| 1 |
#!/usr/bin/perl -w |
| 2 |
|
| 3 |
# indexer, Dobrica Pavlinusic <dpavlin@rot13.org> 2002-06-19 |
| 4 |
# options: -q quiet |
| 5 |
# -d debug |
| 6 |
# -v verbose |
| 7 |
# -l limit regex |
| 8 |
|
| 9 |
# This indexer output xml data which is used to index content with |
| 10 |
# swish-e 2.2, http://www.swish-e.org/ |
| 11 |
# |
| 12 |
# xml is output is on STDOUT and informational oputput (for humas) is |
| 13 |
# on STDERR |
| 14 |
# |
| 15 |
|
| 16 |
use strict; |
| 17 |
use Getopt::Std; |
| 18 |
require Unicode::Map8; |
| 19 |
use GDBM_File; |
| 20 |
|
| 21 |
my $sadrzaj=0; |
| 22 |
my $nr=0; |
| 23 |
my $naslov=""; |
| 24 |
|
| 25 |
my $br; ## broj NN |
| 26 |
my $god; ## godina NN |
| 27 |
my $aname; ## ancor name na originalnim stranicama |
| 28 |
|
| 29 |
my $nn_dir="../"; # dir u kojem su wget-ani fileovi |
| 30 |
#my $path_fmt="http://www.nn.hr/CijeliBrojS.asp?god=%d&br=%s&mid=%s#%d"; |
| 31 |
|
| 32 |
# configure gdbm files here |
| 33 |
my $gdbm_brzakona="$nn_dir/swish/brzakona.gdbm"; |
| 34 |
my $gdbm_file2title="$nn_dir/swish/file2title.gdbm"; |
| 35 |
|
| 36 |
# where to drop full text URLs |
| 37 |
my $full_url_list="$nn_dir/sluzbeno/in.url"; |
| 38 |
|
| 39 |
# URL to original site |
| 40 |
my $full_url="http://www.nn.hr/clanci/sluzbeno/"; |
| 41 |
my $full_filename_fmt="%04d/%04s.htm"; |
| 42 |
my $path_fmt = $full_filename_fmt; |
| 43 |
|
| 44 |
# regex for filenames |
| 45 |
my $broj_html_re = qr/^(CijeliBrojS|pregled.asp)/; |
| 46 |
|
| 47 |
my %opts; |
| 48 |
getopts("vqdl:", \%opts); |
| 49 |
|
| 50 |
my $brojeva=0; |
| 51 |
my $zakona=0; |
| 52 |
my $zak_u_broju; |
| 53 |
|
| 54 |
|
| 55 |
my $l2_map = Unicode::Map8->new("ISO-8859-2") || die; |
| 56 |
|
| 57 |
my %br_zakona; |
| 58 |
tie %br_zakona, 'GDBM_File', $gdbm_brzakona.".temp", &GDBM_NEWDB, 0644; |
| 59 |
my %file2title; |
| 60 |
tie %file2title, 'GDBM_File', $gdbm_file2title.".temp", &GDBM_NEWDB, 0644; |
| 61 |
|
| 62 |
#-------------------------------------------------------------------- |
| 63 |
|
| 64 |
sub save_br_zak { |
| 65 |
my $god = shift || return; |
| 66 |
my $br = shift || return; |
| 67 |
my $zak_u_broju = shift || return; |
| 68 |
print STDERR "[$god/$br: $zak_u_broju zakona]\n" if (! $opts{q}); |
| 69 |
if (! $br_zakona{sprintf("%04d",$god)}) { |
| 70 |
$br_zakona{sprintf("%04d",$god)} = $zak_u_broju; |
| 71 |
} else { |
| 72 |
$br_zakona{sprintf("%04d",$god)} += $zak_u_broju; |
| 73 |
} |
| 74 |
} |
| 75 |
|
| 76 |
#-------------------------------------------------------------------- |
| 77 |
|
| 78 |
sub dump_to_swish { |
| 79 |
my $xml = shift @_; |
| 80 |
my ($god,$br,$nr,$aname) = @_; |
| 81 |
|
| 82 |
use utf8; |
| 83 |
|
| 84 |
# print "Path-Name: ".sprintf($path_fmt,$god,$br,$nr,$aname)."\n". |
| 85 |
print "Path-Name: ".sprintf($path_fmt,$god,$nr)."\n". |
| 86 |
"Content-Length: ".length($xml)."\n". |
| 87 |
"Document-Type: XML\n". |
| 88 |
"\n$xml"; |
| 89 |
} |
| 90 |
|
| 91 |
#-------------------------------------------------------------------- |
| 92 |
|
| 93 |
open(URL,"> $full_url_list") || warn "can't open URL list file '$full_url_list': $!"; |
| 94 |
|
| 95 |
opendir(DIR,$nn_dir) || warn "opendir: $!"; |
| 96 |
my @files; |
| 97 |
if ($opts{l}) { |
| 98 |
# add limit regex |
| 99 |
@files = grep { $_ =~ $broj_html_re && /$opts{l}/ && -f "$nn_dir/$_" } readdir(DIR); |
| 100 |
print STDERR "Using limit regex which is '$opts{l}'\n"; |
| 101 |
} else { |
| 102 |
@files = grep { $_ =~ $broj_html_re && -f "$nn_dir/$_" } readdir(DIR); |
| 103 |
} |
| 104 |
closedir(DIR); |
| 105 |
|
| 106 |
foreach my $file (sort @files) { |
| 107 |
