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; |
|
|
|
|
|
$|=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; |
|
|
|
|
|
my %opts; |
|
|
|
|
|
# usage: |
|
|
# -d directory name |
|
|
# -m multiple directories |
|
|
# -q quiet |
|
|
# -s run swish |
|
|
|
|
|
getopts('d:m:qs', \%opts); |
|
4 |
|
|
5 |
my $path; # this is name of database |
all2all.pl - basic script for all WebPAC needs |
6 |
|
|
7 |
Text::Iconv->raise_error(0); # Conversion errors don't raise exceptions |
=cut |
8 |
|
|
9 |
# this is encoding of all files on disk, including import_xml/*.xml file and |
use strict; |
10 |
# filter/*.pm files! It will be used to store strings in perl internally! |
use Data::Dumper; |
11 |
my $codepage = 'ISO-8859-2'; |
use Carp; |
12 |
|
|
13 |
my $utf2cp = Text::Iconv->new('UTF-8',$codepage); |
use lib './lib'; |
14 |
# this function will convert data from XML files to local encoding |
use WebPAC; |
15 |
sub x { |
use WebPAC::jsFind; |
16 |
return $utf2cp->convert($_[0]); |
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 |
# decode isis/excel or other import codepage |
my $thes; |
|
my $import2cp; |
|
35 |
|
|
36 |
# outgoing xml must be in UTF-8 |
$|=1; |
|
my $cp2utf = Text::Iconv->new($codepage,'UTF-8'); |
|
37 |
|
|
38 |
# mapping between data type and tag which specify |
my $maxmfn = $webpac->open_isis( |
39 |
# format in XML file |
filename => shift @ARGV || '/data/hidra/THS/THS', |
40 |
my %type2tag = ( |
lookup => [ |
41 |
'isis' => 'isis', |
{ 'key' => 'd:v900', 'val' => 'v250^a' }, |
42 |
'excel' => 'column' |
# { '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 |
sub data2xml { |
$log->debug("isis file ",$webpac->{'isis_filename'}," opened"); |
54 |
|
|
55 |
use xmlify; |
$log->info("rows: $maxmfn"); |
56 |
|
|
57 |
my $type = shift @_; |
$webpac->open_import_xml(type => 'isis_hidra_ths'); |
|
my $row = shift @_; |
|
|
my $add_xml = shift @_; |
|
58 |
|
|
59 |
my $xml; |
$log->debug("data: ",Dumper($webpac->{'data'})); |
60 |
|
|
61 |
use parse_format; |
while (my $rec = $webpac->fetch_rec) { |
62 |
|
|
63 |
my $html = ""; # html formatted display output |
my @ds = $webpac->data_structure($rec); |
64 |
|
|
65 |
my %field_usage; # counter for usage of each field |
if (0 && $log->is_debug) { |
66 |
|
$log->debug("rec = ",Dumper($rec)); |
67 |
# sort subrouting using order="" attribute |
$log->debug("ds = ",Dumper(\@ds)); |
|
sub by_order { |
|
|
return 0 if (! $config->{indexer}->{$a}->{order}); |
|
|
return 0 if (! $config->{indexer}->{$b}->{order}); |
|
|
|
|
|
return $config->{indexer}->{$a}->{order} <=> |
|
|
$config->{indexer}->{$b}->{order} ; |
|
68 |
} |
} |
69 |
|
|
70 |
foreach my $field (sort by_order keys %{$config->{indexer}}) { |
next if (! @ds); |
|
|
|
|
$field=x($field); |
|
|
|
|
|
$field_usage{$field}++; |
|
71 |
|
|
72 |
my $swish_data = ""; |
my $filename = $webpac->{'current_filename'} || $log->logdie("no current_filename in webpac object"); |
|
my $display_data = ""; |
|
|
my $line_delimiter; |
|
73 |
|
|
