my ($class, $db, $name) = @_;

  my $ret = $registry_register{lc($db->species())}{lc($db->group())}{'_special'}{lc($name)};
  $registry_register{lc($db->species())}{lc($db->group())}{'_special'}{lc($name)} = undef;

  return $ret;
}

=head2 get_db

  Arg [1]    : db (DBAdaptor) to get adaptor from.
  Arg [2]    : name to get the adaptor for in the registry.
  Example    : my $db = Bio::EnsEMBL::Registry->get_db("Human", "core", "lite");
  Returntype : adaptor
  Exceptions : none
  Status     : At Risk.
             : This is here for backwards compatibility only and may be removed 
             : eventually. Solution is to make sure the db and the adaptor have
             : the same species then call get_DBAdaptor instead.

=cut

sub get_db{
  my ($class, $db, $name) = @_;

  my $ret = Bio::EnsEMBL::Registry->get_DBAdaptor(lc($db->species),lc($name));

  if(defined($ret)){
    return $ret;
  }
  return $registry_register{lc($db->species())}{lc($db->group())}{'_special'}{lc($name)};
}

=head2 get_all_db_adaptors

  Arg [1]    : db (DBAdaptor) to get all the adaptors from.
  Example    : my $db = Bio::EnsEMBL::Registry->get_all_db_adaptors($db);
  Returntype : adaptor
  Exceptions : none
  Status     : At Risk.
             : This is here for backwards compatibility only and may be removed 
             : eventually. Solution is to make sure the dbs all have
             : the same species then call get_all_DBAdaptors(-species => "human");


=cut

sub get_all_db_adaptors {
  my ( $class, $db ) = @_;
  my %ret = ();

  # we now also want to add all the DBAdaptors for the same species.
  # as add_db_adaptor does not add if it is from the same species.

  foreach my $dba ( @{ $registry_register{'_DBA'} } ) {
    if ( lc( $dba->species() ) eq lc( $db->species() ) ) {
      $ret{ $dba->group() } = $dba;
    }
  }

  foreach my $key (
             keys %{
               $registry_register{ $class->get_alias( $db->species() ) }
                 { lc( $db->group() ) }{'_special'} } )
  {
    $ret{$key} =
      $registry_register{ $class->get_alias( $db->species() ) }
      { lc( $db->group() ) }{'_special'}{$key};
  }

  return \%ret;
} ## end sub get_all_db_adaptors


#
# DBAdaptors
#

=head2 add_DBAdaptor

  Arg [1]    : name of the species to add the adaptor to in the registry.
  Arg [2]    : name of the group to add the adaptor to in the registry.
  Arg [3]    : The DBAaptor to be added to the registry.
  Example    : Bio::EnsEMBL::Registry->add_DBAdaptor("Human", "core", $dba);
  Returntype : none
  Exceptions : none
  caller     : internal
  Status     : Stable

=cut

sub add_DBAdaptor{
  my ($class, $species, $group, $adap) = @_;

  if(!($class->alias_exists($species))){
    $class->add_alias($species,$species);
  }
  

  $species = $class->get_alias($species);

  $registry_register{$species}{lc($group)}{'_DB'} = $adap;

  if(!defined($registry_register{'_DBA'})){
    my @list =();
    push(@list,$adap);
    $registry_register{'_DBA'}= \@list;
  }
  else{
    push(@{$registry_register{'_DBA'}},$adap);
  }

}



=head2 get_DBAdaptor

  Arg [1]    : name of the species to get the adaptor for in the registry.
  Arg [2]    : name of the group to get the adaptor for in the registry.
  Example    : $dba = Bio::EnsEMBL::Registry->get_DBAdaptor("Human", "core");
  Returntype : DBAdaptor
  Exceptions : none
  Status     : Stable

=cut

sub get_DBAdaptor{
  my ($class, $species, $group) = @_;

  $species = $class->get_alias($species);

  return  $registry_register{$species}{lc($group)}{'_DB'};

}

=head2 get_all_DBAdaptors

  Arg [SPECIES]: (optional) string 
                  species name to get adaptors for
  Arg [GROUP]  : (optional) string 
                  group name to get adaptors for
  Example      : @dba = @{Bio::EnsEMBL::Registry->get_all_DBAdaptors()};
               : @human_dbas = @{Bio::EnsEMBL::Registry->get_all_DBAdaptors(-species => 'human')};
  Returntype   : list of DBAdaptors
  Exceptions   : none
  Status       : Stable

