--- trunk/all2xml.pl 2003/07/04 20:11:48 62 +++ trunk/all2xml.pl 2003/07/13 22:29:14 98 @@ -58,6 +58,7 @@ 'isis' => 'isis', 'excel' => 'column', 'marc' => 'marc', + 'feed' => 'feed' ); sub data2xml { @@ -81,17 +82,17 @@ # sort subrouting using order="" attribute sub by_order { - return 0 if (! $config->{indexer}->{$a}->{order}); - return 0 if (! $config->{indexer}->{$b}->{order}); + my $va = $config->{indexer}->{$a}->{order} || + $config->{indexer}->{$a}; + my $vb = $config->{indexer}->{$b}->{order} || + $config->{indexer}->{$b}; - return $config->{indexer}->{$a}->{order} <=> - $config->{indexer}->{$b}->{order} ; + return $va <=> $vb; } foreach my $field (sort by_order keys %{$config->{indexer}}) { $field=x($field); - $field_usage{$field}++; my $swish_data = ""; @@ -123,8 +124,18 @@ # init vars so that we go into while... ($swish,$display) = (1,1); - if ($swish || $display) { + # placeholder for all repeatable entries for index + my @index_data; + my $index_filter; + + # while because of repeatable fields + while ($swish || $display) { ($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; + } + # filter="name" ; filter this field through # filter/[name].pm my $filter = $x->{filter}; @@ -161,14 +172,21 @@ # type="index" ; insert into index if ($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); + push @index_data, $display; + $index_filter = $filter if ($filter); + } + } + + # fill data in index + if (@index_data) { + if ($index_filter) { + no strict 'refs'; + foreach my $d (&$index_filter(@index_data)) { + $index->insert($field, $d, $path); + } + } else { + foreach my $d (@index_data) { + $index->insert($field, $d, $path); } } } @@ -230,7 +248,8 @@ # dump formatted output in if ($html) { - $xml .= xmlify("html",$html); + #$xml .= xmlify("html",$html); + $xml .= ""; } if ($xml) { @@ -257,6 +276,8 @@ $cfg_global->val('global', 'dbi_passwd') || '', ); +my $show_progress = $cfg_global->val('global', 'show_progress'); + foreach my $database ($cfg->Sections) { my $type = lc($cfg -> val($database, 'type')) || die "$database doesn't have 'type' defined"; @@ -273,7 +294,7 @@ # output current progress indicator my $last_p = 0; sub progress { - #return if (! $opts{q}); # FIXME + # XXX return if ($show_progress ne ""); my $current = shift; my $total = shift || 1; my $p = int($current * 100 / $total); @@ -283,6 +304,19 @@ } } + my $fake_dir = 1; + sub fakeprogress { + 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: $type...\n"; @@ -373,7 +407,7 @@ } } } elsif ($type_base eq "marc") { - ## XXX + use MARC; $import2cp = Text::Iconv->new($config->{marc_codepage},$codepage); @@ -382,8 +416,6 @@ # optional argument is format my $format = x($config->{format}) || 'usmarc'; - my %id_stored; # to aviod duplicates - print STDERR "Reading MARC file '$marc_file'\n"; my $marc = new MARC; @@ -392,17 +424,13 @@ }) || die "Can't open MARC file '$marc_file'"; my $i=0; # record nr. - my $inc=1; - my $max_i=1000; my $rec; while ($marc->nextmarc(1)) { # XXX - progress($i, $max_i); - $i += $inc; - $inc = -$inc if ($i > $max_i || $i < 0); + fakeprogress($i++); my $swishpath = $database."#".$i; @@ -414,6 +442,44 @@ 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 $line=1; + + while () { + chomp; + + if (/^$/) { + 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); + + } } } @@ -426,15 +492,21 @@ =head1 NAME -isis2xml.pl - read isis file and dump XML +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 and -create XML file for usage with I -indexer. Dispite it's name, this script B -from isis files (isis allready has something like that). Output of this -script is tailor-made for SWISH-E. +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 indexer. Dispite it's name, +this script B 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