#!/usr/bin/perl -w
use strict;
use Biblio::Isis 0.23;
use Getopt::Std;
use Data::Dumper;
use XML::Simple;
use Text::Iconv;
use Config::IniFiles;
use Encode;
use GDBM_File;
use Fcntl; # for O_RDWR
#use TDB_File;
use Carp;
$|=1;
my $config_file = $0;
$config_file =~ s/\.pl$/.conf/;
$config_file = $ARGV[0] if ($ARGV[0] && -f $ARGV[0]);
die "FATAL: can't find configuration file '$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
use index_DBI_filter; # filter support for indexes
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
Text::Iconv->raise_error(0); # Conversion errors don't raise exceptions
# this is encoding of all files on disk, including import_xml/*.xml file and
# filter/*.pm files! It will be used to store strings in perl internally!
my $codepage = 'ISO-8859-2';
my $utf2cp = Text::Iconv->new('UTF-8',$codepage);
# this function will convert data from XML files to local encoding
sub x {
return $utf2cp->convert($_[0]);
}
# decode isis/excel or other import codepage
my $import2cp;
# outgoing xml must be in UTF-8
my $cp2utf = Text::Iconv->new($codepage,'UTF-8');
# mapping between data type and tag which specify
# format in XML file
my %type2tag = (
'isis' => 'isis',
'excel' => 'column',
'marc' => 'marc',
'feed' => 'feed',
);
my $cache; # for cacheing
# 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;
my $last_field_name; # cache to prevent repeated fields
my $broken_cdata = XMLin('<foo><![CDATA[<bar>]]></foo>') eq '<bar>>';
warn "XML::Simple on this system seems broken with <![CDATA[..]]>.\n" if ($broken_cdata);
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 @_;
my $xml;
use parse_format;
my $html = ""; # html formatted display output
my %field_usage; # counter for usage of each field
# sort subrouting using order="" attribute
sub by_order {
my $va = $config->{indexer}->{$a}->{order} ||
$config->{indexer}->{$a};
my $vb = $config->{indexer}->{$b}->{order} ||
$config->{indexer}->{$b};
return $va <=> $vb;
}
my @sorted_tags;
if ($cache->{tags_by_order}) {
@sorted_tags = @{$cache->{tags_by_order}};
} else {
@sorted_tags = sort by_order keys %{$config->{indexer}};
$cache->{tags_by_order} = \@sorted_tags;
}
if (! @sorted_tags) {
print STDERR "WARNING: no tags for this type found in import_xml file!\n";
}
# lookup key
my $lookup_key;
# cache for field in pages
delete $cache->{display_data};
delete $cache->{swish_data};
delete $cache->{swish_exact_data};
delete $cache->{index_data};
delete $cache->{index_delimiter};
delete $cache->{distinct};
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) {
$field_name = x($field_name);
if (! $last_field_name) {
$last_field_name = $field_name;
return $last_field_name;
} elsif ($field_name ne $last_field_name) {
$last_field_name = $field_name;
return $last_field_name;
}
}
}
# init variables for different types
sub init_visible_type($) {
my $type = shift;
# swish, swish_exact, display, index, index_lookup
# swish and display defaults
my ($s,$se,$d,$i,$il) = (1,0,1,0,0);
if (lc($type) eq "display") {
$s = 0;
} elsif (lc($type) eq "swish") {
$d = 0;
} elsif (lc($type) eq "index") {
($s,$se,$d,$i) = (0,1,0,1);
} elsif (lc($type) eq "swish_exact") {
($s,$se,$d,$i) = (0,1,0,0);
} elsif (lc($type) =~ /^lookup/) {
($s,$se,$d,$i,$il) = (0,1,0,0,1);
} elsif ($type) {
print STDERR "WARNING: unknown type: $type\n";
}
return ($s,$se,$d,$i,$il);
}
# convert
#
# <tag>
# <delimiter>, </delimiter>
# <value>200a</value>
# </tag>
#
# to
#
# <tag delimiter=", ">200a</tag>
#
# but without loosing spaces in delimiter (becasue
# new XML::Simple strips spaces in attribute values
# as defined in XML specification)
#
sub unroll_x($) {
my $x = shift;
if (defined $x->{value}) {
my ($v,$d) = ($x->{value}->{content}, $x->{delimiter}->{content});
delete $x->{value};
delete $x->{delimiter};
$x->{content} = $v;
$d =~ s#>$## if ($d && $broken_cdata);
$x->{delimiter} = $d;
}
return $x;
}
# begin real work: go field by field
foreach my $field (@sorted_tags) {
$field=x($field);
$field_usage{$field}++;
my $swish_data = "";