=cut

sub get_all_DBAdaptors{
  my ($class,@args)=@_;
  my @ret;

  my ($species, $group) = 
    rearrange([qw(SPECIES GROUP)], @args);
  if(defined($species)){
    $species = $class->get_alias($species);
  }
  foreach my $dba (@{$registry_register{'_DBA'}}){
    if(!defined($species) || lc($species) eq lc($dba->species)){
      if(!defined($group) || lc($group) eq lc($dba->group)){
	push @ret, $dba;
      }
    }
  }


  return \@ret;
}

=head2 get_all_DBAdaptors_by_connection

  Arg [1]    :dbconnection to use to find DBAdaptors
  Returntype : reference to list of DBAdaptors
  Exceptions : none.
  Example    : @dba = @{Bio::EnsEMBL::Registry->get_all_DBAdaptors_by_connection($dbc);
  Status     : Stable

=cut

sub get_all_DBAdaptors_by_connection{
  my ($self, $dbc_orig) = @_;
  my @return;

  foreach my $dba ( @{$registry_register{'_DBA'}}){
    my $dbc = $dba->dbc;
    if($dbc && $dbc->can('equals') && $dbc->equals($dbc_orig)){
      push @return, $dba;
    }
  }
  return \@return;
}

=head2 remove_DBAdaptor

  Arg [1]    : name of the species to get the adaptor for in the registry.
  Arg [2]    : name of the group to get the adaptor for in the registry.
  Example    : $dba = Bio::EnsEMBL::Registry->remove_DBAdaptor("Human", "core");
  Returntype : none
  Exceptions : none
  Status     : At risk

=cut

sub remove_DBAdaptor{
  my ($class, $species, $group) = @_;

  $species = $class->get_alias($species);
  
 
  delete $registry_register{$species}{$group};
  # This will remove the DBAdaptor and all the other adaptors

  # Now remove if from the _DBA array
  my $index;


  foreach my $i(0..$#{$registry_register{'_DBA'}}){
    my $dba = $registry_register{'_DBA'}->[$i];

    if(($dba->species eq $species) &&
       $dba->group eq $group){
      $index = $i;

      last;
    }
  }
  
  # Now remove from _DBA cache
  splice(@{$registry_register{'_DBA'}}, $index, 1);

  return;
}



=head2 reset_DBAdaptor

  Arg [1]:     string - species e.g. homo_sapiens
  Arg [2]:     string - DB group e.g. core
  Arg [3]:     string - new dbname
  Args [4-7]:  string - optional DB parameters, defaults to current db params if omitted
  Usage :      $reg->reset_registry_db('homo_sapiens', 'core', 'homo_sapiens_core_37_35j');
  Description: Resets a DB within the registry.
  Exceptions:  Throws if mandatory params not supplied
               Throws if species name is not already seen by the registry
               Throws if no current DB for species/group available
  Status :     At risk

=cut

sub reset_DBAdaptor{
  my ($self, $species, $group, $dbname, $host, $port, $user, $pass) = @_;

  # Check mandatory params
  if(! (defined $species && defined $group && defined $dbname)){
	throw('Must provide at least a species, group and dbname parmeter to redefine a DB in the registry');
  }
  
  # Validate species here
  my $alias = $self->get_alias($species);
  throw("Could not find registry alias for species:\t$species") if(! defined $alias);
 

  # Get all current defaults if not defined
  my $current_db = $self->get_DBAdaptor($alias, $group);
  
  if(! defined $current_db){
	throw("There is not current registry DB for:\t${alias}\t${group}");
  }


  $host ||= $current_db->dbc->host;
  $port ||= $current_db->dbc->port;
  $user ||= $current_db->dbc->username;
  $pass ||= $current_db->dbc->password;
  my $class = ref($current_db);

  $self->remove_DBAdaptor($alias, $group);
  
  my @adaptors = @{$self->get_all_adaptors};


  # ConfigRegistry should automatically add this to the Registry
  my $db = $class->new(
					   -user => $user,
					   -host => $host,
					   -port => $port,
					   -pass => $pass,
					   -dbname => $dbname,
					   -species => $alias,
					   -group    => $group,
					  );

  return $db;
}


#
# DNA Adaptors
#

=head2 add_DNAAdaptor

  Arg [1]    : name of the species to add the adaptor to in the registry.
  Arg [2]    : name of the group to add the adaptor to in the registry.
  Arg [3]    : name of the species to get the dna from
  Arg [4]    : name of the group to get the dna from
  Example    : Bio::EnsEMBL::Registry->add_DNAAdaptor("Human", "estgene", "Human", "core");
  Returntype : none
  Exceptions : none
  Status     : Stable

