webpac2/trunk/run.pl

#!/usr/bin/perl -w

use strict;

use Cwd qw/abs_path/;
use File::Temp qw/tempdir/;
use lib './lib';

use WebPAC::Common 0.02;
use WebPAC::Parser 0.04;
use WebPAC::Lookup 0.03;
use WebPAC::Input 0.11;
use WebPAC::Store 0.03;
use WebPAC::Normalize 0.11;
use WebPAC::Output::TT;
use WebPAC::Validate 0.06;
use WebPAC::Output::MARC;
use WebPAC::Config;
use Getopt::Long;
use File::Path;
use Time::HiRes qw/time/;
use File::Slurp;
use Data::Dump qw/dump/;
use Storable qw/dclone/;

use Proc::Queue size => 1;
use POSIX ":sys_wait_h"; # imports WNOHANG

=head1 NAME

run.pl - start WebPAC indexing

B<this command will probably go away. Don't get used to it!>

Options:

=over 4

=item --offset 42

start loading (all) databases at offset 42

=item --limit 100

limit loading to 100 records

=item --clean

remove database and Hyper Estraier index before indexing

=item --only=database_name/input_filter

reindex just single database (legacy name is --one)

C</input_filter> is optional part which can be C<name>
or C<type> from input

=item --config conf/config.yml

path to YAML configuration file

=item --stats

disable indexing, modify_* in configuration and dump statistics about field
and subfield usage for each input

=item --validate path/to/validation_file

turn on extra validation of imput records, see L<WebPAC::Validation>

=item --marc-normalize conf/normalize/mapping.pl

This option specifies normalisation file for MARC creation

=item --marc-output out/marc/test.marc

Optional path to output file

=item --marc-lint

By default turned on if C<--marc-normalize> is used. You can disable lint
messages with C<--no-marc-lint>.

=item --marc-dump

Force dump or input and marc record for debugging.

=item --parallel 4

Run databases in parallel (aproximatly same as number of processors in
machine if you want to use full load)

=item --only-links

Create just links

=item --merge

Create merged index of databases which have links

=back

=cut

my $offset;
my $limit;

my $clean = 0;
my $config_path;
my $debug = 0;
my $only_filter;
my $stats = 0;
my $validate_path;
my ($marc_normalize, $marc_output);
my $marc_lint = 1;
my $marc_dump = 0;
my $parallel = 0;
my $only_links = 0;
my $merge = 0;

my $log = _new WebPAC::Common()->_get_logger();

GetOptions(
        "limit=i" => \$limit,
        "offset=i" => \$offset,
        "clean" => \$clean,
        "one=s" => \$only_filter,
        "only=s" => \$only_filter,
        "config" => \$config_path,
        "debug+" => \$debug,
        "stats" => \$stats,
        "validate=s" => \$validate_path,
        "marc-normalize=s" => \$marc_normalize,
        "marc-output=s" => \$marc_output,
        "marc-lint!" => \$marc_lint,
        "marc-dump!" => \$marc_dump,
        "parallel=i" => \$parallel,
        "only-links!" => \$only_links,
        "merge" => \$merge,
);

my $config = new WebPAC::Config( path => $config_path );

#print "config = ",dump($config) if ($debug);

die "no databases in config file!\n" unless ($config->databases);

$log->info( "-" x 79 );


my $estcmd_fh;
my $estcmd_path = './estcmd-merge.sh';
if ($merge) {
        open($estcmd_fh, '>', $estcmd_path) || $log->logdie("can't open $estcmd_path: $!");
        print $estcmd_fh 'cd /data/estraier/_node/ || exit 1',$/;
        print $estcmd_fh 'sudo /etc/init.d/hyperestraier stop',$/;
        $log->info("created merge batch file $estcmd_path");
}


my $validate;
$validate = new WebPAC::Validate(
        path => $validate_path,
) if ($validate_path);


my $use_indexer = $config->use_indexer;
if ($stats) {
        $log->debug("option --stats disables update of indexing engine...");
        $use_indexer = undef;
} else {
        $log->info("using $use_indexer indexing engine...");
}

# disable indexing when creating marc
$use_indexer = undef if ($marc_normalize);

# parse normalize files and create source files for lookup and normalization

my $parser = new WebPAC::Parser( config => $config );

my $total_rows = 0;
my $start_t = time();

my @links;

if ($parallel) {
        $log->info("Using $parallel processes for speedup");
        Proc::Queue::size($parallel);
}

while (my ($database, $db_config) = each %{ $config->databases }) {

        my ($only_database,$only_input) = split(m#/#, $only_filter) if ($only_filter);
        next if ($only_database && $database !~ m/$only_database/i);

        if ($parallel) {
                my $f=fork;
                if(defined ($f) and $f==0) {
                        $log->info("Created processes $$ for speedup");
                } else {
                        next;
                }
        }

        my $indexer;
        if ($use_indexer) {

                my $cfg_name = $use_indexer;
                $cfg_name =~ s/\-.*$//;

                my $indexer_config = $config->get( $cfg_name ) || $log->logdie("can't find '$cfg_name' part in confguration");
                $indexer_config->{database} = $database;
                $indexer_config->{clean} = $clean;
                $indexer_config->{label} = $db_config->{name};

