#!/usr/bin/perl
# Import an iso2709 file into Koha 3

use strict;
use warnings;
#use diagnostics;
BEGIN {
    # find Koha's Perl modules
    # test carefully before changing this
    use FindBin;
    eval { require "$FindBin::Bin/../kohalib.pl" };
}

# Koha modules used
use MARC::File::USMARC;
use MARC::File::XML;
use MARC::Record;
use MARC::Batch;
use MARC::Charset;

use C4::Context;
use C4::Biblio;
use C4::Koha;
use C4::Debug;
use C4::Charset;
use C4::Items;
use Unicode::Normalize;
use Time::HiRes qw(gettimeofday);
use Getopt::Long;
use IO::File;
use Pod::Usage;

binmode STDOUT, ':encoding(UTF-8)';
my ( $input_marc_file, $number, $offset) = ('',0,0);
my ($version, $delete, $test_parameter, $skip_marc8_conversion, $char_encoding, $verbose, $commit, $fk_off,$format,$biblios,$authorities,$keepids,$match, $isbn_check, $logfile);
my $cleanisbn = 1;
my ($sourcetag,$sourcesubfield,$idmapfl, $dedup_barcode);

$|=1;

GetOptions(
    'commit:f'    => \$commit,
    'file:s'    => \$input_marc_file,
    'n:f' => \$number,
    'o|offset:f' => \$offset,
    'h' => \$version,
    'd' => \$delete,
    't' => \$test_parameter,
    's' => \$skip_marc8_conversion,
    'c:s' => \$char_encoding,
    'v:s' => \$verbose,
    'fk' => \$fk_off,
    'm:s' => \$format,
    'l:s' => \$logfile,
    'k|keepids:s' => \$keepids,
    'b|biblios' => \$biblios,
    'a|authorities' => \$authorities,
    'match=s@'    => \$match,
    'i|isbn' => \$isbn_check,
    'x:s' => \$sourcetag,
    'y:s' => \$sourcesubfield,
    'idmap:s' => \$idmapfl,
    'cleanisbn!'     => \$cleanisbn,
    'dedupbarcode' => \$dedup_barcode,
);
$biblios=!$authorities||$biblios;

if ($version || ($input_marc_file eq '')) {
    pod2usage( -verbose => 2 );
    exit;
}

if (defined $idmapfl) {
  open(IDMAP,">$idmapfl") or die "cannot open $idmapfl \n";
}

if ((not defined $sourcesubfield) && (not defined $sourcetag)){
  $sourcetag="910";
  $sourcesubfield="a";
}

my $dbh = C4::Context->dbh;

# save the CataloguingLog property : we don't want to log a bulkmarcimport. It will slow the import & 
# will create problems in the action_logs table, that can't handle more than 1 entry per second per user.
my $CataloguingLog = C4::Context->preference('CataloguingLog');
$dbh->do("UPDATE systempreferences SET value=0 WHERE variable='CataloguingLog'");

if ($fk_off) {
	$dbh->do("SET FOREIGN_KEY_CHECKS = 0");
}


if ($delete) {
	if ($biblios){
    	print "deleting biblios\n";
    	$dbh->do("truncate biblio");
    	$dbh->do("truncate biblioitems");
    	$dbh->do("truncate items");
	}
	else {
    	print "deleting authorities\n";
    	$dbh->do("truncate auth_header");
	}
    $dbh->do("truncate zebraqueue");
}



if ($test_parameter) {
    print "TESTING MODE ONLY\n    DOING NOTHING\n===============\n";
}

my $marcFlavour = C4::Context->preference('marcflavour') || 'MARC21';