my $swish_exact_data = "";
my $display_data = "";
my @index_data;
my $line_delimiter;
my ($swish,$display);
my $tag = $cfg->val($database, 'import_xml_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/>';
$cache->{index_delimiter}->{$field} = $config->{indexer}->{$field}->{index_delimiter};
my $distinct = $config->{indexer}->{$field}->{distinct};
if ($distinct && !$iterate_by_page) {
warn "WARNING: distinct is currently not supported without iterate_by_page!\n";
$distinct = 0;
}
my $format_name = $config->{indexer}->{$field}->{format_name};
my $format_delimiter = $config->{indexer}->{$field}->{format_delimiter};
if ($format_name && $format_delimiter) {
$cache->{format}->{$field}->{format_name} = $format_name;
$cache->{format}->{$field}->{format_delimiter} = $format_delimiter;
}
foreach my $x (@{$config->{indexer}->{$field}->{$tag}}) {
$x = unroll_x($x);
my $format = x($x->{content});
my $delimiter = x($x->{delimiter}) || ' ';
my $repeat_off = 0; # init repeatable offset
my ($s,$se,$d,$i,$il) = init_visible_type($x->{type});
# what will separate last line from this one?
if ($display_data && $x->{append}) {
$line_delimiter = $delimiter;
} elsif ($display_data) {
$line_delimiter = '<br/>';
}
# init vars so that we go into while...
($swish,$display) = (1,1);
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) {
my $tmp = $swish;
if ($filter && ($s || $se)) {
no strict 'refs';
$tmp = join(" ",&$filter($tmp)) if ($s || $se);
}
$swish_data .= $tmp if ($s && $tmp);
$swish_exact_data .= "xxbxx $tmp xxexx " if ($tmp && $tmp ne "" && $se);
}
# type="display" ; field for display
if ($d && $display) {
my $ldel = $delimiter;
if ($line_delimiter && $display_data) {
$ldel = $line_delimiter;
}
if ($filter) {
no strict 'refs';
my @arr;
foreach my $tmp (&$filter($display)) {
my $tmp2 = mkformat($x,$tmp);
push @arr,$tmp2 if ($tmp2);
}
$display_data .= $ldel if ($display_data && @arr);
$display_data .= join($delimiter,@arr);
} else {
$display_data .= $ldel if ($display_data);
my $tmp = mkformat($x,$display);
$display_data .= $tmp if ($tmp);
}
}
# type="index" ; insert into index
my $idisplay;
if ($i && $display) {
$idisplay = $display;
if ($filter) {
no strict 'refs';
$idisplay = &$filter($idisplay);
}
push @index_data, $idisplay if ($idisplay && !$iterate_by_page);
}
# store fields in lookup
if ($il && $display) {
if (lc($x->{type}) eq "lookup_key") {
if ($lookup_key) {
print STDERR "WARNING: try to redefine lookup_key (keys shouldn't be repeatable fields!)";
} else {
if ($filter) {
no strict 'refs';
$lookup_key = &$filter($display);
} else {
$lookup_key = $display;
}
}
} elsif (lc($x->{type}) eq "lookup_val") {
if ($lookup_key) {
if ($filter) {
no strict 'refs';
$lhash{$lookup_key} = &$filter($display);
} else {
$lhash{$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 $ldel = $page_line_delimiter;
$ldel = " " if ($append);
#print STDERR "line delimiter: ",Dumper($ldel) if ($ldel);
if (! $cache->{$what}->{$field}->[$page]) {
push @{$cache->{$what}->{$field}->[$page]}, {
data => $data,
delimiter => $ldel,
};
}
}
if ($display_data) {
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 = "";
}
if ($idisplay) {
my $ldel=$page_line_delimiter;
my @index_data;
if ($cache->{index_data}->{$field}->[$page]) {
@index_data = @{$cache->{index_data}->{$field}->[$page]};
}
if ($x->{append}) {
if (@index_data) {
$index_data[$#index_data] .= $idisplay;
} else {
push @index_data, $idisplay;
}
} else {
push @index_data, $idisplay;
}
$idisplay = "";
@{$cache->{index_data}->{$field}->[$page]} = @index_data;
}
}
}
if (! $iterate_by_page) {
my $idel = $x->{index_delimiter};
# fill data in index
foreach my $tmp (@index_data) {
my $i = $d = $tmp;
if ($idel && $tmp =~ m/$idel/) {
($i,$d) = split(/$idel/,$tmp);
}
$index->insert($field, $i, $d, $path);
}
@index_data = ();
}
}
# now try to parse variables from configuration file
foreach my $x (@{$config->{indexer}->{$field}->{'config'}}) {
$x = unroll_x($x);
my $delimiter = x($x->{delimiter}) || ' ';
my $val = $cfg->val($database, x($x->{content}));