                # force clean if database has links
                $indexer_config->{clean} = 1 if ($db_config->{links});

                if ($use_indexer eq 'hyperestraier') {

                        # open Hyper Estraier database
                        use WebPAC::Output::Estraier '0.10';
                        $indexer = new WebPAC::Output::Estraier( %{ $indexer_config } );
                
                } elsif ($use_indexer eq 'hyperestraier-native') {

                        # open Hyper Estraier database
                        use WebPAC::Output::EstraierNative;
                        $indexer = new WebPAC::Output::EstraierNative( %{ $indexer_config } );

                } elsif ($use_indexer eq 'kinosearch') {

                        # open KinoSearch
                        use WebPAC::Output::KinoSearch;
                        $indexer_config->{clean} = 1 unless (-e $indexer_config->{index_path});
                        $indexer = new WebPAC::Output::KinoSearch( %{ $indexer_config } );

                } else {
                        $log->logdie("unknown use_indexer: $use_indexer");
                }

                $log->logide("can't continue without valid indexer") unless ($indexer);
        }


        #
        # store Hyper Estraier links to other databases
        #
        if (ref($db_config->{links}) eq 'ARRAY' && $use_indexer) {
                foreach my $link (@{ $db_config->{links} }) {
                        if ($use_indexer eq 'hyperestraier') {
                                if ($merge) {
                                        print $estcmd_fh 'sudo -u www-data estcmd merge ' . $database . ' ' . $link->{to},$/;
                                } else {
                                        $log->info("saving link $database -> $link->{to} [$link->{credit}]");
                                        push @links, sub {
                                                $log->info("adding link $database -> $link->{to} [$link->{credit}]");
                                                $indexer->add_link(
                                                        from => $database,
                                                        to => $link->{to},
                                                        credit => $link->{credit},
                                                );
                                        };
                                }
                        } else {
                                $log->warn("NOT IMPLEMENTED WITH $use_indexer: adding link $database -> $link->{to} [$link->{credit}]");
                        }
                }
        }
        next if ($only_links);


        #
        # now WebPAC::Store
        #
        my $abs_path = abs_path($0);
        $abs_path =~ s#/[^/]*$#/#;

        my $db_path = $config->get('webpac')->{db_path} . '/' . $database;

        if ($clean) {
                $log->info("creating new database '$database' in $db_path");
                rmtree( $db_path ) || $log->warn("can't remove $db_path: $!");
        } else {
                $log->info("working on database '$database' in $db_path");
        }

        my $db = new WebPAC::Store(
                path => $db_path,
                database => $database,
                debug => $debug,
        );


        #
        # now, iterate through input formats
        #

        my @inputs;
        if (ref($db_config->{input}) eq 'ARRAY') {
                @inputs = @{ $db_config->{input} };
        } elsif ($db_config->{input}) {
                push @inputs, $db_config->{input};
        } else {
                $log->info("database $database doesn't have inputs defined");
        }

        foreach my $input (@inputs) {

                next if ($only_input && ($input->{name} !~ m#$only_input#i && $input->{type} !~ m#$only_input#i));

                my $type = lc($input->{type});

                die "I know only how to handle input types ", join(",", $config->webpac('inputs') ), " not '$type'!\n" unless (grep(/$type/, $config->webpac('inputs')));

                my $input_module = $config->webpac('inputs')->{$type};

                $log->info("working on input '$input->{name}' in $input->{path} [type: $input->{type}] using $input_module",
                        $input->{lookup} ? "lookup '$input->{lookup}'" : ""
                );

                if ($stats) {
                        # disable modification of records if --stats is in use
                        delete($input->{modify_records});
                        delete($input->{modify_file});
                }

                warn "parser->depends = ", dump( $parser->{depends} );
                warn "depends on: ", dump( $parser->depends($database, $input->{name}) );
                warn "lookup_create_rules = ", dump( $parser->lookup_create_rules($database, $input->{name}) );
                warn "parser->_lookup_create = ", dump( $parser->{_lookup_create} );

                my $lookup;

