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; |
|
|
|
|
|
$|=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; # there is no other, right now ;-) |
|
|
|
|
|
my $index = new index_DBI(); # open index |
|
4 |
|
|
5 |
my %opts; |
all2all.pl - basic script for all WebPAC needs |
6 |
|
|
7 |
# usage: |
=cut |
|
# -d directory name |
|
|
# -m multiple directories |
|
|
# -q quiet |
|
|
# -s run swish |
|
8 |
|
|
9 |
getopts('d:m:qs', \%opts); |
use strict; |
10 |
|
use Data::Dumper; |
11 |
|
use Carp; |
12 |
|
|
13 |
my $db_dir; |
use lib './lib'; |
14 |
|
use WebPAC; |
15 |
|
use WebPAC::jsFind; |
16 |
|
use WebPAC::Index; |
17 |
|
|
18 |
|
my $webpac = new WebPAC( |
19 |
|
code_page => 'ISO-8859-2', |
20 |
|
# limit_mfn => 100, |
21 |
|
# debug => 1, |
22 |
|
low_mem => 1, |
23 |
|
) || die; |
24 |
|
|
25 |
|
my $log = $webpac->_get_logger() || die "can't get logger"; |
26 |
|
|
27 |
|
$log->debug("creating WebPAC::jsFind object"); |
28 |
|
|
29 |
|
my $index = new WebPAC::jsFind( |
30 |
|
index_path => './out/index', |
31 |
|
keys => 10, |
32 |
|
) || die; |
33 |
|
|
34 |
#die "usage: $0 -d [database_dir] -m [database1,database2] " if (! %opts); |
my $thes; |
35 |
|
|
36 |
#print Dumper($config->{indexer}); |
$|=1; |
|
#print "-" x 70,"\n"; |
|
37 |
|
|
38 |
Text::Iconv->raise_error(1); # Conversion errors raise exceptions |
my $maxmfn = $webpac->open_isis( |
39 |
|
filename => shift @ARGV || '/data/hidra/THS/THS', |
40 |
|
lookup => [ |
41 |
|
{ 'key' => 'd:v900', 'val' => 'v250^a' }, |
42 |
|
# { 'eval' => '"v901^a" eq "Područje"', 'key' => 'pa:v561^4:v562^4:v461^1', 'val' => 'v900' }, |
43 |
|
# { 'eval '=> '"v901^a" eq "Mikrotezaurus"', 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' }, |
44 |
|
# { 'eval' => '"v901^a" eq "Deskriptor"', 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' }, |
45 |
|
{ 'key' => 'a:v561^4:v562^4:v461^1', 'val' => 'v900' }, |
46 |
|
{ 'key' => '900_mfn:v900', 'val' => 'v000' }, |
47 |
|
# tree structure |
48 |
|
{ 'eval' => 'length("v251") == 2', 'key' => 'root:v251', 'val' => 'v900' }, |
49 |
|
{ 'eval' => '"v251"', 'key' => 'code:v900', 'val' => 'v561^4:v251' }, |
50 |
|
], |
51 |
|
); |
52 |
|
|
53 |
|
$log->debug("isis file ",$webpac->{'isis_filename'}," opened"); |
54 |
|
|
55 |
|
$log->info("rows: $maxmfn"); |
56 |
|
|
57 |
|
$webpac->open_import_xml(type => 'isis_hidra_ths'); |
58 |
|
|
59 |
|
$log->debug("data: ",Dumper($webpac->{'data'})); |
60 |
|
|
61 |
|
while (my $rec = $webpac->fetch_rec) { |
62 |
|
|
63 |
|
my @ds = $webpac->data_structure($rec); |
64 |
|
|
65 |
|
if (0 && $log->is_debug) { |
66 |
|
$log->debug("rec = ",Dumper($rec)); |
67 |
|
$log->debug("ds = ",Dumper(\@ds)); |
68 |
|
} |
69 |
|
|
70 |
my $isis_codepage; |
next if (! @ds); |
|
my $index_codepage; |
|
|
my $cludge_codepage = Text::Iconv->new('UTF8','ISO8859-1'); |
|
|
my $xml_codepage; |
|
71 |
|
|
72 |