# FIXME index_lookup is not supported!
my ($s,$se,$d,$i,$il) = init_visible_type($x->{type});
if ($val) {
$display_data .= $delimiter.$val if ($d);
$swish_data .= " ".$val if ($s);
$index->insert($field, $val, $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) {
push @{$cache->{display_data}->{$field}->[$page]}, { data => $display_data };
$display_data = "";
}
if ($swish_data) {
push @{$cache->{swish_data}->{$field}->[$page]}, { data => $swish_data };
$swish_data = "";
}
if ($swish_exact_data) {
push @{$cache->{swish_exact_data}->{$field}->[$page]}, { data => $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});
my $seen; # used for distinct
for (my $page=0; $page <= $nr_pages; $page++) {
my $display_data;
my $delimiter = '';
foreach my $element (@{ $cache->{display_data}->{$field}->[$page] }) {
my $data = $element->{data};
die "BUG! no data in element?" unless ($data);
if ($distinct) {
next if ($cache->{distinct}->{$field}->{ $data });
$cache->{distinct}->{$field}->{ $data } = 1;
}
if ($cache->{format}->{$field}) {
my $tmp = mkformat($cache->{format}->{$field},$data);
$display_data .= $delimiter . $tmp if ($tmp);
} else {
$display_data .= $delimiter . $data;
}
$delimiter = $element->{delimiter} if ($element->{delimiter});
}
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 = join(" ",map { $_->{data} } @{ $cache->{swish_data}->{$field}->[$page] });
if ($swish_data) {
# remove extra spaces
$swish_data =~ s/ +/ /g;
$swish_data =~ s/ +$//g;
$xml .= xmlify($field."_swish", my_unac_string($codepage,$swish_data));
}
my $swish_exact_data = join(" ", map { $_->{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", my_unac_string($codepage,$swish_exact_data));
}
my $idel = $cache->{index_delimiter}->{$field};
foreach my $tmp (@{$cache->{index_data}->{$field}->[$page]}) {
my $i = $tmp;
my $d = $tmp;
if ($idel && $tmp =~ m/$idel/) {
($i,$d) = split(/$idel/,$tmp);
}
$index->insert($field, $i, $d, $path);
#print STDERR "index [$idel] $field: $i --> $d [$path]\n";
}
}
}
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", my_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", my_unac_string($codepage,$swish_exact_data));
}
}
}
# dump formatted output in <html>
if ($html) {
#$xml .= xmlify("html",$html);
$xml .= "<html><![CDATA[ $html ]]></html>";
}
if ($xml) {
$xml .= $add_xml if ($add_xml);
return "<xml>\n$xml</xml>\n";
} else {
return;
}
}
##########################################################################
# 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 $my_unac_filter = $cfg_global->val('global', 'my_unac_filter');
if ($my_unac_filter) {
print STDERR "using $my_unac_filter to filter characters for search\n";
require $my_unac_filter;
} else {
print STDERR "### fallback to default my_unac_string!\n";
eval q{
sub main::my_unac_string($$) {
my ($charset, $string) = (@_);
return $string;
}
};
}
foreach my $database ($cfg->Sections) {
# save database name in global variable path for later
# (need for index filter creation)
$path = $database;
my $type = lc($cfg -> val($database, 'type')) || die "$database doesn't have 'type' defined";
my $add_xml = $cfg -> val($database, 'xml'); # optional
# create new lookup file
my $lookup_file = $cfg -> val($database, 'lookup_newfile'); # optional