print "Characteristic MARC flavour: $marcFlavour\n" if $verbose;
my $starttime = gettimeofday;
my $batch;
my $fh = IO::File->new($input_marc_file); # don't let MARC::Batch open the file, as it applies the ':utf8' IO layer
if (defined $format && $format =~ /XML/i) {
    # ugly hack follows -- MARC::File::XML, when used by MARC::Batch,
    # appears to try to convert incoming XML records from MARC-8
    # to UTF-8.  Setting the BinaryEncoding key turns that off
    # TODO: see what happens to ISO-8859-1 XML files.
    # TODO: determine if MARC::Batch can be fixed to handle
    #       XML records properly -- it probably should be
    #       be using a proper push or pull XML parser to
    #       extract the records, not using regexes to look
    #       for <record>.*</record>.
    $MARC::File::XML::_load_args{BinaryEncoding} = 'utf-8';
    my $recordformat= ($marcFlavour eq "MARC21"?"USMARC":uc($marcFlavour));
#UNIMARC Authorities have a different way to manage encoding than UNIMARC biblios.
    $recordformat=$recordformat."AUTH" if ($authorities and $marcFlavour ne "MARC21");
    $MARC::File::XML::_load_args{RecordFormat} = $recordformat;
    $batch = MARC::Batch->new( 'XML', $fh );
} else {
    $batch = MARC::Batch->new( 'USMARC', $fh );
}
$batch->warnings_off();
$batch->strict_off();
my $i=0;
my $commitnum = $commit ? $commit : 50;


# Skip file offset
if ( $offset ) {
    print "Skipping file offset: $offset records\n";
    $batch->next() while ($offset--);
}

my ($tagid,$subfieldid);
if ($authorities){
	  $tagid='001';
}
else {
   ( $tagid, $subfieldid ) =
            GetMarcFromKohaField( "biblio.biblionumber", '' );
	$tagid||="001";
}

# the SQL query to search on isbn
my $sth_isbn = $dbh->prepare("SELECT biblionumber,biblioitemnumber FROM biblioitems WHERE isbn=?");

$dbh->{AutoCommit} = 0;
my $loghandle;
if ($logfile){
   $loghandle= IO::File->new($logfile,"w") ;
   print $loghandle "id;operation;status\n";
}
RECORD: while (  ) {
    my $record;
    # get records
    eval { $record = $batch->next() };
    if ( $@ ) {
        print "Bad MARC record $i: $@ skipped\n";
        # FIXME - because MARC::Batch->next() combines grabbing the next
        # blob and parsing it into one operation, a correctable condition
        # such as a MARC-8 record claiming that it's UTF-8 can't be recovered
        # from because we don't have access to the original blob.  Note
        # that the staging import can deal with this condition (via
        # C4::Charset::MarcToUTF8Record) because it doesn't use MARC::Batch.
        next;
    }
    # skip if we get an empty record (that is MARC valid, but will result in AddBiblio failure
    last unless ( $record );
    $i++;
    print ".";
    print "\r$i" unless $i % 100;
    