my $XML_CHARSET = 'UTF8'; |
my $filename = $webpac->{'current_filename'} || $log->logdie("no current_filename in webpac object"); |
73 |
|
|
74 |
|
if ($filename) { |
75 |
|
$webpac->output_file( |
76 |
|
file => $filename, |
77 |
|
template => 'html.tt', |
78 |
|
data => \@ds, |
79 |
|
headline => $webpac->{'headline'}, |
80 |
|
); |
81 |
|
} else { |
82 |
|
print $webpac->output( |
83 |
|
template => 'text.tt', |
84 |
|
data => \@ds, |
85 |
|
headline => $webpac->{'headline'}, |
86 |
|
); |
87 |
|
} |
88 |
|
|
89 |
sub isis2xml { |
my $headline = $webpac->{'headline'}; |
90 |
|
|
91 |
use xmlify; |
my $f = $filename; |
92 |
|
$f =~ s!out/!!; |
93 |
|
|
94 |
my $row = shift @_; |
# save into index |
95 |
my $add_xml = shift @_; |
foreach my $ds (@ds) { |
96 |
|
next if (! $ds->{'swish'}); |
97 |
|
|
98 |
|
$index->insert( |
99 |
|
index_name => $ds->{'tag'}, |
100 |
|
path => $f, |
101 |
|
headline => $headline, |
102 |
|
words => join(" ",@{$ds->{'swish'}}) |
103 |
|
); |
104 |
|
} |
105 |
|
|
106 |
my $xml; |
# save into sorted index (thesaurus) |
107 |
|
foreach my $ds (@ds) { |
108 |
|
next if (! $ds->{'index'}); |
109 |
|
|
110 |
|
$thes->{$ds->{'tag'}} ||= new WebPAC::Index; |
111 |
|
|
112 |
|
foreach my $h (@{$ds->{'index'}}) { |
113 |
|
$thes->{$ds->{'tag'}}->insert( |
114 |
|
path => $f, |
115 |
|
headline => $h, |
116 |
|
); |
117 |
|
} |
118 |
|
} |
119 |
|
|
120 |
use parse_format; |
# print Dumper(\@ds); |
121 |
|
|
122 |
my $html = ""; # html formatted display output |
} |
123 |
|
|
124 |
my %field_usage; # counter for usage of each field |
foreach my $t (keys %{$thes}) { |
125 |
|
|
126 |
|
my @e = $thes->{$t}->elements; |
127 |
|
if (! @e) { |
128 |
|
$log->logwarn("no elements in sorted index $t?"); |
129 |
|
next; |
130 |
|
} |
131 |
|
|
132 |
foreach my $field (keys %{$config->{indexer}}) { |
my $file = "./out/bfilter/$t.txt"; |
133 |
|
$log->info("saving sorted index $t to '$file' [".scalar(@e)." elements]"); |
134 |
|
|
135 |
$field_usage{$field}++; |
$webpac->output_file( |
136 |
|
file => $file, |
137 |
|
template => 'index.tt', |
138 |
|
data => \@e, |
139 |
|
index_name => $t, |
140 |
|
); |
141 |
|
} |
142 |
|
|
143 |
my $swish_data = ""; |
if (0 && $log->is_debug) { |
144 |
my $display_data = ""; |
$log->debug("lookup hash: ",Dumper($webpac->{'lookup'})); |
145 |
my $line_delimiter = ""; |
$log->debug("data hash: ",Dumper($webpac->{'data'})); |
146 |
|
foreach my $t (keys %{$thes}) { |
147 |
|
$log->debug("thesaurus $t hash: ",Dumper($thes->{$t})); |
148 |
|
} |
149 |
|
} |
150 |
|
|
151 |
foreach my $x (@{$config->{indexer}->{$field}->{isis}}) { |
$log->info("lookup hash: ",Dumper($webpac->{'lookup'})); |
152 |
|
|
153 |
my $format = $x->{content}; |
# |
154 |
my $delimiter = $x->{delimiter} || ' '; |
# tree dump implementation which shouldn't be here :-) |
155 |
|
# |
156 |
|
|
157 |