if ($lookup_file) {
if (! -e $lookup_file) {
open(LOOKUP, "> $lookup_file") || die "can't create $lookup_file': $!";
close(LOOKUP);
}
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";
# delete memory cache for lookup file
delete $cache->{lhash};
}
# 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";
}
my $import_xml_type = $cfg->val($database, 'import_xml_file') || $type;
my $import_xml_file = "./import_xml/$import_xml_type.xml";
if (! -r $import_xml_file) {
print STDERR "ERROR: file $import_xml_file not readable skipping!\n";
next;
}
print STDERR "reading $import_xml_file\n";
# extract just type basic
my $type_base = $type;
$type_base =~ s/_.+$//g;
my $tag = $cfg->val($database, 'import_xml_tag') || $type2tag{$type_base} || die "can't find which tag to use for type $type";
$config=XMLin($import_xml_file, ForceArray => [ $tag, 'config', 'format' ], ForceContent => 1 );
# check for broken XML::Simple
if ( $broken_cdata ) {
map {
$config->{format}->{$_}->{content} =~ s#>$##;
} keys %{ $config->{format} };
}
# helper for progress bar
sub fmt_time {
my $t = shift || 0;
my $out = "";
my ($ss,$mm,$hh) = gmtime($t);
$out .= "${hh}h" if ($hh);
$out .= sprintf("%02d:%02d", $mm,$ss);
$out .= " " if ($hh == 0);
return $out;
}
# output current progress indicator
my $last_p = 0;
my $start_t = time();
sub progress {
return if (! $show_progress);
my $current = shift;
my $total = shift || 1;
my $p = int($current * 100 / $total);
if ($p < $last_p || $current == 1) {
$start_t = time();
$last_p = 0;
} elsif ($p != $last_p) {
my $rate = ($current / (time() - $start_t || 1));
my $eta = ($total-$current) / ($rate || 1);
printf STDERR ("%5d [%-38s] %-5d %0.1f/s %s\r",$current,"=" x ($p/3)."$p%>", $total, $rate, fmt_time($eta));
$last_p = $p;
}
}
my $fake_dir = 1;
my $fake_pos = 0;
my $last_fake_t = time();
sub fakeprogress {
return if (! $show_progress);
my $current = shift @_;
my @ind = ('-','\\','|','/','-','\\','|','/');
if ($current < $fake_pos) {
$start_t = time();
$last_fake_t = 0;
$fake_dir = 1;
$fake_pos = 0;
}
if (time()-$last_fake_t >= 1) {
$last_fake_t = time();
$fake_pos += $fake_dir;
$fake_dir = -$fake_dir if ($fake_pos > 38);
}
if ($current % 10 == 0) {
my $rate = ($current / (time() - $start_t || 1));
printf STDERR ("%5d [%-38s] %0.1f/s\r",$current, " " x $fake_pos .$ind[($current / 10) % 8], $rate);
}
}
# now read database
# 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 = new Biblio::Isis(
isisdb => $isis_db,
join_subfields_with => ' ; ',
);
if (! $db) {
print STDERR "FATAL: can't read ISIS database: $isis_db, skipping...\n";
next;
}
my $max_rowid = $db->count if ($db);
print STDERR "Reading database: $isis_db [$max_rowid rows]\n";
for (my $row_id = 1; $row_id <= $max_rowid; $row_id++ ) {
my $row = $db->to_hash( $row_id );
if ($row) {
$row->{mfn} = $row_id;
$row->{record} = $db->{record};
progress($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$xml\n";
}
}
}
print STDERR "\n";
} elsif ($type_base eq "excel") {
require Spreadsheet::ParseExcel;
require Spreadsheet::ParseExcel::Utility;
import 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}) || die "no start_row in $type.xml";
$start_row--;