open(IN,"$nn_dir/$file") || die "can't open '$nn_dir/$file': $!"; |
| 108 |
|
| 109 |
if ($file=~m/(?:god|godina)=(\d+)\&(?:br|broj)=(\d+)/) { |
| 110 |
if ($god && $br && $zak_u_broju) { |
| 111 |
save_br_zak($god,$br,$zak_u_broju); |
| 112 |
} |
| 113 |
($br,$god) = ($2,$1); |
| 114 |
$brojeva++; |
| 115 |
$zak_u_broju = 0; |
| 116 |
print STDERR "$file $god $br\n" if (! $opts{q}); |
| 117 |
} |
| 118 |
|
| 119 |
my $insert_in_swish = 0; |
| 120 |
|
| 121 |
while(<IN>) { |
| 122 |
chomp; |
| 123 |
s/\015//g; # kill cr |
| 124 |
tr/ðèæÐÈÆ/¹ð¾èæ©Ð®ÈÆ/; # 1250 -> iso8859-2 |
| 125 |
|
| 126 |
# |
| 127 |
# parse old pages (CijeliBrojS.asp) with <div class=sadrzaj> |
| 128 |
# |
| 129 |
|
| 130 |
if (m,<div class=sadrzaj>,) { |
| 131 |
$sadrzaj++; |
| 132 |
next; |
| 133 |
} |
| 134 |
|
| 135 |
if ($sadrzaj) { |
| 136 |
if (s/<a href="#([^"]+)">\s*(\S+)\.\s*<[^>]+>//i) { |
| 137 |
($aname,$nr) = ($1,$2); |
| 138 |
} elsif (s/<a href="Javascript:Mojdok\((\d+),(\d+),'*(\w+)'*,(\d+)\)[^>]*>//i) { |
| 139 |
($nr,$aname) = ($3,$4); |
| 140 |
die "conflict in godina: $1 != $god" if ($god != $1); |
| 141 |
die "conflict in broj: $2 != $br" if ($br != $2); |
| 142 |
} else { |
| 143 |
die "can't find nr in line: $_ [$file]"; |
| 144 |
} |
| 145 |
$naslov.=$_; |
| 146 |
$naslov=~s/^\s*$nr\.*\s*//g; |
| 147 |
$sadrzaj = 0; |
| 148 |
$insert_in_swish = 1; |
| 149 |
} |
| 150 |
|
| 151 |
# |
| 152 |
# new pregled.asp format |
| 153 |
# |
| 154 |
|
| 155 |
if (m#<A TARGET="ispis" HREF="/clanci/sluzbeno/(\d+)/(\d+).htm">\s*(\d+)\.*\s+([^<]+)</A>#) { |
| 156 |
($god, $nr, $aname, $naslov) = ($1,$2,$3,$4); |
| 157 |
$naslov=~s/^\s*$nr\.*\s*//g; |
| 158 |
$insert_in_swish = 1; |
| 159 |
} |
| 160 |
|
| 161 |
if ($insert_in_swish) { |
| 162 |
$insert_in_swish = 0; |
| 163 |
$naslov=~s/\s+/ /g; |
| 164 |
$naslov=~s/<[^>]+>//g; |
| 165 |
$naslov=~s/^\s+//g; |
| 166 |
$naslov=~s/\s+$//g; |
| 167 |
print STDERR "$god $br $nr: $naslov\n" if ($opts{v}); |
| 168 |
my $naslov_czs = lc($naslov); |
| 169 |
$naslov_czs =~ tr/¹©ðÐèÈæƾ®/sSdDcCcCzZ/; |
| 170 |
$naslov_czs =~ tr/a-zA-Z0-9/ /cs; # non a-z -> space |
| 171 |
# $naslov_czs = $hr->minimal(split(/ /,$naslov_czs)); |
| 172 |
my $xml="<nn>\n<br>$br</br>\n<god>$god</god>\n<nr>$nr</nr>\n<aname>$aname</aname>\n"; |
| 173 |
my $naslov_utf=$l2_map->tou($naslov)->utf8; |
| 174 |
|
| 175 |
# Escape <, >, & and ", and to produce valid XML |
| 176 |
my %escape = ('<'=>'<', '>'=>'>', '&'=>'&', '"'=>'"'); |
| 177 |
my $escape_re = join '|' => keys %escape; |
| 178 |
$naslov_utf =~ s/($escape_re)/$escape{$1}/g; |
| 179 |
|
| 180 |
$xml.="<naslov>$naslov_utf</naslov>\n"; |
| 181 |
$xml.="<naslov_czs>$naslov_czs</naslov_czs>\n</nn>\n\n"; |
| 182 |
dump_to_swish($xml,$god,$br,$nr,$aname); |
| 183 |
|
| 184 |
my $file = sprintf($full_filename_fmt,$god,$nr); |
| 185 |
if (! -e "$nn_dir/sluzbeno/$file") { |
| 186 |
print URL $full_url.$file."\n"; |
| 187 |
} |
| 188 |
$file2title{$file}="$god $br $nr $aname $naslov"; |
| 189 |
|
| 190 |
$naslov=""; |
| 191 |
$nr=0; |
| 192 |
$zakona++; |
| 193 |
$zak_u_broju++; |
| 194 |
} |
| 195 |
|
| 196 |
|
| 197 |
} |
| 198 |
|
| 199 |
close(IN); |
| 200 |
} |
| 201 |
|
| 202 |
save_br_zak($god,$br,$zak_u_broju); |
| 203 |
print STDERR "Ukupno $brojeva brojeva NN, sa $zakona zakona...\n" if (! $opts{q}); |
| 204 |
|
| 205 |
untie %br_zakona; |
| 206 |
|
| 207 |
# rename temp gdbm files |
| 208 |
rename $gdbm_brzakona.".temp",$gdbm_brzakona || die "can't rename $gdbm_brzakona: $!"; |
| 209 |
rename $gdbm_file2title.".temp",$gdbm_file2title || die "can't rename $gdbm_file2title: $!"; |
Parent Directory
| 
Revision Log