1 |
#!/usr/bin/perl -w |
#!/usr/bin/perl -w |
2 |
|
|
3 |
use strict; |
=head1 NAME |
|
use OpenIsis; |
|
|
use Getopt::Std; |
|
|
use Data::Dumper; |
|
|
use XML::Simple; |
|
|
use Text::Unaccent 1.02; # 1.01 won't compile on my platform, |
|
|
use Text::Iconv; |
|
|
use Config::IniFiles; |
|
|
use Encode; |
|
|
#use GDBM_File; |
|
|
use Fcntl; # for O_RDWR |
|
|
use TDB_File; |
|
|
|
|
|
$|=1; |
|
|
|
|
|
my $config_file = $0; |
|
|
$config_file =~ s/\.pl$/.conf/; |
|
|
die "FATAL: can't find configuration file '$config_file'" if (! -e $config_file); |
|
|
|
|
|
my $config; |
|
|
|
|
|
#use index_DBI; # default DBI module for index |
|
|
use index_DBI_cache; # faster DBI module using memory cache |
|
|
my $index; |
|
|
|
|
|
my %opts; |
|
|
|
|
|
# usage: |
|
|
# -d directory name |
|
|
# -m multiple directories |
|
|
# -q quiet |
|
|
# -s run swish |
|
|
|
|
|
getopts('d:m:qs', \%opts); |
|
|
|
|
|
my $path; # this is name of database |
|
4 |
|
|
5 |
Text::Iconv->raise_error(0); # Conversion errors don't raise exceptions |
all2all.pl - basic script for all WebPAC needs |
6 |
|
|
7 |
# this is encoding of all files on disk, including import_xml/*.xml file and |
=cut |
|
# filter/*.pm files! It will be used to store strings in perl internally! |
|
|
my $codepage = 'ISO-8859-2'; |
|
8 |
|
|
9 |
my $utf2cp = Text::Iconv->new('UTF-8',$codepage); |
use strict; |
10 |
# this function will convert data from XML files to local encoding |
use Data::Dumper; |
11 |
sub x { |
use Carp; |
|
return $utf2cp->convert($_[0]); |
|
|
} |
|
12 |
|
|
13 |
# decode isis/excel or other import codepage |
use lib './lib'; |
14 |
my $import2cp; |
use WebPAC; |
15 |
|
use WebPAC::jsFind; |
16 |
|
|
17 |
|
my $webpac = new WebPAC( |
18 |
|
code_page => 'ISO-8859-2', |
19 |
|
limit_mfn => 100, |
20 |
|
# debug => 1, |
21 |
|
) || die; |
22 |
|
|
23 |
|
my $log = $webpac->_get_logger(); |
24 |
|
|
25 |
|
my $index = new WebPAC::jsFind( |
26 |
|
index_path => './out/index', |
27 |
|
) || die; |
28 |
|
|
29 |
# outgoing xml must be in UTF-8 |
$|=1; |
|
my $cp2utf = Text::Iconv->new($codepage,'UTF-8'); |
|
30 |
|
|
31 |
# mapping between data type and tag which specify |
my $maxmfn = $webpac->open_isis( |
32 |
# format in XML file |
filename => shift @ARGV || '/data/hidra/THS/THS', |
33 |
my %type2tag = ( |
lookup => [ |
34 |
'isis' => 'isis', |
{ 'key' => 'd:v900', 'val' => 'v250^a' }, |
35 |
'excel' => 'column', |
# { 'eval' => '"v901^a" eq "Područje"', 'key' => 'pa:v561^4:v562^4:v461^1', 'val' => 'v900' }, |
36 |
'marc' => 'marc', |
# { 'eval '=> '"v901^a" eq "Mikrotezaurus"', 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' }, |
37 |
'feed' => 'feed' |
# { 'eval' => '"v901^a" eq "Deskriptor"', 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' }, |
38 |
|
{ 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' }, |
39 |
|
{ 'key' => '900_mfn:v900', 'val' => 'v000' }, |
40 |
|
], |
41 |
); |
); |
42 |
|
|
43 |
my $cache; # for cacheing |
$log->info("rows: $maxmfn"); |
|
|
|
|
# lookup hash (tied to file) |
|
|
my %lhash; |
|
|
# this option will cache all lookup entries in memory. |
|
|
# if you are tight on memory, turn this off |
|
|
my $use_lhash_cache = 1; |
|
|
|
|
|
sub data2xml { |
|
|
|
|
|
use xmlify; |
|
|
|
|
|
my $type = shift @_; |
|
|
my $row = shift @_; |
|
|
my $add_xml = shift @_; |
|
|
# needed to read values from configuration file |
|
|
my $cfg = shift @_; |
|
|
my $database = shift @_; |
|
44 |
|
|
45 |
my $xml; |
$webpac->open_import_xml(type => 'isis_hidra_ths'); |
46 |
|
|
47 |