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) {
# this conversion is a cludge.
# Files from Excell could have
# characters which don't fit into
# destination encoding.
$row->{int2col($iC)} = $utf2cp->convert($cell->Value) || $cell->Value;
}
}
progress($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$xml\n";
}
}
print STDERR "\n";
} elsif ($type_base eq "marc") {
require MARC::File::USMARC;
$import2cp = Text::Iconv->new($config->{marc_codepage},$codepage);
my $marc_file = $cfg -> val($database, 'marc_file') || die "$database doesn't have 'marc_file' defined!";
# optional argument is format
warn "marc_format is no longer used!" if ($config->{marc_format});
print STDERR "Reading MARC file '$marc_file'\n";
my $marc = MARC::File::USMARC->in( $marc_file );
if (! $marc) {
print STDERR "FATAL: can't read MARC file: $marc_file, skipping...\n";
next;
}
# count records in MARC file
sub marc_count {
my $filename = shift || die;
my $file = MARC::File::USMARC->in($filename) || return;
my $count = 0;
while ($file->skip()) {
$count++;
}
return $count;
}
my $count = marc_count($marc_file) || warn "no records in '$marc_file'?";
my $i = 1;
while( my $rec = $marc->next() ) {
progress($i,$count);
my $swishpath = $database."#".$i;
if (my $xml = data2xml($type_base,$rec,$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";
}
$i++;
}
print STDERR "\n";
} elsif ($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);
} elsif ($type_base eq "dbf") {
my $dbf_file = $cfg -> val($database, 'dbf_file') || die "$database doesn't have 'dbf_file' defined!";
my $dbf_codepage = $cfg -> val($database, 'dbf_codepage') || die "$database doesn't have 'dbf_codepage' defined!";
my $dbf_mapping = $cfg -> val($database, 'dbf_mapping') || die "$database doesn't have 'dbf_mapping' defined!";
$import2cp = Text::Iconv->new($dbf_codepage,$codepage);
require XBase;
my $db = new XBase $dbf_file;
if (! $db) {
print STDERR "ERROR: can't read DBF database: $dbf_file, skipping...\n";
next;
}
my $max_rowid = $db->last_record;
print STDERR "Reading database: $dbf_file [$max_rowid rows]\n";
my %dbf2iso;
foreach my $m (split(/[\n\r]+/,$dbf_mapping)) {
my ($col,$fld) = split(/\s+/,$m,2);
$dbf2iso{$col} = $fld;
}
#print STDERR "## dbf2iso: ",Dumper(\%dbf2iso),"\n## /dbf2iso\n";
# bad, bad...
require "to_hash.pm";
foreach my $row_id (0 .. $max_rowid) {
my $dbf_row = $db->get_record_as_hash($row_id);
if ($dbf_row) {
#print STDERR "## dbf_row: ",Dumper($dbf_row),"\n## /dbf_row\n";
# apply mapping from config file
# all unspecified records will get _ in
# front of them - _DELETE will be __DELETE
my $rec;
map {
my $new_fld = $dbf2iso{$_} || '_'.$_;
my $data = $dbf_row->{$_};
push @{ $rec->{$new_fld} }, $data if ($data && $data !~ /^(?:\s+|\$a\.|)$/);
} keys %{$dbf_row};
#print STDERR "## rec: ",Dumper($rec),"\n## /rec\n";
my $row = to_hash($row_id+1, $rec);
$row->{mfn} = $row_id+1;
$row->{record} = $rec;
#print STDERR "## row: ",Dumper($row),"\n## /row\n";
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";
}
}
}
print STDERR "\n";
}
}
# call this to commit index
$index->close;
1;
__END__
##########################################################################
=head1 NAME
all2xml.pl - read various file formats and dump XML for SWISH-E
=head1 SYNOPSYS
$ all2xml.pl [test.conf]
=head1 DESCRIPTION
This command will read ISIS data file using Biblio::Isis perl module, MARC
records using MARC::File 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.
If no configuration file is specified, it will use default one called
C<all2xml.conf>.
=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
RSS Feed