=cut

sub add_DNAAdaptor{
  my ($class, $species, $group, $dnadb_species, $dnadb_group) = @_;

  $species = $class->get_alias($species);
  $dnadb_species = $class->get_alias($dnadb_species);
  if($dnadb_group->isa('Bio::EnsEMBL::DBSQL::DBAdaptor')){
    deprecated("");
  }
  else{
    $registry_register{$species}{lc($group)}{'_DNA'} = $dnadb_group;
    $registry_register{$species}{lc($group)}{'_DNA2'} = $dnadb_species;
  }
}

=head2 get_DNAAdaptor

  Arg [1]    : name of the species to get the adaptor for in the registry.
  Arg [2]    : name of the group to get the adaptor for in the registry.
  Example    : $dnaAdap = Bio::EnsEMBL::Registry->get_DNAAdaptor("Human", "core");
  Returntype : adaptor
  Exceptions : none
  Status     : Stable

=cut

sub get_DNAAdaptor{
  my ($class, $species, $group) = @_;

  $species = $class->get_alias($species);
  my $new_group = $registry_register{$species}{lc($group)}{'_DNA'};
  my $new_species = $registry_register{$species}{lc($group)}{'_DNA2'};
  if( defined $new_group ) {
    return  $class->get_DBAdaptor($new_species,$new_group);
  } else {
    return undef;
  }
}

#
# General Adaptors
#

=head2 add_adaptor

  Arg [1]    : name of the species to add the adaptor to in the registry.
  Arg [2]    : name of the group to add the adaptor to in the registry.
  Arg [3]    : name of the type to add the adaptor to in the registry.
  Arg [4]    : The DBAaptor to be added to the registry.
  Arg [5]    : (optional) if set okay to overwrite.
  Example    : Bio::EnsEMBL::Registry->add_adaptor("Human", "core", "Gene", $adap);
  Returntype : none
  Exceptions : none
  Caller     : internal
  Status     : Stable


=cut

sub add_adaptor {
  my ( $class, $species, $group, $type, $adap, $reset ) = @_;

  $species = $class->get_alias($species);

  # Since the adaptors are not stored initially, only their class paths
  # when the adaptors are obtained, we need to store these instead.  It
  # is not necessarily an error if the registry is overwritten without
  # the reset set but it is an indication that we are overwriting a
  # database which should be a warning for now

  if ( defined($reset) )
  {    # JUST REST THE HASH VALUE NO MORE PROCESSING NEEDED
    $registry_register{$species}{ lc($group) }{ lc($type) } = $adap;
    return;
  }
  if (
    defined( $registry_register{$species}{ lc($group) }{ lc($type) } ) )
  {
  # print STDERR (
  #      "Overwriting Adaptor in Registry for $species $group $type\n");
    $registry_register{$species}{ lc($group) }{ lc($type) } = $adap;
    return;
  }
  $registry_register{$species}{ lc($group) }{ lc($type) } = $adap;

  if ( !defined( $registry_register{$species}{'list'} ) ) {
    $registry_register{$species}{'list'} = [$type];
  } else {
    push( @{ $registry_register{$species}{'list'} }, $type );
  }

  if ( !defined( $registry_register{ lc($type) }{$species} ) ) {
    $registry_register{ lc($type) }{$species} = [$type];
  } else {
    push( @{ $registry_register{ lc($type) }{$species} }, $adap );
  }

} ## end sub add_adaptor


=head2 get_adaptor

  Arg [1]    : name of the species to add the adaptor to in the registry.
  Arg [2]    : name of the group to add the adaptor to in the registry.
  Arg [3]    : name of the type to add the adaptor to in the registry.
  Example    : $adap = Bio::EnsEMBL::Registry->get_adaptor("Human", "core", "Gene");
  Returntype : adaptor
  Exceptions : none
  Status     : Stable

=cut

sub get_adaptor{
  my ($class,$species,$group,$type)= @_;