    # transcode the record to UTF8 if needed & applicable.
    if ($record->encoding() eq 'MARC-8' and not $skip_marc8_conversion) {
        # FIXME update condition
        my ($guessed_charset, $charset_errors);
         ($record, $guessed_charset, $charset_errors) = MarcToUTF8Record($record, $marcFlavour.(($authorities and $marcFlavour ne "MARC21")?'AUTH':''));
        if ($guessed_charset eq 'failed') {
            warn "ERROR: failed to perform character conversion for record $i\n";
            next RECORD;            
        }
    }
    my $isbn;
    # remove trailing - in isbn (only for biblios, of course)
    if ($biblios && $cleanisbn) {
        my $tag = $marcFlavour eq 'UNIMARC' ? '010' : '020';
        my $field = $record->field($tag);
        my $isbn = $field && $field->subfield('a');
        if ( $isbn ) {
            $isbn =~ s/-//g;
            $field->update('a' => $isbn);
        }
    }
    my $id;
    # search for duplicates (based on Local-number)
    if ($match){
       require C4::Search;
       my $query=build_query($match,$record);
       my $server=($authorities?'authorityserver':'biblioserver');
       my ($error, $results,$totalhits)=C4::Search::SimpleSearch( $query, 0, 3, [$server] );
       die "unable to search the database for duplicates : $error" if (defined $error);
       #warn "$query $server : $totalhits";
       if ( @{$results} == 1 ){
           my $marcrecord = MARC::File::USMARC::decode($results->[0]);
	   	   $id=GetRecordId($marcrecord,$tagid,$subfieldid);
       } 
       elsif  ( @{$results} > 1){
           $debug && warn "more than one match for $query";
       } 
       else {
           $debug && warn "nomatch for $query";
       }
    }
	my $originalid;
    if ($keepids){
	  $originalid=GetRecordId($record,$tagid,$subfieldid);
      if ($originalid){
		 my $storeidfield;
		 if (length($keepids)==3){
		 	$storeidfield=MARC::Field->new($keepids,$originalid);
		 }
		 else  {
			$storeidfield=MARC::Field->new(substr($keepids,0,3),"","",substr($keepids,3,1),$originalid);
		 }
         $record->insert_fields_ordered($storeidfield);
	     $record->delete_field($record->field($tagid));
      }
    }
    unless ($test_parameter) {
        if ($authorities){
            use C4::AuthoritiesMarc;
            my $authtypecode=GuessAuthTypeCode($record);
            my $authid= ($id?$id:GuessAuthId($record));
            if ($authid && GetAuthority($authid)){
            ## Authority has an id and is in database : Replace
                eval { ( $authid ) = ModAuthority($authid,$record, $authtypecode) };
                if ($@){
                    warn "Problem with authority $authid Cannot Modify";
					printlog({id=>$originalid||$id||$authid, op=>"edit",status=>"ERROR"}) if ($logfile);
                }
				else{
					printlog({id=>$originalid||$id||$authid, op=>"edit",status=>"ok"}) if ($logfile);
				}
            }  
            elsif (defined $authid) {
            ## An authid is defined but no authority in database : add
                eval { ( $authid ) = AddAuthority($record,$authid, $authtypecode) };
                if ($@){
                    warn "Problem with authority $authid Cannot Add ".$@;
					printlog({id=>$originalid||$id||$authid, op=>"insert",status=>"ERROR"}) if ($logfile);
                }
   				else{
					printlog({id=>$originalid||$id||$authid, op=>"insert",status=>"ok"}) if ($logfile);
				}
            }
	        else {
            ## True insert in database
                eval { ( $authid ) = AddAuthority($record,"", $authtypecode) };
                if ($@){
                    warn "Problem with authority $authid Cannot Add".$@;
					printlog({id=>$originalid||$id||$authid, op=>"insert",status=>"ERROR"}) if ($logfile);
                }
   				else{
					printlog({id=>$originalid||$id||$authid, op=>"insert",status=>"ok"}) if ($logfile);
				}
 	        }
        }
        else {
            my ( $biblionumber, $biblioitemnumber, $itemnumbers_ref, $errors_ref );
            $biblionumber = $id;
            # check for duplicate, based on ISBN (skip it if we already have found a duplicate with match parameter
            if (!$biblionumber && $isbn_check && $isbn) {
    #         warn "search ISBN : $isbn";
                $sth_isbn->execute($isbn);
                ($biblionumber,$biblioitemnumber) = $sth_isbn->fetchrow;
            }
        	if (defined $idmapfl) {
			 	if ($sourcetag < "010"){
					if ($record->field($sourcetag)){
					  my $source = $record->field($sourcetag)->data();
					  printf(IDMAP "%s|%s\n",$source,$biblionumber);
					}
			    } else {
					my $source=$record->subfield($sourcetag,$sourcesubfield);
					printf(IDMAP "%s|%s\n",$source,$biblionumber);
			  }
			}
					# create biblio, unless we already have it ( either match or isbn )
            if ($biblionumber) {
				eval{$biblioitemnumber=GetBiblioData($biblionumber)->{biblioitemnumber};}
			}
			else 
			{
                eval { ( $biblionumber, $biblioitemnumber ) = AddBiblio($record, '', { defer_marc_save => 1 }) };
            }
            if ( $@ ) {
                warn "ERROR: Adding biblio $biblionumber failed: $@\n";
				printlog({id=>$id||$originalid||$biblionumber, op=>"insert",status=>"ERROR"}) if ($logfile);
                next RECORD;
            } 
 			else{
				printlog({id=>$id||$originalid||$biblionumber, op=>"insert",status=>"ok"}) if ($logfile);
			}
            eval { ( $itemnumbers_ref, $errors_ref ) = AddItemBatchFromMarc( $record, $biblionumber, $biblioitemnumber, '' ); };
            my $error_adding = $@;
            # Work on a clone so that if there are real errors, we can maybe
            # fix them up later.
			my $clone_record = $record->clone();
            C4::Biblio::_strip_item_fields($clone_record, '');
            # This sets the marc fields if there was an error, and also calls
            # defer_marc_save.
            ModBiblioMarc( $clone_record, $biblionumber, '' );
            if ( $error_adding ) {
                warn "ERROR: Adding items to bib $biblionumber failed: $error_adding";
				printlog({id=>$id||$originalid||$biblionumber, op=>"insertitem",status=>"ERROR"}) if ($logfile);
                # if we failed because of an exception, assume that 
                # the MARC columns in biblioitems were not set.
                next RECORD;
            }
 			else{
				printlog({id=>$id||$originalid||$biblionumber, op=>"insert",status=>"ok"}) if ($logfile);
			}
            if ($dedup_barcode && grep { exists $_->{error_code} && $_->{error_code} eq 'duplicate_barcode' } @$errors_ref) {
                # Find the record called 'barcode'
                my ($tag, $sub) = C4::Biblio::GetMarcFromKohaField('items.barcode', '');
                # Now remove any items that didn't have a duplicate_barcode error,
                # erase the barcodes on items that did, and re-add those items.
                my %dupes;
                foreach my $i (0 .. $#{$errors_ref}) {
                    my $ref = $errors_ref->[$i];
                    if ($ref && ($ref->{error_code} eq 'duplicate_barcode')) {
                        $dupes{$ref->{item_sequence}} = 1;
                        # Delete the error message because we're going to
                        # retry this one.
                        delete $errors_ref->[$i];
                    }
                }
                my $seq = 0;
                foreach my $field ($record->field($tag)) {
                    $seq++;
                    if ($dupes{$seq}) {
                        # Here we remove the barcode
                        $field->delete_subfield(code => $sub);
                    } else {
                        # otherwise we delete the field because we don't want
                        # two of them
                        $record->delete_fields($field);
                    }
                }
                # Now re-add the record as before, adding errors to the prev list
                my $more_errors;
                eval { ( $itemnumbers_ref, $more_errors ) = AddItemBatchFromMarc( $record, $biblionumber, $biblioitemnumber, '' ); };
                if ( $@ ) {
                    warn "ERROR: Adding items to bib $biblionumber failed: $@\n";
                    printlog({id=>$id||$originalid||$biblionumber, op=>"insertitem",status=>"ERROR"}) if ($logfile);
                    # if we failed because of an exception, assume that
                    # the MARC columns in biblioitems were not set.
                    ModBiblioMarc( $record, $biblionumber, '' );
                    next RECORD;
                } else {
                    printlog({id=>$id||$originalid||$biblionumber, op=>"insert",status=>"ok"}) if ($logfile);
                }
                push @$errors_ref, @{ $more_errors };
            }
            if ($#{ $errors_ref } > -1) {
                report_item_errors($biblionumber, $errors_ref);
            }
        }
        $dbh->commit() if (0 == $i % $commitnum);
    }
    last if $i == $number;
}
$dbh->commit();