# FIX: this is ugly, UGLY, cludge string is returned |
my $file = 'out/browse.html'; |
158 |
# in UTF8 encoding , but as if source charset |
open(HTML, "> $file") || $log->logdie("can't open '$file'"); |
|
# is ISO8859-1 and not some other. This break other |
|
|
# character encodings, so we convert it first |
|
|
# back to ISO8859-1 |
|
|
$format = $xml_codepage->convert($format); |
|
|
$delimiter = $xml_codepage->convert($delimiter) if ($delimiter); |
|
159 |
|
|
160 |
my $isis_i = 0; # isis repeatable offset |
print HTML qq{ |
161 |
|
<html> |
162 |
|
<head> |
163 |
|
<title>Browse</title> |
164 |
|
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-2"/> |
165 |
|
</head> |
166 |
|
<body> |
167 |
|
<ul> |
168 |
|
}; |
169 |
|
|
170 |
my ($s,$d,$i) = (1,1,0); # swish, display default |
foreach my $code (sort keys %{$webpac->{'lookup'}}) { |
|
$s = 0 if (lc($x->{type}) eq "display"); |
|
|
$d = 0 if (lc($x->{type}) eq "swish"); |
|
|
($s,$d,$i) = (0,0,1) if (lc($x->{type}) eq "index"); |
|
|
#print STDERR "## s: $s d: $d i: $i ## $format ##\n"; |
|
171 |
|
|
172 |
$display_data .= $line_delimiter if ($display_data && $display_data !~ /$line_delimiter$/); |
my $l = $webpac->{'lookup'} || $log->logconfess("can't find lookup"); |
173 |
|
|
174 |
my ($swish,$display) = (1,1); |
if ($code =~ m/^root:/) { |
175 |
|
my $v900 = shift @{ $l->{$code} } || $log->logconfess("can't lookup '$code'"); |
176 |
|
$code =~ s/^root:// || die; |
177 |
|
|
178 |
while ($swish || $display) { |
my $term = shift @{ $l->{"d:${v900}"} } || die; |
179 |
($swish,$display) = parse_format($format,$row,$isis_i++); |
my $mfn = shift @{ $l->{"900_mfn:${v900}"} } || die; |
|
#print STDERR "s: $swish\nd: $display\n" if ($swish); |
|
180 |
|
|
181 |
#print STDERR "swish: $swish<-- display: $display<--\n"; |
$log->debug("$code -> $v900 : $term [$mfn]"); |
182 |
|
|
183 |
# filter="name" ; filter this field through |
print HTML qq{ <li><a href="thes/$mfn.html">$term</a></li>\n} if (-e "out/thes/$mfn.html"); |
|
# filter/[name].pm |
|
|
my $filter = $x->{filter}; |
|
|
if ($filter) { |
|
|
require "filter/".$filter.".pm"; |
|
|
} |
|
|
# type="swish" ; field for swish |
|
|
if ($s && $swish) { |
|
|
if ($filter) { |
|
|
print STDERR "using filter '$filter'\n"; |
|
|
no strict 'refs'; |
|
|
$swish_data .= join(" ",&$filter($swish)); |
|
|
} else { |
|
|
$swish_data .= $swish; |
|
|
} |
|
|
} |
|
184 |
|
|
185 |
# type="display" ; field for display |
unless ($l->{"a:${code}::"}) { |
186 |
if ($d && $display) { |
$log->warn("can't find 'a:${code}::'"); |
187 |
if ($filter) { |
next; |
|
no strict 'refs'; |
|
|
$display_data .= join($delimiter,&$filter($display)); |
|
|
} else { |
|
|
if ($display_data) { |
|
|
$display_data .= $delimiter.$display; |
|
|
} else { |
|
|
$display_data .= $display; |
|
|
} |
|
|
} |
|
|
} |
|
|
|
|
|
# type="index" ; insert into index |
|
|
if ($i && $display) { |
|
|