use parse_format; |
while (my $rec = $webpac->fetch_rec) { |
48 |
|
|
49 |
my $html = ""; # html formatted display output |
my @ds = $webpac->data_structure($rec); |
50 |
|
|
51 |
my %field_usage; # counter for usage of each field |
if ($log->is_debug) { |
52 |
|
$log->debug("rec = ",Dumper($rec)); |
53 |
|
$log->debug("ds = ",Dumper(\@ds)); |
54 |
|
} |
55 |
|
|
56 |
# sort subrouting using order="" attribute |
next if (! @ds); |
|
sub by_order { |
|
|
my $va = $config->{indexer}->{$a}->{order} || |
|
|
$config->{indexer}->{$a}; |
|
|
my $vb = $config->{indexer}->{$b}->{order} || |
|
|
$config->{indexer}->{$b}; |
|
57 |
|
|
58 |
return $va <=> $vb; |
my $filename = $webpac->{'current_filename'}; |
|
} |
|
59 |
|
|
60 |
my @sorted_tags; |
if ($filename) { |
61 |
if ($cache->{tags_by_order}) { |
open(OUT,"> $filename") || $log->logdie("can't open output '$filename': $!"); |
62 |
@sorted_tags = @{$cache->{tags_by_order}}; |
print OUT $webpac->output( |
63 |
|
template => 'html.tt', |
64 |
|
data => \@ds, |
65 |
|
); |
66 |
|
close(OUT); |
67 |
} else { |
} else { |
68 |
@sorted_tags = sort by_order keys %{$config->{indexer}}; |
print $webpac->output( |
69 |
$cache->{tags_by_order} = \@sorted_tags; |
template => 'text.tt', |
70 |
|
data => \@ds, |
71 |
|
); |
72 |
} |
} |
73 |
|
|
74 |
# lookup key |
my $headline; |
75 |
my $lookup_key; |
foreach my $ds (@ds) { |
76 |
|
if ($ds->{'tag'} eq 'headline') { |
77 |
# cache for field in pages |
$headline = join(" ",@{$ds->{'display'}}); |
78 |
delete $cache->{display_data}; |
last; |
|
delete $cache->{swish_data}; |
|
|
delete $cache->{swish_exact_data}; |
|
|
my @page_fields; # names of fields |
|
|
|
|
|
|
|
|
# subs used to produce output |
|
|
|
|
|
sub get_field_name($$$) { |
|
|
my ($config,$field,$field_usage) = @_; |
|
|
|
|
|
# find field name (signular, plural) |
|
|
my $field_name = ""; |
|
|
if ($config->{indexer}->{$field}->{name_singular} && $field_usage == 1) { |
|
|
$field_name = $config->{indexer}->{$field}->{name_singular}; |
|
|
} elsif ($config->{indexer}->{$field}->{name_plural}) { |
|
|
$field_name = $config->{indexer}->{$field}->{name_plural}; |
|
|
} elsif ($config->{indexer}->{$field}->{name}) { |
|
|
$field_name = $config->{indexer}->{$field}->{name}; |
|
|
} else { |
|
|
print STDERR "WARNING: field '$field' doesn't have 'name' attribute!"; |
|
|
} |
|
|
if ($field_name) { |
|
|
return x($field_name); |
|
79 |
} |
} |
80 |
} |
} |
81 |
|
|
82 |
|
my $f = $filename; |
83 |
|
$f =~ s!out/!!; |
84 |
|
|
85 |
# begin real work: go field by field |
# save into index |
86 |
foreach my $field (@sorted_tags) { |
foreach my $ds (@ds) { |
87 |
|
next if (! $ds->{'swish'}); |
|
$field=x($field); |
|
|
$field_usage{$field}++; |
|
|
|
|
|
my $swish_data = ""; |
|
|
my $swish_exact_data = ""; |
|
|
my $display_data = ""; |
|
|
my $line_delimiter; |
|
|
|
|
|
my ($swish,$display); |
|
|
|
|
|
my $tag = $type2tag{$type} || die "can't find which tag to use for type $type"; |
|
|
|
|
|
# is this field page-by-page? |
|
|
my $iterate_by_page = $config->{indexer}->{$field}->{iterate_by_page}; |
|
|
push @page_fields,$field if ($iterate_by_page); |
|
|
my %page_max = (); |
|
|
# default line_delimiter if using |
|
|
my $page_line_delimiter = $config->{indexer}->{$field}->{page_line_delimiter} || '<br/>'; |
|
|
|
|
|
foreach my $x (@{$config->{indexer}->{$field}->{$tag}}) { |
|
|
|
|
|
my $format = x($x->{content}); |
|
|
my $delimiter = x($x->{delimiter}) || ' '; |
|
|
|
|
|
my $repeat_off = 0; # init repeatable offset |
|
|
|
|
|