  $species = $class->get_alias($species);

  my %dnadb_adaptors = qw(sequence  1 assemblymapper 1  karyotypeband 1 repeatfeature 1 coordsystem 1  assemblyexceptionfeature 1 );

  my $dnadb_group =  $registry_register{$species}{lc($group)}{_DNA};

  if( defined($dnadb_group) && defined($dnadb_adaptors{lc($type)}) ) {
      $species = $registry_register{$species}{lc($group)}{'_DNA2'};
      $group = $dnadb_group;
  }

  my $ret = $registry_register{$species}{ lc($group) }{ lc($type) };
  if ( !defined($ret) ) { return undef; }

  if(!ref($ret)){ # not instantiated yet
    my $dba = $registry_register{$species}{lc($group)}{'_DB'};
    my $module = $ret;
    eval "require $module";

    if($@) {
      warning("$module cannot be found.\nException $@\n");
      return undef;
    }
    if(!defined($registry_register{$species}{lc($group)}{'CHECKED'})){
      $registry_register{$species}{lc($group)}{'CHECKED'} = 1;
      $class->version_check($dba);
    }
    my $adap = "$module"->new($dba);
    Bio::EnsEMBL::Registry->add_adaptor($species, $group, $type, $adap, "reset");
    $ret = $adap;
  }

  return $ret;
}

=head2 get_all_adaptors

  Arg [SPECIES] : (optional) string 
                  species name to get adaptors for
  Arg [GROUP] : (optional) string 
                  group name to get adaptors for
  Arg [TYPE] : (optional) string 
                  type to get adaptors for
  Example    : @adaps = @{Bio::EnsEMBL::Registry->get_all_adaptors()};
  Returntype : ref to list of adaptors
  Exceptions : none
  Status     : Stable

=cut

sub get_all_adaptors{
  my ($class,@args)= @_;
  my ($species, $group, $type);
  my @ret=();
  my (%species_hash, %group_hash, %type_hash);


  if(@args == 1){ # Old species only one parameter
    warn("-SPECIES argument should now be used to get species adaptors");
    $species = $args[0];
  }
  else{
    # new style -SPECIES, -GROUP, -TYPE
    ($species, $group, $type) =
      rearrange([qw(SPECIES GROUP TYPE)], @args);
  }

  if(defined($species)){
    $species_hash{$species} = 1;
  }
  else{
    # get list of species
    foreach my $dba (@{$registry_register{'_DBA'}}){
      $species_hash{lc($dba->species())} = 1;
    }
  }
  if(defined($group)){
    $group_hash{$group} = 1;
  }
  else{
    foreach my $dba (@{$registry_register{'_DBA'}}){
      $group_hash{lc($dba->group())} = 1;
    }
  }
  if(defined($type)){
    $type_hash{$type} =1;
  }
  else{
    foreach my $dba (@{$registry_register{'_DBA'}}){ 
	foreach my $ty (@{$registry_register{lc($dba->species)}{'list'}}){
	  $type_hash{lc($ty)} = 1;
	}
      }
  }
  
  ### NOW NEED TO INSTANTIATE BY CALLING get_adaptor
  foreach my $sp (keys %species_hash){
    foreach my $gr (keys %group_hash){
      foreach my $ty (keys %type_hash){
	my $temp = $class->get_adaptor($sp,$gr,$ty);
	if(defined($temp)){
	  push @ret, $temp;
	}
      }
    }
  }
  return (\@ret);
}


=head2 add_alias

  Arg [1]    : name of the species to add alias for
  Arg [2]    : name of the alias
  Example    : Bio::EnsEMBL::Registry->add_alias("Homo Sapiens","Human");
  Description: add alternative name for the species.
  Returntype : none
  Exceptions : none
  Status     : Stable

=cut

sub add_alias{
  my ($class, $species,$key) = @_;

  $registry_register{'_ALIAS'}{lc($key)} = lc($species);
}

=head2 get_alias

  Arg [1]    : name of the possible alias to get species for
  Example    : Bio::EnsEMBL::Registry->get_alias("Human");
  Description: get proper species name.
  Returntype : species name
  Exceptions : none
  Status     : Stable

=cut

sub get_alias{
  my ($class, $key) = @_;

  if(!defined($registry_register{'_ALIAS'}{lc($key)})){
    return $key;
  }
  return $registry_register{'_ALIAS'}{lc($key)};
}

=head2 alias_exists

  Arg [1]    : name of the possible alias to get species for
  Example    : Bio::EnsEMBL::Registry->alias_exists("Human");
  Description: does the species name exist.
  Returntype : 1 if exists else 0
  Exceptions : none
  Status     : Stable

=cut

sub alias_exists{
  my ($class, $key) = @_;

  if(defined($registry_register{'_ALIAS'}{lc($key)})){
    return 1;
  }
  return 0;
}