74 |
my ($swish,$display); |
if ($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 |
my $tag = $type2tag{$type} || die "can't find which tag to use for type $type"; |
my $headline = $webpac->{'headline'}; |
|
foreach my $x (@{$config->{indexer}->{$field}->{$tag}}) { |
|
90 |
|
|
91 |
my $format = x($x->{content}); |
my $f = $filename; |
92 |
my $delimiter = x($x->{delimiter}) || ' '; |
$f =~ s!out/!!; |
93 |
|
|
94 |
my $repeat_off = 0; # repeatable offset |
# save into index |
95 |
|
foreach my $ds (@ds) { |
96 |
|
next if (! $ds->{'swish'}); |
97 |
|
|
98 |
my ($s,$d,$i) = (1,1,0); # swish, display default |
$index->insert( |
99 |
$s = 0 if (lc($x->{type}) eq "display"); |
index_name => $ds->{'tag'}, |
100 |
$d = 0 if (lc($x->{type}) eq "swish"); |
path => $f, |
101 |
($s,$d,$i) = (0,0,1) if (lc($x->{type}) eq "index"); |
headline => $headline, |
102 |
|
words => join(" ",@{$ds->{'swish'}}) |
103 |
|
); |
104 |
|
} |
105 |
|
|
106 |
# what will separate last line from this one? |
# save into sorted index (thesaurus) |
107 |
if ($display_data && $x->{append} && $x->{append} eq "1") { |
foreach my $ds (@ds) { |
108 |
$line_delimiter = ' '; |
next if (! $ds->{'index'}); |
|
} elsif ($display_data) { |
|
|
$line_delimiter = '<br/>'; |
|
|
} |
|
109 |
|
|
110 |
# init vars so that we go into while... |
$thes->{$ds->{'tag'}} ||= new WebPAC::Index; |
|
($swish,$display) = (1,1); |
|
111 |
|
|
112 |
if ($swish || $display) { |
foreach my $h (@{$ds->{'index'}}) { |
113 |
($swish,$display) = parse_format($type, $format,$row,$repeat_off++,$import2cp); |
$thes->{$ds->{'tag'}}->insert( |
114 |
# filter="name" ; filter this field through |
path => $f, |
115 |
# filter/[name].pm |
headline => $h, |
116 |
my $filter = $x->{filter}; |
); |
|
if ($filter) { |
|
|
require "filter/".$filter.".pm"; |
|
|
} |
|
|
# type="swish" ; field for swish |
|
|
if ($s && $swish) { |
|
|
if ($filter) { |
|
|
no strict 'refs'; |
|
|
$swish_data .= join(" ",&$filter($swish)); |
|
|
} else { |
|
|
$swish_data .= $swish; |
|
|
} |
|
|
} |
|
|
|
|
|
# type="display" ; field for display |
|
|
if ($d && $display) { |
|
|
if ($line_delimiter && $display_data) { |
|
|
$display_data .= $line_delimiter; |
|
|
undef $line_delimiter; |
|
|
} |
|
|
if ($filter) { |
|
|
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 = $display; |
|
|
if ($filter) { |
|
|
no strict 'refs'; |
|
|
foreach my $d (&$filter($index_data)) { |
|
|
$index->insert($field, $d, $path); |
|
|
} |
|
|
} else { |
|
|
$index->insert($field, $index_data, $path); |
|
|
} |
|
|
} |
|
|
} |
|
117 |
} |
} |
118 |
|
} |
119 |
|
|
120 |
|
# print Dumper(\@ds); |
121 |
|
|
122 |
if ($display_data) { |
} |
|
|
|
|
if ($field eq "headline") { |
|
|
$xml .= xmlify("headline", $display_data); |
|
|
} else { |
|
|
|
|
|
# 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) { |
|
|
$html .= x($field_name); |
|
|
} |
|
|
$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)); |
|
|
} |
|
123 |
|
|
124 |
|
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 |