# swish, swish_exact, display, index, index_lookup |
|
|
# swish and display defaults |
|
|
my ($s,$se,$d,$i,$il) = (1,0,1,0,0); |
|
|
$s = 0 if (lc($x->{type}) eq "display"); |
|
|
$d = 0 if (lc($x->{type}) eq "swish"); |
|
|
$se = 1 if (lc($x->{type}) eq "swish_exact"); |
|
|
($s,$d,$i) = (0,0,1) if (lc($x->{type}) eq "index"); |
|
|
$il = 1 if (lc($x->{type}) =~ /^lookup/); |
|
|
|
|
|
|
|
|
# what will separate last line from this one? |
|
|
if ($display_data && $x->{append} && $x->{append} eq "1") { |
|
|
$line_delimiter = ' '; |
|
|
} elsif ($display_data) { |
|
|
$line_delimiter = '<br/>'; |
|
|
} |
|
|
|
|
|
# init vars so that we go into while... |
|
|
($swish,$display) = (1,1); |
|
|
|
|
|
# placeholder for all repeatable entries for index |
|
|
my @index_data; |
|
|
|
|
|
sub mkformat { |
|
|
my $x = shift || die "mkformat needs tag reference"; |
|
|
my $data = shift || return; |
|
|
my $format_name = x($x->{format_name}) || return $data; |
|
|
my $fmt = x($config->{format}->{$format_name}->{content}) || die "<format name=\"$format_name\"> is not defined!"; |
|
|
my $format_delimiter = x($x->{format_delimiter}); |
|
|
my @data; |
|
|
if ($format_delimiter) { |
|
|
@data = split(/$format_delimiter/,$data); |
|
|
} else { |
|
|
push @data,$data; |
|
|
} |
|
|
|
|
|
if ($fmt) { |
|
|
my $nr = scalar $fmt =~ s/%s/%s/g; |
|
|
if (($#data+1) == $nr) { |
|
|
return sprintf($fmt,@data); |
|
|
} else { |
|
|
print STDERR "mkformat: [$data] can't be split on [$format_delimiter] to $nr fields!\n"; |
|
|
return $data; |
|
|
} |
|
|
} else { |
|
|
print STDERR "usage of link '$format_name' without defined format (<link> tag)\n"; |
|
|
} |
|
|
} |
|
|
|
|
|
# while because of repeatable fields |
|
|
while ($swish || $display) { |
|
|
my $page = $repeat_off; |
|
|
$page_max{$field} = $page if ($iterate_by_page && $page > ($page_max{$field} || 0)); |
|
|
($swish,$display) = parse_format($type, $format,$row,$repeat_off++,$import2cp); |
|
|
if ($repeat_off > 1000) { |
|
|
print STDERR "loop (more than 1000 repeatable fields) deteced in $row, $format\n"; |
|
|
last; |
|
|
} |
|
|
|
|
|
# is this field is lookup? |
|
|
if ($display && $x->{lookup}) { |
|
|
my $null = "<!-- null -->"; |
|
|
if ($use_lhash_cache) { |
|
|
if (!defined($cache->{lhash}->{$display})) { |
|
|
my $new_display = $lhash{$display}; |
|
|
if (defined($new_display)) { |
|
|
#print STDERR "lookup cache store '$display' = '$new_display'\n"; |
|
|
$display = $new_display; |
|
|
$cache->{lhash}->{$display} = $new_display; |
|
|
} else { |
|
|
print STDERR "WARNING: lookup for '$display' didn't find anything.\n"; |
|
|
$display = ""; |
|
|
$cache->{lhash}->{$display} = $null; |
|
|
} |
|
|
} else { |
|
|
$display = $cache->{lhash}->{$display}; |
|
|
} |
|
|
} else { |
|
|
$display = $lhash{$display} || $null; |
|
|
} |
|
|
} |
|
|
|
|
|
# filter="name" ; filter this field through |
|
|
# filter/[name].pm |
|
|
my $filter = $x->{filter}; |
|
|
if ($filter && !$cache->{filter_loaded}->{$filter}) { |
|
|
require "filter/".$filter.".pm"; |
|
|
$cache->{filter_loaded}->{$filter}++; |
|
|
} |
|
|
# type="swish" ; field for swish |
|
|
if ($swish) { |
|
|
if ($filter && ($s || $se)) { |
|
|
no strict 'refs'; |
|
|
my $tmp = join(" ",&$filter($swish)) if ($s || $se); |
|
|
$swish_data .= $tmp if ($s); |
|
|
$swish_exact_data .= $tmp if ($se); |
|
|
|
|
|
} else { |
|
|
$swish_data .= $swish if ($s); |
|
|
$swish_exact_data .= $swish if ($se); |
|
|
} |
|
|
} |
|
|
|
|
|
# type="display" ; field for display |
|
|
if ($d && $display) { |
|
|