                my $input_db = new WebPAC::Input(
                        module => $input_module,
                        encoding => $config->webpac('webpac_encoding'),
                        limit => $limit || $input->{limit},
                        offset => $offset,
                        lookup_coderef => sub {
                                my $rec = shift || return;
                                $lookup->add( $rec );
                        },
                        recode => $input->{recode},
                        stats => $stats,
                        modify_records => $input->{modify_records},
                        modify_file => $input->{modify_file},
                );
                $log->logdie("can't create input using $input_module") unless ($input);

                my $maxmfn = $input_db->open(
                        path => $input->{path},
                        code_page => $input->{encoding},        # database encoding
                        %{ $input },
                );

                my $report_fh;
                if ($stats || $validate) {
                        my $path = "out/report/" . $database . '-' . $input->{name} . '.txt';
                        open($report_fh, '>', $path) || $log->logdie("can't open $path: $!");

                        print $report_fh "Report for database '$database' input '$input->{name}' records ",
                                $offset || 1, "-", $limit || $input->{limit} || $maxmfn, "\n\n";
                        $log->info("Generating report file $path");
                }

                my @norm_array = ref($input->{normalize}) eq 'ARRAY' ?
                        @{ $input->{normalize} } : ( $input->{normalize} );

                if ($marc_normalize) {
                        @norm_array = ( {
                                path => $marc_normalize,
                                output => $marc_output || 'out/marc/' . $database . '-' . $input->{name} . '.marc',
                        } );
                }

                foreach my $normalize (@norm_array) {

                        my $normalize_path = $normalize->{path} || $log->logdie("can't find normalize path in config");

                        $log->logdie("Found '$normalize_path' as normalization file which isn't supported any more!") unless ( $normalize_path =~ m!\.pl$!i );

                        my $rules = read_file( $normalize_path ) or die "can't open $normalize_path: $!";

                        $log->info("Using $normalize_path for normalization...");

                        my $marc = new WebPAC::Output::MARC(
                                path => $normalize->{output},
                                lint => $marc_lint,
                                dump => $marc_dump,
                        ) if ($normalize->{output});

                        # reset position in database
                        $input_db->seek(1);

                        # generate name of config key for indexer (strip everything after -)
                        my $indexer_config = $use_indexer;
                        $indexer_config =~ s/^(\w+)-?.*$/$1/g if ($indexer_config);

                        foreach my $pos ( 0 ... $input_db->size ) {

                                my $row = $input_db->fetch || next;

                                my $mfn = $row->{'000'}->[0];

                                if (! $mfn || $mfn !~ m#^\d+$#) {
                                        $log->warn("record $pos doesn't have valid MFN but '$mfn', using $pos");
                                        $mfn = $pos;
                                        push @{ $row->{'000'} }, $pos;
                                }


                                if ($validate) {
                                        if ( my $errors = $validate->validate_errors( $row, $input_db->dump ) ) {
                                                $log->error( "MFN $mfn validation error:\n",
                                                        $validate->report_error( $errors )
                                                );
                                        }
                                }

                                my $ds_config = dclone($db_config);

                                # default values -> database key
                                $ds_config->{_} = $database;

                                # current mfn
                                $ds_config->{_mfn} = $mfn;

                                # attach current input
                                $ds_config->{input} = $input;

                                my $ds = WebPAC::Normalize::data_structure(
                                        row => $row,
                                        rules => $rules,
                                        lookup => $lookup ? $lookup->lookup_hash : undef,
                                        config => $ds_config,
                                        marc_encoding => 'utf-8',
                                );

                                $db->save_ds(
                                        id => $mfn,
                                        ds => $ds,
                                        prefix => $input->{name},
                                ) if ($ds && !$stats);

                                $indexer->add(
                                        id => $input->{name} . "/" . $mfn,
                                        ds => $ds,
                                        type => $config->get($indexer_config)->{type},
                                ) if ($indexer && $ds);

                                if ($marc) {
                                        my $i = 0;

                                        while (my $fields = WebPAC::Normalize::_get_marc_fields( fetch_next => 1 ) ) {
                                                $marc->add(
                                                        id => $mfn . ( $i ? "/$i" : '' ),
                                                        fields => $fields,
                                                        leader => WebPAC::Normalize::marc_leader(),
                                                        row => $row,
                                                );
                                                $i++;
                                        }

                                        $log->info("Created $i instances of MFN $mfn\n") if ($i > 1);
                                }

                                $total_rows++;
                        }

                        if ($validate) {
                                my $errors = $validate->report;
                                if ($errors) {
                                        $log->info("validation errors:\n$errors\n" );
                                        print $report_fh "$errors\n" if ($report_fh);
                                }
                        }