=head2 set_disconnect_when_inactive

  Example    : Bio::EnsEMBL::Registry->set_disconnect_when_inactive();
  Description: Set the flag to make sure that the database connection is dropped if
               not being used on each database.
  Returntype : none
  Exceptions : none
  Status     : Stable

=cut

sub set_disconnect_when_inactive{
  foreach my $dba ( @{get_all_DBAdaptors()}){
    my $dbc = $dba->dbc;
    # Disconnect if connected
    $dbc->disconnect_if_idle() if $dbc->connected();
    $dbc->disconnect_when_inactive(1);
  }
}


=head2 disconnect_all

  Example    : Bio::EnsEMBL::Registry->disconnect_all();
  Description: disconnect from all the databases.
  Returntype : none
  Exceptions : none
  Status     : Stable

=cut

sub disconnect_all {
  foreach my $dba ( @{get_all_DBAdaptors()||[]} ){
    my $dbc = $dba->dbc;
    next unless $dbc;
    # Disconnect if connected
    $dbc->disconnect_if_idle() if $dbc->connected();
  }
}

=head2 change_access

  Will change the username and password for a set of databases.
  if host,user or database names are missing then these are not checked.
  So for example if you do not specify a database then ALL databases on
  the specified  host and port will be changed.

  Arg [1]    : name of the host to change access on
  Arg [2]    : port number to change access on
  Arg [3]    : name of the user to change access on
  Arg [4]    : name of the database to change access on
  Arg [5]    : name of the new user
  Arg [6]    : new password

  Example    : Bio::EnsEMBL::Registry->get_alias("Human");
  Description: change username and password on one or more databases
  Returntype : none
  Exceptions : none
  Status     : Stable

=cut

sub change_access{
my $self = shift;
    my ($host,$port,$user,$dbname,$new_user,$new_pass) = @_;
    foreach my $dba ( @{$registry_register{'_DBA'}}){
	my $dbc = $dba->dbc;
	if((!defined($host) or $host eq $dbc->host) and
	   (!defined($port) or $port eq $dbc->port) and
	   (!defined($user) or $user eq $dbc->username) and
	   (!defined($dbname) or $dbname eq $dbc->dbname)){
	    if($dbc->connected()){
		$dbc->db_handle->disconnect();
		$dbc->connected(undef);
	    }
	    # over write the username and password
	    $dbc->username($new_user);
	    $dbc->password($new_pass);
	}
    }
}



=head2 load_registry_from_url

  Arg [1] : string $url
  Arg [2] : (optional) integer
               If not 0, will print out all information.
  Arg [3] : (optional) integer
               This option will turn off caching for slice features, so, 
               every time a set of features is retrieved, they will come from
               the database instead of the cache. This option is only recommended
               for advanced users, specially if you need to store and retrieve
               features. It might reduce performance when querying the database if 
               not used properly. If in doubt, do not use it or ask in ensembl-dev           
  Example : load_registry_from_url(mysql://anonymous@ensembldb.ensembl.org:3306);
  Description: Will load the correct versions of the ensembl databases for the
               software release it can find on a database instance into the 
               registry. Also adds a set of standard aliases. The url format is:
               mysql://[[username][:password]@]hostname[:port].
               You can also request a specific version for the databases by adding
               a slash and the version number but your script may crash as the API
               version won't match the DB version.
  Exceptions : None.
  Status     : Stable
 
=cut

sub load_registry_from_url {
  my ($self, $url, $verbose, $no_cache) = @_;

  if ($url =~ /mysql\:\/\/([^\@]+\@)?([^\:\/]+)(\:\d+)?(\/\d+)?/) {
    my $user_pass = $1;
    my $host = $2;
    my $port = $3;
    my $version = $4;

    $user_pass =~ s/\@$//;
    my ($user, $pass) = $user_pass =~ m/([^\:]+)(\:.+)?/;
    $pass =~ s/^\:// if ($pass);
    $port =~ s/^\:// if ($port);
    $version =~ s/^\/// if ($version);

    $self->load_registry_from_db(
        -host=> $host,
        -user => $user,
        -pass => $pass,
        -port => $port,
        -db_version => $version,
        -verbose => $verbose,
	-no_cache => $no_cache);
  } else {
    throw("Only MySQL URLs are accepted at the moment");
  }
}


=head2 load_registry_from_db

  Arg [HOST] : The domain name of the database host to connect to.
               