if ($line_delimiter && $display_data) { |
|
|
$display_data .= $line_delimiter; |
|
|
} |
|
|
if ($filter) { |
|
|
no strict 'refs'; |
|
|
if ($display_data) { |
|
|
$display_data .= $delimiter.join($delimiter,mkformat($x,&$filter($display))); |
|
|
} else { |
|
|
$display_data = join($delimiter,mkformat($x,&$filter($display))); |
|
|
} |
|
|
} else { |
|
|
if ($display_data) { |
|
|
$display_data .= $delimiter.mkformat($x,$display); |
|
|
} else { |
|
|
$display_data = mkformat($x,$display); |
|
|
} |
|
|
} |
|
|
} |
|
|
|
|
|
# type="index" ; insert into index |
|
|
if ($i && $display) { |
|
|
if ($filter) { |
|
|
no strict 'refs'; |
|
|
$display = &$filter($display); |
|
|
} |
|
|
if ($x->{append} && @index_data) { |
|
|
$index_data[$#index_data].=$display; |
|
|
} else { |
|
|
push @index_data, $display; |
|
|
} |
|
|
} |
|
|
|
|
|
# store fields in lookup |
|
|
if ($il && $display) { |
|
|
if (lc($x->{type}) eq "lookup_key") { |
|
|
if ($lookup_key) { |
|
|
print STDERR "WARNING: try to redefine lookup_key (keys shouldn't be repeatable fields!)"; |
|
|
} else { |
|
|
$lookup_key = $display; |
|
|
} |
|
|
} elsif (lc($x->{type}) eq "lookup_val") { |
|
|
if ($lookup_key) { |
|
|
$lhash{$lookup_key} = $display; |
|
|
} else { |
|
|
print STDERR "WARNING: no lookup_key defined for '$display'?"; |
|
|
} |
|
|
} |
|
|
|
|
|
} |
|
|
|
|
|
# store data for page-by-page repeatable fields |
|
|
if ($iterate_by_page) { |
|
|
sub iterate_fld($$$$$$) { |
|
|
my ($cache,$what,$field,$page,$data,$append) = @_; |
|
|
return if (!$data); |
|
|
my $line_delimiter = $page_line_delimiter; |
|
|
$line_delimiter = " " if ($append); |
|
|
if (! $cache->{$what}->{$field}->[$page]) { |
|
|
$cache->{$what}->{$field}->[$page] = $data; |
|
|
} else { |
|
|
$cache->{$what}->{$field}->[$page] .= $line_delimiter.$data; |
|
|
} |
|
|
} |
|
|
|
|
|
if ($display_data) { |
|
|
print STDERR "line delimiter: ",Dumper($line_delimiter) if ($line_delimiter); |
|
|
iterate_fld($cache,'display_data',$field,$page,$display_data,$x->{append}); |
|
|
} |
|
|
$display_data = ""; |
|
|
if ($swish_data) { |
|
|
iterate_fld($cache,'swish_data',$field,$page,$swish_data,$x->{append}); |
|
|
$swish_data = ""; |
|
|
} |
|
|
if ($swish_exact_data) { |
|
|
iterate_fld($cache,'swish_exact_data',$field,$page,$swish_exact_data,$x->{append}); |
|
|
$swish_exact_data = ""; |
|
|
} |
|
|
} |
|
|
} |
|
|
|
|
|
# fill data in index |
|
|
foreach my $d (@index_data) { |
|
|
$index->insert($field, $d, $path); |
|
|
} |
|
|
} |
|
|
|
|
|
# now try to parse variables from configuration file |
|
|
foreach my $x (@{$config->{indexer}->{$field}->{'config'}}) { |
|
|
|
|
|
my $delimiter = x($x->{delimiter}) || ' '; |
|
|
my $val = $cfg->val($database, x($x->{content})); |
|
|
|
|
|
my ($s,$d,$i) = (1,1,0); # swish, display default |
|
|
$s = 0 if (lc($x->{type}) eq "display"); |
|
|
$d = 0 if (lc($x->{type}) eq "swish"); |
|
|
# no support for swish exact in config. |
|
|
# IMHO, it's useless |
|
|
($s,$d,$i) = (0,0,1) if (lc($x->{type}) eq "index"); |
|
|
|
|
|
if ($val) { |
|
|
$display_data .= $delimiter.$val if ($d); |
|
|
$swish_data .= $val if ($s); |
|
|
$index->insert($field, $val, $path) if ($i); |
|
|
} |
|
|
|
|
|
if ($iterate_by_page) { |
|
|
# FIXME data from config tag will appear just |
|
|
# on first page!!! |
|
|
my $page = 0; |
|
|
if ($display_data) { |
|
|
$cache->{display_data}->{$field}->[$page] = $display_data; |
|
|
$display_data = ""; |
|
|
} |
|
|
if ($swish_data) { |
|
|
$cache->{swish_data}->{$field}->[$page] = $swish_data; |
|
|