                        if ($stats) {
                                my $s = $input_db->stats;
                                $log->info("statistics of fields usage:\n$s");
                                print $report_fh "Statistics of fields usage:\n$s" if ($report_fh);
                        }

                        # close MARC file
                        $marc->finish if ($marc);

                        # close report
                        close($report_fh) if ($report_fh)
                }

        }

        eval { $indexer->finish } if ($indexer && $indexer->can('finish'));

        my $dt = time() - $start_t;
        $log->info("$total_rows records ", $indexer ? "indexed " : "",
                sprintf("in %.2f sec [%.2f rec/sec]",
                        $dt, ($total_rows / $dt)
                )
        );


        # end forked process
        if ($parallel) {
                $log->info("parallel process $$ finished");
                exit(0);
        }

}

if ($parallel) {
        # wait all children to finish
        sleep(1) while wait != -1;
        $log->info("all parallel processes finished");
}

#
# handle links or merge after indexing
#

if ($merge) {
        print $estcmd_fh 'sudo /etc/init.d/hyperestraier start',$/;
        close($estcmd_fh);
        chmod 0700, $estcmd_path || $log->warn("can't chmod 0700 $estcmd_path: $!");
        system $estcmd_path;
} else {
        foreach my $link (@links) {
                $log->logdie("coderef in link ", Dumper($link), " is ", ref($link), " and not CODE") unless (ref($link) eq 'CODE');
                $link->();
        }
}
1	dpavlin	74	#!/usr/bin/perl -w
2
3			use strict;
4
5			use Cwd qw/abs_path/;
6			use File::Temp qw/tempdir/;
7			use lib './lib';
8
9	dpavlin	255	use WebPAC::Common 0.02;
10	dpavlin	698	use WebPAC::Parser 0.04;
11	dpavlin	588	use WebPAC::Lookup 0.03;
12	dpavlin	619	use WebPAC::Input 0.11;
13	dpavlin	209	use WebPAC::Store 0.03;
14	dpavlin	581	use WebPAC::Normalize 0.11;
15	dpavlin	74	use WebPAC::Output::TT;
16	dpavlin	652	use WebPAC::Validate 0.06;
17	dpavlin	578	use WebPAC::Output::MARC;
18	dpavlin	684	use WebPAC::Config;
19	dpavlin	301	use Getopt::Long;
20			use File::Path;
21	dpavlin	389	use Time::HiRes qw/time/;
22	dpavlin	492	use File::Slurp;
23	dpavlin	556	use Data::Dump qw/dump/;
24	dpavlin	595	use Storable qw/dclone/;
25	dpavlin	74
26	dpavlin	606	use Proc::Queue size => 1;
27			use POSIX ":sys_wait_h"; # imports WNOHANG
28
29	dpavlin	301	=head1 NAME
30	dpavlin	76
31	dpavlin	301	run.pl - start WebPAC indexing
32	dpavlin	141
33	dpavlin	301	B<this command will probably go away. Don't get used to it!>
34	dpavlin	141
35	dpavlin	301	Options:
36
37			=over 4
38
39			=item --offset 42
40
41			start loading (all) databases at offset 42
42
43			=item --limit 100
44
45			limit loading to 100 records
46
47			=item --clean
48
49			remove database and Hyper Estraier index before indexing
50
51	dpavlin	510	=item --only=database_name/input_filter
52	dpavlin	335
53	dpavlin	423	reindex just single database (legacy name is --one)
54	dpavlin	335
55	dpavlin	510	C</input_filter> is optional part which can be C<name>
56			or C<type> from input
57
58	dpavlin	301	=item --config conf/config.yml
59
60			path to YAML configuration file
61
62	dpavlin	507	=item --stats
63
64	dpavlin	638	disable indexing, modify_* in configuration and dump statistics about field
65			and subfield usage for each input
66	dpavlin	507
67	dpavlin	516	=item --validate path/to/validation_file
68
69			turn on extra validation of imput records, see L<WebPAC::Validation>
70
71	dpavlin	552	=item --marc-normalize conf/normalize/mapping.pl
72
73			This option specifies normalisation file for MARC creation
74
75			=item --marc-output out/marc/test.marc
76
77			Optional path to output file
78
79	dpavlin	556	=item --marc-lint
80