|
my $file = "./out/bfilter/$t.txt"; |
133 |
|
$log->info("saving sorted index $t to '$file' [".scalar(@e)." elements]"); |
134 |
|
|
135 |
|
$webpac->output_file( |
136 |
|
file => $file, |
137 |
|
template => 'index.tt', |
138 |
|
data => \@e, |
139 |
|
index_name => $t, |
140 |
|
); |
141 |
|
} |
142 |
|
|
143 |
# dump formatted output in <html> |
if (0 && $log->is_debug) { |
144 |
if ($html) { |
$log->debug("lookup hash: ",Dumper($webpac->{'lookup'})); |
145 |
$xml .= xmlify("html",$html); |
$log->debug("data hash: ",Dumper($webpac->{'data'})); |
146 |
} |
foreach my $t (keys %{$thes}) { |
147 |
|
$log->debug("thesaurus $t hash: ",Dumper($thes->{$t})); |
|
if ($xml) { |
|
|
$xml .= $add_xml if ($add_xml); |
|
|
return "<xml>\n$xml</xml>\n"; |
|
|
} else { |
|
|
return; |
|
148 |
} |
} |
149 |
} |
} |
150 |
|
|
151 |
########################################################################## |
$log->info("lookup hash: ",Dumper($webpac->{'lookup'})); |
152 |
|
|
153 |
# read configuration for this script |
# |
154 |
my $cfg = new Config::IniFiles( -file => $config_file ); |
# tree dump implementation which shouldn't be here :-) |
155 |
|
# |
156 |
|
|
157 |
# read global.conf configuration |
my $file = 'out/browse.html'; |
158 |
my $cfg_global = new Config::IniFiles( -file => 'global.conf' ); |
open(HTML, "> $file") || $log->logdie("can't open '$file'"); |
159 |
|
|
160 |
# open index |
print HTML qq{ |
161 |
$index = new index_DBI( |
<html> |
162 |
$cfg_global->val('global', 'dbi_dbd'), |
<head> |
163 |
$cfg_global->val('global', 'dbi_dsn'), |
<title>Browse</title> |
164 |
$cfg_global->val('global', 'dbi_user'), |
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-2"/> |
165 |
$cfg_global->val('global', 'dbi_passwd') || '', |
</head> |
166 |
); |
<body> |
167 |
|
<ul> |
168 |
|
}; |
169 |
|
|
170 |
foreach my $database ($cfg->Sections) { |
foreach my $code (sort keys %{$webpac->{'lookup'}}) { |
171 |
|
|
172 |
my $type = lc($cfg -> val($database, 'type')) || die "$database doesn't have 'type' defined"; |
my $l = $webpac->{'lookup'} || $log->logconfess("can't find lookup"); |
|
my $add_xml = $cfg -> val($database, 'xml'); # optional |
|
173 |
|
|
174 |
print STDERR "reading ./import_xml/$type.xml\n"; |
if ($code =~ m/^root:/) { |
175 |
|
my $v900 = shift @{ $l->{$code} } || $log->logconfess("can't lookup '$code'"); |
176 |
|
$code =~ s/^root:// || die; |
177 |
|
|
178 |
$config=XMLin("./import_xml/$type.xml", forcearray => [ $type2tag{$type} ], forcecontent => 1); |
my $term = shift @{ $l->{"d:${v900}"} } || die; |
179 |
|
my $mfn = shift @{ $l->{"900_mfn:${v900}"} } || die; |
180 |
|
|
181 |
# output current progress indicator |
$log->debug("$code -> $v900 : $term [$mfn]"); |
|
my $last_p = 0; |
|
|
sub progress { |
|
|
#return if (! $opts{q}); # FIXME |
|
|
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 = $p; |
|
|
} |
|
|
} |
|
182 |
|
|
183 |
# now read database |
print HTML qq{ <li><a href="thes/$mfn.html">$term</a></li>\n} if (-e "out/thes/$mfn.html"); |