my $index_data = $index_codepage->convert($display) || $display; |
|
|
if ($filter) { |
|
|
no strict 'refs'; |
|
|
foreach my $d (&$filter($index_data)) { |
|
|
$index->insert($field, $d, $db_dir); |
|
|
} |
|
|
} else { |
|
|
$index->insert($field, $index_data, $db_dir); |
|
|
} |
|
|
} |
|
|
} |
|
|
if ($x->{append}) { |
|
|
$line_delimiter = ' '; |
|
|
} else { |
|
|
$line_delimiter = '<br/>'; |
|
|
} |
|
188 |
} |
} |
189 |
|
|
190 |
|
print HTML qq{ <ul>\n}; |
191 |
|
|
192 |
#print STDERR "s_d: $swish_data\nd_d: $display_data\n" if ($swish_data); |
foreach my $l2_v900 (@{ $l->{"a:${code}::"} }) { |
193 |
if ($display_data) { |
|
194 |
|
my $l2_code = shift @{ $l->{"code:${l2_v900}"} } || die; |
195 |
# remove last <br> |
my $l2_term = shift @{ $l->{"d:$l2_v900"} } || die; |
196 |
$display_data =~ s/$line_delimiter$//; |
my $l2_mfn = shift @{ $l->{"900_mfn:${l2_v900}"} } || die; |
197 |
|
|
198 |
$display_data = $isis_codepage->convert($display_data) || die "Can't convert '$display_data' !"; |
$log->debug("$l2_code -> $l2_v900 : $l2_term [$l2_mfn]"); |
199 |
# FIX: this is removed and replaced by html tag. |
|
200 |
#$xml .= xmlify($field."_display", $display_data); |
print HTML qq{ <li><a href="thes/$l2_mfn.html">$l2_term</a></li>\n} if (-e "out/thes/$l2_mfn.html"); |
201 |
|
|
202 |
if ($field eq "headline") { |
unless ($l->{"a:${l2_code}:"}) { |
203 |
$xml .= xmlify("headline", $display_data); |
$log->warn("can't find 'a:${l2_code}:'"); |
204 |
} else { |
next; |
|
|
|
|
# find field name (signular, plural) |
|
|
my $field_name = ""; |
|
|
if ($config->{indexer}->{$field}->{name_singular} && $field_usage{$field} == 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) { |
|
|
$html .= $xml_codepage->convert($field_name); |
|
|
# $html .= "-->".$field_name."<--"; |
|
|
} |
|
|
$html .= $display_data."###\n"; |
|
205 |
} |
} |
|
} |
|
|
if ($swish_data) { |
|
|
my $i = Text::Iconv->new($config->{isis_codepage},'ISO8859-2'); |
|
|
# remove extra spaces |
|
|
$swish_data =~ s/ +/ /g; |
|
|
$swish_data =~ s/ +$//g; |
|
|
|
|
|
$swish_data = $i->convert($swish_data); |
|
|
$xml .= xmlify($field."_swish",unac_string('ISO8859-2',$swish_data)); |
|
|
#$swish_data = $isis_codepage->convert($swish_data)."##" || $swish_data; |
|
|
#$xml .= xmlify($field."_swish",unac_string($config->{isis_codepage},$swish_data)); |
|
|
} |
|
|
|
|
|
|
|
|
} |
|
206 |
|
|
207 |
# dump formatted output in <html> |
print HTML qq{ <ul>\n}; |
|
if ($html) { |
|
|
$xml .= xmlify("html",$html); |
|
|
} |
|
|
|
|
|
if ($xml) { |
|
|
#print STDERR "x: $xml\n"; |
|
|
$xml .= $add_xml if ($add_xml); |
|
|
return "<xml>\n$xml</xml>\n"; |
|
|
} else { |
|
|
return; |
|
|
} |
|
|
} |
|
208 |
|
|
209 |
########################################################################## |
foreach my $l3_v900 (@{ $l->{"a:${l2_code}:"} }) { |
210 |
|
|
211 |
my $cfg = new Config::IniFiles( -file => $config_file ); |