81			By default turned on if C<--marc-normalize> is used. You can disable lint
82	dpavlin	558	messages with C<--no-marc-lint>.
83	dpavlin	556
84	dpavlin	559	=item --marc-dump
85
86			Force dump or input and marc record for debugging.
87
88	dpavlin	606	=item --parallel 4
89
90			Run databases in parallel (aproximatly same as number of processors in
91			machine if you want to use full load)
92
93	dpavlin	607	=item --only-links
94
95			Create just links
96
97	dpavlin	608	=item --merge
98
99			Create merged index of databases which have links
100
101	dpavlin	301	=back
102
103			=cut
104
105			my $offset;
106			my $limit;
107
108			my $clean = 0;
109	dpavlin	684	my $config_path;
110	dpavlin	301	my $debug = 0;
111	dpavlin	510	my $only_filter;
112	dpavlin	507	my $stats = 0;
113	dpavlin	516	my $validate_path;
114	dpavlin	552	my ($marc_normalize, $marc_output);
115	dpavlin	556	my $marc_lint = 1;
116	dpavlin	559	my $marc_dump = 0;
117	dpavlin	606	my $parallel = 0;
118	dpavlin	607	my $only_links = 0;
119	dpavlin	608	my $merge = 0;
120	dpavlin	606
121	dpavlin	611	my $log = _new WebPAC::Common()->_get_logger();
122
123	dpavlin	301	GetOptions(
124			"limit=i" => \$limit,
125			"offset=i" => \$offset,
126			"clean" => \$clean,
127	dpavlin	510	"one=s" => \$only_filter,
128			"only=s" => \$only_filter,
129	dpavlin	684	"config" => \$config_path,
130	dpavlin	560	"debug+" => \$debug,
131	dpavlin	507	"stats" => \$stats,
132	dpavlin	516	"validate=s" => \$validate_path,
133	dpavlin	552	"marc-normalize=s" => \$marc_normalize,
134			"marc-output=s" => \$marc_output,
135	dpavlin	556	"marc-lint!" => \$marc_lint,
136	dpavlin	559	"marc-dump!" => \$marc_dump,
137	dpavlin	606	"parallel=i" => \$parallel,
138	dpavlin	607	"only-links!" => \$only_links,
139	dpavlin	608	"merge" => \$merge,
140	dpavlin	301	);
141
142	dpavlin	684	my $config = new WebPAC::Config( path => $config_path );
143	dpavlin	301
144	dpavlin	611	#print "config = ",dump($config) if ($debug);
145	dpavlin	301
146	dpavlin	684	die "no databases in config file!\n" unless ($config->databases);
147	dpavlin	431
148	dpavlin	509	$log->info( "-" x 79 );
149	dpavlin	431
150	dpavlin	608
151			my $estcmd_fh;
152			my $estcmd_path = './estcmd-merge.sh';
153			if ($merge) {
154			open($estcmd_fh, '>', $estcmd_path) \|\| $log->logdie("can't open $estcmd_path: $!");
155			print $estcmd_fh 'cd /data/estraier/_node/ \|\| exit 1',$/;
156			print $estcmd_fh 'sudo /etc/init.d/hyperestraier stop',$/;
157			$log->info("created merge batch file $estcmd_path");
158			}
159
160
161	dpavlin	516	my $validate;
162			$validate = new WebPAC::Validate(
163			path => $validate_path,
164			) if ($validate_path);
165
166	dpavlin	608
167	dpavlin	684	my $use_indexer = $config->use_indexer;
168	dpavlin	509	if ($stats) {
169			$log->debug("option --stats disables update of indexing engine...");
170			$use_indexer = undef;
171			} else {
172			$log->info("using $use_indexer indexing engine...");
173			}
174	dpavlin	141
175	dpavlin	552	# disable indexing when creating marc
176			$use_indexer = undef if ($marc_normalize);
177
178	dpavlin	698	# parse normalize files and create source files for lookup and normalization
179
180			my $parser = new WebPAC::Parser( config => $config );
181
182	dpavlin	213	my $total_rows = 0;
183	dpavlin	389	my $start_t = time();
184	dpavlin	213
185	dpavlin	529	my @links;
186
187	dpavlin	606	if ($parallel) {
188			$log->info("Using $parallel processes for speedup");
189			Proc::Queue::size($parallel);
190			}
191
192	dpavlin	684	while (my ($database, $db_config) = each %{ $config->databases }) {
193	dpavlin	141
194	dpavlin	512	my ($only_database,$only_input) = split(m#/#, $only_filter) if ($only_filter);
195	dpavlin	510	next if ($only_database && $database !~ m/$only_database/i);
196	dpavlin	335
197	dpavlin	606	if ($parallel) {
198			my $f=fork;
199			if(defined ($f) and $f==0) {
200			$log->info("Created processes $$ for speedup");