if ($fk_off) {
	$dbh->do("SET FOREIGN_KEY_CHECKS = 1");
}

# restore CataloguingLog
$dbh->do("UPDATE systempreferences SET value=$CataloguingLog WHERE variable='CataloguingLog'");

my $timeneeded = gettimeofday - $starttime;
print "\n$i MARC records done in $timeneeded seconds\n";
if ($logfile){
  print $loghandle "file : $input_marc_file\n";
  print $loghandle "$i MARC records done in $timeneeded seconds\n";
  $loghandle->close;
}
exit 0;

sub GetRecordId{
	my $marcrecord=shift;
	my $tag=shift;
	my $subfield=shift;
	my $id;
	if ($tag lt "010"){
		return $marcrecord->field($tag)->data() if $marcrecord->field($tag);
	} 
	elsif ($subfield){
		if ($marcrecord->field($tag)){
			return $marcrecord->subfield($tag,$subfield);
		}
	}
	return $id;
}
sub build_query {
	my $match = shift;
	my $record=shift;
        my @searchstrings;
	foreach my $matchingpoint (@$match){
	  my $string = build_simplequery($matchingpoint,$record);
	  push @searchstrings,$string if (length($string)>0);
        }
	return join(" and ",@searchstrings);
}
sub build_simplequery {
	my $element=shift;
	my $record=shift;
        my ($index,$recorddata)=split /,/,$element;
        my ($tag,$subfields) =($1,$2) if ($recorddata=~/(\d{3})(.*)/);
        my @searchstrings;
        foreach my $field ($record->field($tag)){
		  if (length($field->as_string("$subfields"))>0){
          	push @searchstrings,"$index,wrdl=\"".$field->as_string("$subfields")."\"";
		  }
        }
	return join(" and ",@searchstrings);
}
sub report_item_errors {
    my $biblionumber = shift;
    my $errors_ref = shift;