|
print STDERR "using: $type...\n"; |
|
184 |
|
|
185 |
if ($type eq "isis") { |
unless ($l->{"a:${code}::"}) { |
186 |
my $isis_db = $cfg -> val($database, 'isis_db') || die "$database doesn't have 'isis_db' defined!"; |
$log->warn("can't find 'a:${code}::'"); |
187 |
|
next; |
188 |
|
} |
189 |
|
|
190 |
$import2cp = Text::Iconv->new($config->{isis_codepage},$codepage); |
print HTML qq{ <ul>\n}; |
|
my $db = OpenIsis::open( $isis_db ); |
|
191 |
|
|
192 |
my $max_rowid = OpenIsis::maxRowid( $db ); |
foreach my $l2_v900 (@{ $l->{"a:${code}::"} }) { |
193 |
|
|
194 |
|
my $l2_code = shift @{ $l->{"code:${l2_v900}"} } || die; |
195 |
|
my $l2_term = shift @{ $l->{"d:$l2_v900"} } || die; |
196 |
|
my $l2_mfn = shift @{ $l->{"900_mfn:${l2_v900}"} } || die; |
197 |
|
|
198 |
print STDERR "Reading database: $isis_db [$max_rowid rows]\n"; |
$log->debug("$l2_code -> $l2_v900 : $l2_term [$l2_mfn]"); |
199 |
|
|
200 |
my $path = $database; |
print HTML qq{ <li><a href="thes/$l2_mfn.html">$l2_term</a></li>\n} if (-e "out/thes/$l2_mfn.html"); |
201 |
|
|
202 |
|
unless ($l->{"a:${l2_code}:"}) { |
203 |
|
$log->warn("can't find 'a:${l2_code}:'"); |
204 |
|
next; |
205 |
|
} |
206 |
|
|
207 |
for (my $row_id = 1; $row_id <= $max_rowid; $row_id++ ) { |
print HTML qq{ <ul>\n}; |
|
my $row = OpenIsis::read( $db, $row_id ); |
|
|
if ($row && $row->{mfn}) { |
|
|
|
|
|
progress($row->{mfn}, $max_rowid); |
|
208 |
|
|
209 |
my $swishpath = $path."#".int($row->{mfn}); |
foreach my $l3_v900 (@{ $l->{"a:${l2_code}:"} }) { |
210 |
|
|
211 |
if (my $xml = data2xml($type,$row,$add_xml)) { |
my $l3_term = shift @{ $l->{"d:$l3_v900"} } || die; |
212 |
$xml = $cp2utf->convert($xml); |
my $l3_mfn = shift @{ $l->{"900_mfn:${l3_v900}"} } || die; |
|
use bytes; # as opposed to chars |
|
|
print "Path-Name: $swishpath\n"; |
|
|
print "Content-Length: ".(length($xml)+1)."\n"; |
|
|
print "Document-Type: XML\n\n$xml\n"; |
|
|
} |
|
|
} |
|
|
} |
|
|
print STDERR "\n"; |
|
213 |
|
|
214 |
} elsif ($type eq "excel") { |
$log->debug(" -> $l3_v900 : $l3_term [$l3_mfn]"); |
|
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++; |
|
|
} |
|
215 |
|
|
216 |
my $oWorksheet = $oBook->{Worksheet}[$sheet_nr]; |
print HTML qq{ <li><a href="thes/$l3_mfn.html">$l3_term</a></li>\n} if (-e "out/thes/$l3_mfn.html"); |
|
|
|
|
print STDERR "using sheet: ",$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)} = $cell->Value; |
|
|
} |
|
217 |
} |
} |
218 |
|
|
219 |
progress($iR, $end_row); |
print HTML qq{ </ul>\n}; |
|
|
|
|
# print "row[$iR/$end_row] "; |
|
|
# foreach (keys %{$row}) { |
|
|
# print "$_: ",$row->{$_},"\t"; |
|
|
# } |
|
|
# print "\n"; |
|
|
|
|
|
my $swishpath = $database."#".$iR; |
|
|
|
|
|
next if (! $row); |
|
|
|
|
|
if (my $xml = data2xml($type,$row,$add_xml)) { |
|
|
$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"; |
|
|
} |
|
220 |
} |
} |
221 |
|
print HTML qq{ </ul>\n}; |
222 |
} |
} |
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'})); |