201			} else {
202			next;
203			}
204			}
205
206	dpavlin	607	my $indexer;
207	dpavlin	509	if ($use_indexer) {
208	dpavlin	627
209			my $cfg_name = $use_indexer;
210			$cfg_name =~ s/\-.*$//;
211
212	dpavlin	684	my $indexer_config = $config->get( $cfg_name ) \|\| $log->logdie("can't find '$cfg_name' part in confguration");
213	dpavlin	509	$indexer_config->{database} = $database;
214			$indexer_config->{clean} = $clean;
215			$indexer_config->{label} = $db_config->{name};
216	dpavlin	431
217	dpavlin	608	# force clean if database has links
218			$indexer_config->{clean} = 1 if ($db_config->{links});
219
220	dpavlin	509	if ($use_indexer eq 'hyperestraier') {
221	dpavlin	255
222	dpavlin	509	# open Hyper Estraier database
223			use WebPAC::Output::Estraier '0.10';
224			$indexer = new WebPAC::Output::Estraier( %{ $indexer_config } );
225
226	dpavlin	627	} elsif ($use_indexer eq 'hyperestraier-native') {
227
228			# open Hyper Estraier database
229			use WebPAC::Output::EstraierNative;
230			$indexer = new WebPAC::Output::EstraierNative( %{ $indexer_config } );
231
232	dpavlin	509	} elsif ($use_indexer eq 'kinosearch') {
233	dpavlin	430
234	dpavlin	509	# open KinoSearch
235			use WebPAC::Output::KinoSearch;
236			$indexer_config->{clean} = 1 unless (-e $indexer_config->{index_path});
237			$indexer = new WebPAC::Output::KinoSearch( %{ $indexer_config } );
238	dpavlin	431
239	dpavlin	509	} else {
240			$log->logdie("unknown use_indexer: $use_indexer");
241			}
242
243			$log->logide("can't continue without valid indexer") unless ($indexer);
244	dpavlin	430	}
245
246
247	dpavlin	255	#
248	dpavlin	607	# store Hyper Estraier links to other databases
249			#
250			if (ref($db_config->{links}) eq 'ARRAY' && $use_indexer) {
251			foreach my $link (@{ $db_config->{links} }) {
252			if ($use_indexer eq 'hyperestraier') {
253	dpavlin	608	if ($merge) {
254			print $estcmd_fh 'sudo -u www-data estcmd merge ' . $database . ' ' . $link->{to},$/;
255			} else {
256			$log->info("saving link $database -> $link->{to} [$link->{credit}]");
257			push @links, sub {
258			$log->info("adding link $database -> $link->{to} [$link->{credit}]");
259			$indexer->add_link(
260			from => $database,
261			to => $link->{to},
262			credit => $link->{credit},
263			);
264			};
265			}
266	dpavlin	607	} else {
267			$log->warn("NOT IMPLEMENTED WITH $use_indexer: adding link $database -> $link->{to} [$link->{credit}]");
268			}
269			}
270			}
271			next if ($only_links);
272
273
274			#
275	dpavlin	255	# now WebPAC::Store
276			#
277	dpavlin	210	my $abs_path = abs_path($0);
278			$abs_path =~ s#/[^/]*$#/#;
279	dpavlin	141
280	dpavlin	684	my $db_path = $config->get('webpac')->{db_path} . '/' . $database;
281	dpavlin	74
282	dpavlin	301	if ($clean) {
283	dpavlin	511	$log->info("creating new database '$database' in $db_path");
284	dpavlin	301	rmtree( $db_path ) \|\| $log->warn("can't remove $db_path: $!");
285			} else {
286	dpavlin	511	$log->info("working on database '$database' in $db_path");
287	dpavlin	301	}
288	dpavlin	255
289	dpavlin	210	my $db = new WebPAC::Store(
290			path => $db_path,
291	dpavlin	217	database => $database,
292	dpavlin	301	debug => $debug,
293	dpavlin	210	);
294	dpavlin	74
295	dpavlin	233
296	dpavlin	213	#
297			# now, iterate through input formats
298			#
299	dpavlin	74
300	dpavlin	213	my @inputs;
301			if (ref($db_config->{input}) eq 'ARRAY') {
302			@inputs = @{ $db_config->{input} };
303	dpavlin	255	} elsif ($db_config->{input}) {
304			push @inputs, $db_config->{input};
305	dpavlin	213	} else {
306	dpavlin	255	$log->info("database $database doesn't have inputs defined");
307	dpavlin	213	}
308	dpavlin	74
309	dpavlin	213	foreach my $input (@inputs) {
310	dpavlin	233
311	dpavlin	512	next if ($only_input && ($input->{name} !~ m#$only_input#i && $input->{type} !~ m#$only_input#i));
312	dpavlin	510
