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::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);

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 $lookup;
                if ($input->{lookup}) {
                        $lookup = new WebPAC::Lookup(
                                lookup_file => $input->{lookup},
                        );
                        delete( $input->{lookup} );
                }

                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});
                }

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