$swish_data = ""; |
|
|
} |
|
|
if ($swish_exact_data) { |
|
|
$cache->{swish_exact_data}->{$field}->[$page] = $swish_exact_data; |
|
|
$swish_exact_data = ""; |
|
|
} |
|
|
} |
|
|
} |
|
|
|
|
|
# save data page-by-page |
|
|
foreach my $field (@page_fields) { |
|
|
my $nr_pages = $page_max{$field} || next; |
|
|
#print STDERR "field '$field' iterate over ",($nr_pages || 0)," pages...\n"; |
|
|
#print STDERR Dumper($cache->{display_data}); |
|
|
for (my $page=0; $page <= $nr_pages; $page++) { |
|
|
|
|
|
my $display_data = $cache->{display_data}->{$field}->[$page]; |
|
|
if ($display_data) { # default |
|
|
if ($field eq "headline") { |
|
|
$xml .= xmlify("headline", $display_data); |
|
|
} else { |
|
|
|
|
|
# fallback to empty field name if needed |
|
|
$html .= get_field_name($config,$field,$field_usage{$field}) || ''; |
|
|
$html .= "#-#".$display_data."###\n"; |
|
|
} |
|
|
} |
|
|
|
|
|
my $swish_data = $cache->{swish_data}->{$field}->[$page]; |
|
|
if ($swish_data) { |
|
|
# remove extra spaces |
|
|
$swish_data =~ s/ +/ /g; |
|
|
$swish_data =~ s/ +$//g; |
|
|
|
|
|
$xml .= xmlify($field."_swish", unac_string($codepage,$swish_data)); |
|
|
} |
|
|
|
|
|
my $swish_exact_data = $cache->{swish_exact_data}->{$field}->[$page]; |
|
|
if ($swish_exact_data) { |
|
|
$swish_exact_data =~ s/ +/ /g; |
|
|
$swish_exact_data =~ s/ +$//g; |
|
|
|
|
|
# add delimiters before and after word. |
|
|
# That is required to produce exact match |
|
|
$xml .= xmlify($field."_swish_exact", unac_string($codepage,'xxbxx '.$swish_exact_data.' xxexx')); |
|
|
} |
|
|
} |
|
|
|
|
|
} |
|
|
|
|
|
if (! $iterate_by_page) { |
|
|
if ($display_data) { |
|
|
if ($field eq "headline") { |
|
|
$xml .= xmlify("headline", $display_data); |
|
|
} else { |
|
|
|
|
|
# fallback to empty field name if needed |
|
|
$html .= get_field_name($config,$field,$field_usage{$field}) || ''; |
|
|
$html .= "#-#".$display_data."###\n"; |
|
|
} |
|
|
} |
|
|
if ($swish_data) { |
|
|
# remove extra spaces |
|
|
$swish_data =~ s/ +/ /g; |
|
|
$swish_data =~ s/ +$//g; |
|
|
|
|
|
$xml .= xmlify($field."_swish", unac_string($codepage,$swish_data)); |
|
|
} |
|
|
|
|
|
if ($swish_exact_data) { |
|
|
$swish_exact_data =~ s/ +/ /g; |
|
|
$swish_exact_data =~ s/ +$//g; |
|
|
|
|
|
# add delimiters before and after word. |
|
|
# That is required to produce exact match |
|
|
$xml .= xmlify($field."_swish_exact", unac_string($codepage,'xxbxx '.$swish_exact_data.' xxexx')); |
|
|
} |
|
|
} |
|
|
} |
|
88 |
|
|
89 |
# dump formatted output in <html> |
$index->insert( |
90 |
if ($html) { |
index_name => $ds->{'tag'}, |
91 |
#$xml .= xmlify("html",$html); |
path => $f, |
92 |
$xml .= "<html><![CDATA[ $html ]]></html>"; |
headline => $headline, |
93 |
} |
words => join(" ",@{$ds->{'swish'}}) |
94 |
|
); |
|
if ($xml) { |
|
|
$xml .= $add_xml if ($add_xml); |
|
|
return "<xml>\n$xml</xml>\n"; |
|
|
} else { |
|
|
return; |
|
95 |
} |
} |
|
} |
|
96 |
|
|
97 |
########################################################################## |
# print Dumper(\@ds); |
98 |
|
|
|
# read configuration for this script |
|
|
my $cfg = new Config::IniFiles( -file => $config_file ); |
|
|
|
|
|
# read global.conf configuration |
|
|
my $cfg_global = new Config::IniFiles( -file => 'global.conf' ); |
|
|
|
|
|
# open index |
|
|
$index = new index_DBI( |
|
|
$cfg_global->val('global', 'dbi_dbd'), |
|
|
$cfg_global->val('global', 'dbi_dsn'), |
|
|
$cfg_global->val('global', 'dbi_user'), |
|
|
$cfg_global->val('global', 'dbi_passwd') || '', |
|
|
); |
|
|
|
|
|