313	dpavlin	233	my $type = lc($input->{type});
314
315	dpavlin	684	die "I know only how to handle input types ", join(",", $config->webpac('inputs') ), " not '$type'!\n" unless (grep(/$type/, $config->webpac('inputs')));
316	dpavlin	233
317	dpavlin	684	my $input_module = $config->webpac('inputs')->{$type};
318	dpavlin	74
319	dpavlin	523	$log->info("working on input '$input->{name}' in $input->{path} [type: $input->{type}] using $input_module",
320			$input->{lookup} ? "lookup '$input->{lookup}'" : ""
321			);
322	dpavlin	286
323	dpavlin	638	if ($stats) {
324			# disable modification of records if --stats is in use
325			delete($input->{modify_records});
326			delete($input->{modify_file});
327			}
328
329	dpavlin	699	warn "parser->depends = ", dump( $parser->{depends} );
330			warn "depends on: ", dump( $parser->depends($database, $input->{name}) );
331			warn "lookup_create_rules = ", dump( $parser->lookup_create_rules($database, $input->{name}) );
332			warn "parser->_lookup_create = ", dump( $parser->{_lookup_create} );
333	dpavlin	698
334			my $lookup;
335
336	dpavlin	287	my $input_db = new WebPAC::Input(
337			module => $input_module,
338	dpavlin	684	encoding => $config->webpac('webpac_encoding'),
339	dpavlin	301	limit => $limit \|\| $input->{limit},
340			offset => $offset,
341	dpavlin	585	lookup_coderef => sub {
342			my $rec = shift \|\| return;
343			$lookup->add( $rec );
344			},
345	dpavlin	416	recode => $input->{recode},
346	dpavlin	507	stats => $stats,
347	dpavlin	600	modify_records => $input->{modify_records},
348	dpavlin	636	modify_file => $input->{modify_file},
349	dpavlin	287	);
350	dpavlin	286	$log->logdie("can't create input using $input_module") unless ($input);
351	dpavlin	113
352	dpavlin	287	my $maxmfn = $input_db->open(
353	dpavlin	285	path => $input->{path},
354	dpavlin	213	code_page => $input->{encoding}, # database encoding
355	dpavlin	523	%{ $input },
356	dpavlin	213	);
357	dpavlin	113
358	dpavlin	671	my $report_fh;
359			if ($stats \|\| $validate) {
360			my $path = "out/report/" . $database . '-' . $input->{name} . '.txt';
361			open($report_fh, '>', $path) \|\| $log->logdie("can't open $path: $!");
362
363			print $report_fh "Report for database '$database' input '$input->{name}' records ",
364			$offset \|\| 1, "-", $limit \|\| $input->{limit} \|\| $maxmfn, "\n\n";
365			$log->info("Generating report file $path");
366			}
367
368	dpavlin	539	my @norm_array = ref($input->{normalize}) eq 'ARRAY' ?
369			@{ $input->{normalize} } : ( $input->{normalize} );
370	dpavlin	210
371	dpavlin	552	if ($marc_normalize) {
372			@norm_array = ( {
373			path => $marc_normalize,
374			output => $marc_output \|\| 'out/marc/' . $database . '-' . $input->{name} . '.marc',
375			} );
376			}
377
378	dpavlin	539	foreach my $normalize (@norm_array) {
379	dpavlin	492
380	dpavlin	539	my $normalize_path = $normalize->{path} \|\| $log->logdie("can't find normalize path in config");
381	dpavlin	269
382	dpavlin	539	$log->logdie("Found '$normalize_path' as normalization file which isn't supported any more!") unless ( $normalize_path =~ m!\.pl$!i );
383	dpavlin	210
384	dpavlin	539	my $rules = read_file( $normalize_path ) or die "can't open $normalize_path: $!";
385	dpavlin	213
386	dpavlin	539	$log->info("Using $normalize_path for normalization...");
387	dpavlin	213
388	dpavlin	578	my $marc = new WebPAC::Output::MARC(
389			path => $normalize->{output},
390			lint => $marc_lint,
391			dump => $marc_dump,
392			) if ($normalize->{output});
393	dpavlin	540
394	dpavlin	539	# reset position in database
395			$input_db->seek(1);
396	dpavlin	290
397	dpavlin	674	# generate name of config key for indexer (strip everything after -)
398			my $indexer_config = $use_indexer;
399	dpavlin	675	$indexer_config =~ s/^(\w+)-?.*$/$1/g if ($indexer_config);
400	dpavlin	674
401	dpavlin	539	foreach my $pos ( 0 ... $input_db->size ) {