  Arg [USER] : string
               The name of the database user to connect with
  Arg [PASS] : (optional) string
               The password to be used to connect to the database
  Arg [PORT] : int
               The port to use when connecting to the database
  Arg [VERBOSE]: (optional) Wether to print database messages 
  Arg [DB_VERSION]: (optional) By default, only databases corresponding
               to this API version are loaded. This allows the script to
               use databases from another version although it might not
               work properly. This option should only be used for
               production or testing purposes and if you really know what
               you are doing.
  Arg [WAIT_TIMEOUT]: (optional) integer
                 Time in seconds for the wait timeout to happen. Time after which
                 the connection is deleted if not used. By default this is 28800 (8 hours)
                 So set this to greater than this if your connection are getting deleted.
                 Only set this if you are having problems and know what you are doing.
   Arg [-NO_CACHE]: (optional) int 1
                 This option will turn off caching for slice features, so, 
                 every time a set of features is retrieved, they will come from
                 the database instead of the cache. This option is only recommended
                 for advanced users, specially if you need to store and retrieve
                 features. It might reduce performance when querying the database if 
                 not used properly. If in doubt, do not use it or ask in ensembl-dev       

  Example : load_registry_from_db( -host => 'ensembldb.ensembl.org',
				   -user => 'anonymous',
				   -verbose => "1" );

  Description: Will load the correct versions of the ensembl databases for the
               software release it can find on a database instance into the 
               registry. Also adds a set of standard aliases.

  Exceptions : None.
  Status     : Stable
 
=cut

sub load_registry_from_db {
  my ( $self, @args ) = @_;

  my ( $host, $port, $user, $pass, $verbose, $db_version,
       $wait_timeout, $no_cache ) =
    rearrange(
             [qw(HOST PORT USER PASS VERBOSE DB_VERSION WAIT_TIMEOUT NO_CACHE)],
             @args );

  my $go_version = 0;
  my $compara_version =0;
  my $ancestral_version =0;

  $user ||= "ensro";
  if ( !defined($port) ) {
    $port = 3306;
    if ( $host eq "ensembldb.ensembl.org" ) {
      if ( !defined($db_version) or $db_version >= 48 ) {
        $port = 5306;
      }
    }
  }

  $wait_timeout ||= 0;

  my $dbh =
    DBI->connect( "DBI:mysql:host=$host;port=$port", $user, $pass );

  my $res = $dbh->selectall_arrayref("show databases");
  my @dbnames = map { $_->[0] } @$res;

  my %temp;
  my $software_version = $self->software_version();

  if ( defined($db_version) ) {
    $software_version = $db_version;
  }

  if ($verbose) {
    printf( "Will only load v%d databases\n", $software_version );
  }

  for my $db (@dbnames) {
    if ( $db =~ /^(\w+)_(collection_core_(?:\d+_)?(\d+)_(\w+))/ )
    {    # NEEDS TO BE FIRST
      if ( $3 eq $software_version ) {
        $temp{$1} = $2;
      }
    } elsif ( $db =~ /^([a-z]+_[a-z]+_[a-z]+)_(\d+)_(\d+[a-z]*)/ ) {
      if ( $2 eq $software_version ) {
        $temp{$1} = $2 . "_" . $3;
      }
    } elsif ( $db =~ /^(.+)_(userdata)$/ ) {
      $temp{$1} = $2;
    } elsif ( $db =~ /^ensembl_compara_(\d+)/ ) {
      if ( $1 eq $software_version ) {
        $compara_version = $1;
      }
    } elsif ( $db =~ /^ensembl_ancestral_(\d+)/ ) {
      if ( $1 eq $software_version ) {
        $ancestral_version = $1;
      }
    } elsif ( $db =~ /^ensembl_go_(\d+)/ ) {
      if ( $1 eq $software_version ) {
        $go_version = $1;
      }
    }
  } ## end for my $db (@dbnames)

  @dbnames = ();

  foreach my $key ( keys %temp ) {
    push @dbnames, $key . "_" . $temp{$key};
  }

  # Register Core databases

  my @core_dbs = grep { /^[a-z]+_[a-z]+_core_\d+_/ } @dbnames;

  foreach my $coredb (@core_dbs) {
    next if ($coredb =~ /collection/);  # Skip multi-species databases

    my ( $species, $num ) =
      ( $coredb =~ /(^[a-z]+_[a-z]+)_core_(\d+)/ );

    my $dba =
      Bio::EnsEMBL::DBSQL::DBAdaptor->new(-group        => "core",
                                          -species      => $species,
                                          -host         => $host,
                                          -user         => $user,
                                          -pass         => $pass,
                                          -port         => $port