my $show_progress = $cfg_global->val('global', 'show_progress'); |
|
|
|
|
|
my $unac_filter = $cfg_global->val('global', 'unac_filter'); |
|
|
if ($unac_filter) { |
|
|
require $unac_filter; |
|
99 |
} |
} |
100 |
|
|
101 |
foreach my $database ($cfg->Sections) { |
if ($log->is_debug) { |
102 |
|
$log->debug("lookup hash: ",Dumper($webpac->{'lookup'})); |
103 |
my $type = lc($cfg -> val($database, 'type')) || die "$database doesn't have 'type' defined"; |
$log->debug("data hash: ",Dumper($webpac->{'data'})); |
|
my $add_xml = $cfg -> val($database, 'xml'); # optional |
|
|
|
|
|
# create new lookup file |
|
|
my $lookup_file = $cfg -> val($database, 'lookup_newfile'); # optional |
|
|
if ($lookup_file) { |
|
|
#tie %lhash, 'GDBM_File', $lookup_file, &GDBM_NEWDB, 0644; |
|
|
tie %lhash, 'TDB_File', $lookup_file, TDB_CLEAR_IF_FIRST, O_RDWR, 0644; |
|
|
print STDERR "creating lookup file '$lookup_file'\n"; |
|
|
} |
|
|
|
|
|
# open existing lookup file |
|
|
$lookup_file = $cfg -> val($database, 'lookup_open'); # optional |
|
|
if ($lookup_file) { |
|
|
#tie %lhash, 'GDBM_File', $lookup_file, &GDBM_READER, 0644; |
|
|
tie %lhash, 'TDB_File', $lookup_file, TDB_DEFAULT, O_RDWR, 0644; |
|
|
print STDERR "opening lookup file '$lookup_file'\n"; |
|
|
} |
|
|
|
|
|
print STDERR "reading ./import_xml/$type.xml\n"; |
|
|
|
|
|
# extract just type basic |
|
|
my $type_base = $type; |
|
|
$type_base =~ s/_.+$//g; |
|
|
|
|
|
$config=XMLin("./import_xml/$type.xml", forcearray => [ $type2tag{$type_base}, 'config', 'format' ], forcecontent => 1); |
|
|
|
|
|
# output current progress indicator |
|
|
my $last_p = 0; |
|
|
sub progress { |
|
|
return if (! $show_progress); |
|
|
my $current = shift; |
|
|
my $total = shift || 1; |
|
|
my $p = int($current * 100 / $total); |
|
|
if ($p != $last_p) { |
|
|
printf STDERR ("%5d / %5d [%-51s] %-2d %% \r",$current,$total,"=" x ($p/2).">", $p ); |
|
|
$last_p = $p; |
|
|
} |
|
|
} |
|
|
|
|
|
my $fake_dir = 1; |
|
|
sub fakeprogress { |
|
|
return if (! $show_progress); |
|
|
my $current = shift @_; |
|
|
|
|
|
my @ind = ('-','\\','|','/','-','\\','|','/', '-'); |
|
|
|
|
|
$last_p += $fake_dir; |
|
|
$fake_dir = -$fake_dir if ($last_p > 1000 || $last_p < 0); |
|
|
if ($last_p % 10 == 0) { |
|
|
printf STDERR ("%5d / %5s [%-51s]\r",$current,"?"," " x ($last_p/20).$ind[($last_p/20) % $#ind]); |
|
|
} |
|
|
} |
|
|
|
|
|
# now read database |
|
|
print STDERR "using: $type...\n"; |
|
|
|
|
|
# erase cache for tags by order in this database |
|
|
delete $cache->{tags_by_order}; |
|
|
|
|
|
if ($type_base eq "isis") { |
|
|
|
|
|
my $isis_db = $cfg -> val($database, 'isis_db') || die "$database doesn't have 'isis_db' defined!"; |
|
|
|
|
|
$import2cp = Text::Iconv->new($config->{isis_codepage},$codepage); |
|
|
my $db = OpenIsis::open( $isis_db ); |
|
|
|
|
|
# check if .txt database for OpenIsis is zero length, |
|
|
# if so, erase it and re-open database |
|
|
sub check_txt_db { |
|
|
my $isis_db = shift || die "need isis database name"; |
|
|
my $reopen = 0; |
|
|
|
|
|
if (-e $isis_db.".TXT") { |
|
|
print STDERR "WARNING: removing $isis_db.TXT OpenIsis database...\n"; |
|
|
unlink $isis_db.".TXT" || warn "FATAL: unlink error on '$isis_db.TXT': $!"; |
|
|
$reopen++; |
|
|
} |
|
|
if (-e $isis_db.".PTR") { |
|
|
print STDERR "WARNING: removing $isis_db.PTR OpenIsis database...\n"; |
|
|
unlink $isis_db.".PTR" || warn "FATAL: unlink error on '$isis_db.PTR': $!"; |
|
|
$reopen++; |
|
|
} |
|
|
return OpenIsis::open( $isis_db ) if ($reopen); |
|
|
} |
|
|
|
|
|
# EOF error |
|
|