402	dpavlin	516
403	dpavlin	539	my $row = $input_db->fetch \|\| next;
404	dpavlin	516
405	dpavlin	539	my $mfn = $row->{'000'}->[0];
406	dpavlin	213
407	dpavlin	539	if (! $mfn \|\| $mfn !~ m#^\d+$#) {
408			$log->warn("record $pos doesn't have valid MFN but '$mfn', using $pos");
409			$mfn = $pos;
410			push @{ $row->{'000'} }, $pos;
411			}
412	dpavlin	508
413	dpavlin	213
414	dpavlin	539	if ($validate) {
415	dpavlin	653	if ( my $errors = $validate->validate_errors( $row, $input_db->dump ) ) {
416			$log->error( "MFN $mfn validation error:\n",
417	dpavlin	671	$validate->report_error( $errors )
418	dpavlin	652	);
419			}
420	dpavlin	539	}
421	dpavlin	213
422	dpavlin	595	my $ds_config = dclone($db_config);
423
424			# default values -> database key
425			$ds_config->{_} = $database;
426
427			# current mfn
428			$ds_config->{_mfn} = $mfn;
429
430			# attach current input
431			$ds_config->{input} = $input;
432
433	dpavlin	539	my $ds = WebPAC::Normalize::data_structure(
434			row => $row,
435			rules => $rules,
436			lookup => $lookup ? $lookup->lookup_hash : undef,
437	dpavlin	595	config => $ds_config,
438	dpavlin	541	marc_encoding => 'utf-8',
439	dpavlin	539	);
440	dpavlin	507
441	dpavlin	539	$db->save_ds(
442			id => $mfn,
443			ds => $ds,
444			prefix => $input->{name},
445			) if ($ds && !$stats);
446	dpavlin	210
447	dpavlin	539	$indexer->add(
448			id => $input->{name} . "/" . $mfn,
449			ds => $ds,
450	dpavlin	684	type => $config->get($indexer_config)->{type},
451	dpavlin	540	) if ($indexer && $ds);
452	dpavlin	539
453	dpavlin	580	if ($marc) {
454			my $i = 0;
455	dpavlin	559
456	dpavlin	580	while (my $fields = WebPAC::Normalize::_get_marc_fields( fetch_next => 1 ) ) {
457			$marc->add(
458			id => $mfn . ( $i ? "/$i" : '' ),
459			fields => $fields,
460			leader => WebPAC::Normalize::marc_leader(),
461			row => $row,
462			);
463			$i++;
464			}
465
466			$log->info("Created $i instances of MFN $mfn\n") if ($i > 1);
467			}
468
469	dpavlin	539	$total_rows++;
470			}
471
472	dpavlin	664	if ($validate) {
473			my $errors = $validate->report;
474	dpavlin	671	if ($errors) {
475			$log->info("validation errors:\n$errors\n" );
476			print $report_fh "$errors\n" if ($report_fh);
477			}
478	dpavlin	657	}
479
480	dpavlin	671	if ($stats) {
481			my $s = $input_db->stats;
482			$log->info("statistics of fields usage:\n$s");
483			print $report_fh "Statistics of fields usage:\n$s" if ($report_fh);
484			}
485	dpavlin	539
486	dpavlin	540	# close MARC file
487	dpavlin	578	$marc->finish if ($marc);
488	dpavlin	539
489	dpavlin	671	# close report
490	dpavlin	672	close($report_fh) if ($report_fh)
491	dpavlin	540	}
492
493	dpavlin	539	}
494
495	dpavlin	509	eval { $indexer->finish } if ($indexer && $indexer->can('finish'));
496	dpavlin	434
497	dpavlin	389	my $dt = time() - $start_t;
498	dpavlin	511	$log->info("$total_rows records ", $indexer ? "indexed " : "",
499			sprintf("in %.2f sec [%.2f rec/sec]",
500	dpavlin	389	$dt, ($total_rows / $dt)
501			)
502			);
503	dpavlin	255
504
505	dpavlin	606	# end forked process
506			if ($parallel) {
507			$log->info("parallel process $$ finished");
508			exit(0);
509			}
510
511	dpavlin	210	}
512
513	dpavlin	606	if ($parallel) {
514			# wait all children to finish
515			sleep(1) while wait != -1;
516			$log->info("all parallel processes finished");
517			}
518
519	dpavlin	608	#
520			# handle links or merge after indexing
521			#
522	dpavlin	607
523	dpavlin	608	if ($merge) {
524			print $estcmd_fh 'sudo /etc/init.d/hyperestraier start',$/;
525			close($estcmd_fh);
526			chmod 0700, $estcmd_path \|\| $log->warn("can't chmod 0700 $estcmd_path: $!");
527			system $estcmd_path;
528			} else {
529			foreach my $link (@links) {
530			$log->logdie("coderef in link ", Dumper($link), " is ", ref($link), " and not CODE") unless (ref($link) eq 'CODE');
531			$link->();
532			}
533	dpavlin	529	}