    foreach my $error (@{ $errors_ref }) {
        next if !$error;
        my $msg = "Item not added (bib $biblionumber, item tag #$error->{'item_sequence'}, barcode $error->{'item_barcode'}): ";
        my $error_code = $error->{'error_code'};
        $error_code =~ s/_/ /g;
        $msg .= "$error_code $error->{'error_information'}";
        print $msg, "\n";
    }
}
sub printlog{
	my $logelements=shift;
	print $loghandle join (";",@$logelements{qw<id op status>}),"\n";
}


=head1 NAME

bulkmarcimport.pl - Import bibliographic/authority records into Koha

=head1 USAGE

 $ export KOHA_CONF=/etc/koha.conf
 $ perl misc/migration_tools/bulkmarcimport.pl -d -commit 1000 \\
    -file /home/jmf/koha.mrc -n 3000

=head1 WARNING

Don't use this script before you've entered and checked your MARC parameters
tables twice (or more!). Otherwise, the import won't work correctly and you
will get invalid data.

=head1 DESCRIPTION

=over

=item  B<-h>

This version/help screen

=item B<-b, -biblios>

Type of import: bibliographic records

=item B<-a, -authorities>

Type of import: authority records

=item B<-file>=I<FILE>

The I<FILE> to import

=item  B<-v>

Verbose mode. 1 means "some infos", 2 means "MARC dumping"

=item B<-fk>

Turn off foreign key checks during import.

=item B<-n>=I<NUMBER>

The I<NUMBER> of records to import. If missing, all the file is imported

=item B<-o, -offset>=I<NUMBER>

File offset before importing, ie I<NUMBER> of records to skip.

=item B<-commit>=I<NUMBER>

The I<NUMBER> of records to wait before performing a 'commit' operation

=item B<-l>

File logs actions done for each record and their status into file

=item B<-t>

Test mode: parses the file, saying what he would do, but doing nothing.

=item B<-s>

Skip automatic conversion of MARC-8 to UTF-8.  This option is provided for
debugging.

=item B<-c>=I<CHARACTERISTIC>

The I<CHARACTERISTIC> MARC flavour. At the moment, only I<MARC21> and
I<UNIMARC> are supported. MARC21 by default.

=item B<-d>

Delete EVERYTHING related to biblio in koha-DB before import. Tables: biblio,
biblioitems, items

=item B<-m>=I<FORMAT>

Input file I<FORMAT>: I<MARCXML> or I<ISO2709> (defaults to ISO2709)

=item B<-k, -keepids>=<FIELD>

Field store ids in I<FIELD> (usefull for authorities, where 001 contains the
authid for Koha, that can contain a very valuable info for authorities coming
from LOC or BNF. useless for biblios probably)

=item B<-match>=<FIELD>

I<FIELD> matchindex,fieldtomatch matchpoint to use to deduplicate fieldtomatch
can be either 001 to 999 or field and list of subfields as such 100abcde

=item B<-i,-isbn>

If set, a search will be done on isbn, and, if the same isbn is found, the
biblio is not added. It's another method to deduplicate.  B<-match> & B<-isbn>
can be both set.

=item B<-cleanisbn>

Clean ISBN fields from entering biblio records, ie removes hyphens. By default,
ISBN are cleaned. --nocleanisbn will keep ISBN unchanged.

=item B<-x>=I<TAG>

Source bib I<TAG> for reporting the source bib number

=item B<-y>=I<SUBFIELD>

Source I<SUBFIELD> for reporting the source bib number

=item B<-idmap>=I<FILE>

I<FILE> for the koha bib and source id

=item B<-keepids>

Store ids in 009 (usefull for authorities, where 001 contains the authid for
Koha, that can contain a very valuable info for authorities coming from LOC or
BNF. useless for biblios probably)

=item B<-dedupbarcode>

If set, whenever a duplicate barcode is detected, it is removed and the attempt
to add the record is retried, thereby giving the record a blank barcode. This
is useful when something has set barcodes to be a biblio ID, or similar
(usually other software.)

=back

=cut