if ($db == -1) { |
|
|
$db = check_txt_db($isis_db); |
|
|
if ($db == -1) { |
|
|
print STDERR "FATAL: OpenIsis can't open zero size file $isis_db\n"; |
|
|
next; |
|
|
} |
|
|
} |
|
|
|
|
|
# OpenIsis::ERR_BADF |
|
|
if ($db == -4) { |
|
|
print STDERR "FATAL: OpenIsis can't find file $isis_db\n"; |
|
|
next; |
|
|
# OpenIsis::ERR_IO |
|
|
} elsif ($db == -5) { |
|
|
print STDERR "FATAL: OpenIsis can't access file $isis_db\n"; |
|
|
next; |
|
|
} elsif ($db < 0) { |
|
|
print STDERR "FATAL: OpenIsis unknown error $db with file $isis_db\n"; |
|
|
next; |
|
|
} |
|
|
|
|
|
my $max_rowid = OpenIsis::maxRowid( $db ); |
|
|
|
|
|
# if 0 records, try to rease isis .txt database |
|
|
if ($max_rowid == 0) { |
|
|
# force removal of database |
|
|
$db = check_txt_db($isis_db); |
|
|
$max_rowid = OpenIsis::maxRowid( $db ); |
|
|
} |
|
|
|
|
|
print STDERR "Reading database: $isis_db [$max_rowid rows]\n"; |
|
|
|
|
|
my $path = $database; |
|
|
|
|
|
for (my $row_id = 1; $row_id <= $max_rowid; $row_id++ ) { |
|
|
my $row = OpenIsis::read( $db, $row_id ); |
|
|
if ($row && $row->{mfn}) { |
|
|
|
|
|
progress($row->{mfn}, $max_rowid); |
|
|
|
|
|
my $swishpath = $path."#".int($row->{mfn}); |
|
|
|
|
|
if (my $xml = data2xml($type_base,$row,$add_xml,$cfg,$database)) { |
|
|
$xml = $cp2utf->convert($xml); |
|
|
use bytes; # as opposed to chars |
|
|
print "Path-Name: $swishpath\n"; |
|
|
print "Content-Length: ".(length($xml)+1)."\n"; |
|
|
print "Document-Type: XML\n\n$xml\n"; |
|
|
} |
|
|
} |
|
|
} |
|
|
# for this to work with current version of OpenIsis (0.9.0) |
|
|
# you might need my patch from |
|
|
# http://www.rot13.org/~dpavlin/projects/openisis-0.9.0-perl_close.diff |
|
|
OpenIsis::close($db); |
|
|
print STDERR "\n"; |
|
|
|
|
|
} elsif ($type_base eq "excel") { |
|
|
use Spreadsheet::ParseExcel; |
|
|
use Spreadsheet::ParseExcel::Utility qw(int2col); |
|
|
|
|
|
$import2cp = Text::Iconv->new($config->{excel_codepage},$codepage); |
|
|
my $excel_file = $cfg -> val($database, 'excel_file') || die "$database doesn't have 'excel_file' defined!"; |
|
|
|
|
|
my $sheet = x($config->{sheet}) || die "no sheet in $type.xml"; |
|
|
my $start_row = x($config->{start_row}) - 1 || die "no start_row in $type.xml"; |
|
|
|
|
|
my $oBook = Spreadsheet::ParseExcel::Workbook->Parse($excel_file) || die "can't open Excel file '$excel_file'"; |
|
|
|
|
|
my $sheet_nr = 0; |
|
|
foreach my $oWks (@{$oBook->{Worksheet}}) { |
|
|
#print STDERR "-- SHEET $sheet_nr:", $oWks->{Name}, "\n"; |
|
|
last if ($oWks->{Name} eq $sheet); |
|
|
$sheet_nr++; |
|
|
} |
|
|
|
|
|
my $oWorksheet = $oBook->{Worksheet}[$sheet_nr]; |
|
|
|
|
|
print STDERR "using sheet: ",$oWorksheet->{Name},"\n"; |
|
|
defined ($oWorksheet) || die "can't find sheet '$sheet' in $excel_file"; |
|
|
my $end_row = x($config->{end_row}) || $oWorksheet->{MaxRow}; |
|
|
|
|
|
for(my $iR = $start_row ; defined $end_row && $iR <= $end_row ; $iR++) { |
|
|
my $row; |
|
|
for(my $iC = $oWorksheet->{MinCol} ; defined $oWorksheet->{MaxCol} && $iC <= $oWorksheet->{MaxCol} ; $iC++) { |
|
|
my $cell = $oWorksheet->{Cells}[$iR][$iC]; |
|
|
if ($cell) { |
|
|
$row->{int2col($iC)} = $cell->Value; |
|
|
} |
|
|
} |
|
|
|
|
|
progress($iR, $end_row); |
|
|
|
|
|
# print "row[$iR/$end_row] "; |
|
|
# foreach (keys %{$row}) { |
|
|
# print "$_: ",$row->{$_},"\t"; |
|
|
# } |
|
|
# print "\n"; |
|
|
|
|
|
my $swishpath = $database."#".$iR; |
|
|
|
|
|
next if (! $row); |
|
|
|
|
|
if (my $xml = data2xml($type_base,$row,$add_xml,$cfg,$database)) { |
|
|
$xml = $cp2utf->convert($xml); |