my $l3_term = shift @{ $l->{"d:$l3_v900"} } || die; |
212 |
|
my $l3_mfn = shift @{ $l->{"900_mfn:${l3_v900}"} } || die; |
213 |
|
|
214 |
foreach my $database ($cfg->Sections) { |
$log->debug(" -> $l3_v900 : $l3_term [$l3_mfn]"); |
215 |
|
|
216 |
my $isis_db = $cfg -> val($database, 'isis_db'); |
print HTML qq{ <li><a href="thes/$l3_mfn.html">$l3_term</a></li>\n} if (-e "out/thes/$l3_mfn.html"); |
|
my $type = $cfg -> val($database, 'type'); |
|
|
my $add_xml = $cfg -> val($database, 'xml'); |
|
|
|
|
|
# read configuration for this type |
|
|
$config=XMLin("./import_xml/$type.xml", forcearray => [ 'isis' ], forcecontent => 1); |
|
|
$isis_codepage = Text::Iconv->new($config->{isis_codepage},$XML_CHARSET); |
|
|
$index_codepage = Text::Iconv->new($config->{isis_codepage},$config->{index_codepage}); |
|
|
$xml_codepage = Text::Iconv->new($cfg->val($database,'xml_codepage'),$XML_CHARSET); |
|
|
|
|
|
my $db = OpenIsis::open( $isis_db ); |
|
|
if (0) { |
|
|
# # FIX |
|
|
# if (! $db ) { |
|
|
print STDERR "WARNING: can't open '$isis_db'\n"; |
|
|
next ; |
|
|
} |
|
|
|
|
|
my $max_rowid = OpenIsis::maxRowid( $db ); |
|
|
|
|
|
print STDERR "Reading database: $isis_db [$max_rowid rows]\n"; |
|
|
|
|
|
my $path = $database; # was $isis_db |
|
|
|
|
|
my $last_p = 0; |
|
|
|
|
|
# { my $row_id = 4514; |
|
|
# FIX |
|
|
for (my $row_id = 1; $row_id <= $max_rowid; $row_id++ ) { |
|
|
my $row = OpenIsis::read( $db, $row_id ); |
|
|
if ($row && $row->{mfn}) { |
|
|
#print STDERR "mfn: ",$row->{mfn},"\n"; |
|
|
# output current process indicator |
|
|
my $p = int($row->{mfn} * 100 / $max_rowid); |
|
|
if ($p != $last_p) { |
|
|
printf STDERR ("%5d / %5d [%-51s] %-2d %% \r",$row->{mfn},$max_rowid,"=" x ($p/2).">", $p ) if (! $opts{q}); |
|
|
$last_p = $p; |
|
217 |
} |
} |
218 |
|
|
219 |
if (my $xml = isis2xml($row,$add_xml)) { |
print HTML qq{ </ul>\n}; |
|
#print STDERR "--ret-->$xml\n"; |
|
|
print "Path-Name: $path#".int($row->{mfn})."\n"; |
|
|
print "Content-Length: ".(length($xml)+1)."\n"; |
|
|
print "Document-Type: XML\n\n$xml\n"; |
|
|
} |
|
220 |
} |
} |
221 |
|
print HTML qq{ </ul>\n}; |
222 |
} |
} |
|
print STDERR "\n"; |
|
223 |
} |
} |
224 |
|
|
225 |
# call this to commit index |
print HTML qq{ |
226 |
$index->close; |
</ul> |
227 |
|
</body> |
228 |
1; |
</html> |
229 |
__END__ |
}; |
|
########################################################################## |
|
|
|
|
|
=head1 NAME |
|
|
|
|
|
isis2xml.pl - read isis file and dump XML |
|
|
|
|
|
=head1 DESCRIPTION |
|
|
|
|
|
This command will read ISIS data file using OpenIsis perl module and |
|
|
create 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. |
|
230 |
|
|
231 |
=head1 AUTHOR |
$index->close; |
|
|
|
|
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 |
|
232 |
|
|
233 |
=cut |
$log->info("elapsed time: ",$webpac->fmt_time(time() - $webpac->{'start_t'})); |