|
|
use bytes; # as opposed to chars |
|
|
print "Path-Name: $swishpath\n"; |
|
|
print "Content-Length: ".(length($xml)+1)."\n"; |
|
|
print "Document-Type: XML\n\n$xml\n"; |
|
|
} |
|
|
} |
|
|
} elsif ($type_base eq "marc") { |
|
|
|
|
|
use MARC; |
|
|
|
|
|
$import2cp = Text::Iconv->new($config->{marc_codepage},$codepage); |
|
|
my $marc_file = $cfg -> val($database, 'marc_file') || die "$database doesn't have 'marc_file' defined!"; |
|
|
|
|
|
# optional argument is format |
|
|
my $format = x($config->{format}) || 'usmarc'; |
|
|
|
|
|
print STDERR "Reading MARC file '$marc_file'\n"; |
|
|
|
|
|
my $marc = new MARC; |
|
|
my $nr = $marc->openmarc({ |
|
|
file=>$marc_file, format=>$format |
|
|
}) || die "Can't open MARC file '$marc_file'"; |
|
|
|
|
|
my $i=0; # record nr. |
|
|
|
|
|
my $rec; |
|
|
|
|
|
while ($marc->nextmarc(1)) { |
|
|
|
|
|
# XXX |
|
|
fakeprogress($i++); |
|
|
|
|
|
my $swishpath = $database."#".$i; |
|
|
|
|
|
if (my $xml = data2xml($type_base,$marc,$add_xml,$cfg,$database)) { |
|
|
$xml = $cp2utf->convert($xml); |
|
|
use bytes; # as opposed to chars |
|
|
print "Path-Name: $swishpath\n"; |
|
|
print "Content-Length: ".(length($xml)+1)."\n"; |
|
|
print "Document-Type: XML\n\n$xml\n"; |
|
|
} |
|
|
} |
|
|
} elsif ($type_base eq "feed") { |
|
|
|
|
|
$import2cp = Text::Iconv->new($config->{feed_codepage},$codepage); |
|
|
my $prog = x($config->{prog}) || die "$database doesn't have 'prog' defined!"; |
|
|
|
|
|
print STDERR "Reading feed from program '$prog'\n"; |
|
|
|
|
|
open(FEED,"feeds/$prog |") || die "can't start $prog: $!"; |
|
|
|
|
|
my $i=1; # record nr. |
|
|
|
|
|
my $data; |
|
|
my $line=1; |
|
|
|
|
|
while (<FEED>) { |
|
|
chomp; |
|
|
|
|
|
if (/^$/) { |
|
|
my $swishpath = $database."#".$i++; |
|
|
|
|
|
if (my $xml = data2xml($type_base,$data,$add_xml,$cfg,$database)) { |
|
|
$xml = $cp2utf->convert($xml); |
|
|
use bytes; # as opposed to chars |
|
|
print "Path-Name: $swishpath\n"; |
|
|
print "Content-Length: ".(length($xml)+1)."\n"; |
|
|
print "Document-Type: XML\n\n$xml\n"; |
|
|
} |
|
|
$line = 1; |
|
|
$data = {}; |
|
|
next; |
|
|
} |
|
|
|
|
|
$line = $1 if (s/^(\d+):\s*//); |
|
|
$data->{$line++} = $_; |
|
|
|
|
|
fakeprogress($i); |
|
|
|
|
|
} |
|
|
# close lookup |
|
|
untie %lhash if (%lhash); |
|
|
} |
|
104 |
} |
} |
105 |
|
|
|
# call this to commit index |
|
106 |
$index->close; |
$index->close; |
107 |
|
|
|
1; |
|
|
__END__ |
|
|
########################################################################## |
|
|
|
|
|
=head1 NAME |
|
|
|
|
|
all2xml.pl - read various file formats and dump XML for SWISH-E |
|
|
|
|
|
=head1 DESCRIPTION |
|
|
|
|
|
This command will read ISIS data file using OpenIsis perl module, MARC |
|
|
records using MARC module and optionally Micro$oft Excel files to |
|
|
create one XML file for usage with I<SWISH-E> indexer. Dispite it's name, |
|
|
this script B<isn't general xml generator> from isis files (isis allready |
|
|
has something like that). Output of this script is tailor-made for SWISH-E. |
|
|
|
|
|
=head1 BUGS |
|
|
|
|
|
Documentation is really lacking. However, in true Open Source spirit, source |
|
|
is best documentation. I even made considerable effort to comment parts |
|
|
which are not intuitively clear, so... |
|
|
|
|
|
=head1 AUTHOR |
|
|
|
|
|
Dobrica Pavlinusic <dpavlin@rot13.org> |
|
|
|
|
|
=head1 COPYRIGHT |
|
|
|
|
|
GNU Public License (GPL) v2 or later |
|
|
|
|
|
=head1 SEE ALSO |
|
|
|
|
|
SWISH-E web site at http://www.swish-e.org |
|
|
|
|
|
=cut |
|