# format in XML file
my %type2tag = (
'isis' => 'isis',
- 'excel' => 'column'
+ 'excel' => 'column',
+ 'marc' => 'marc',
+ 'feed' => 'feed'
);
sub data2xml {
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;
# 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 $swish_data = "";
# init vars so that we go into while...
($swish,$display) = (1,1);
- if ($swish || $display) {
+ # placeholder for all repeatable entries for index
+ my @index_data;
+ my $index_filter;
+
+ 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 || $display) {
($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 = $x->{filter};
}
if ($filter) {
no strict 'refs';
- $display_data .= join($delimiter,&$filter($display));
+ if ($display_data) {
+ $display_data .= $delimiter.join($delimiter,mkformat($x,&$filter($display)));
+ } else {
+ $display_data = join($delimiter,mkformat($x,&$filter($display)));
+ }
} else {
if ($display_data) {
- $display_data .= $delimiter.$display;
+ $display_data .= $delimiter.mkformat($x,$display);
} else {
- $display_data .= $display;
+ $display_data = mkformat($x,$display);
}
}
}
# type="index" ; insert into index
if ($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);
+ 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_data) {
+ $index->insert($field, &$index_filter($d), $path);
+ }
+ } else {
+ foreach my $d (@index_data) {
+ $index->insert($field, $d, $path);
}
}
}
}
+ # now try to parse variables from configuration file
+ foreach my $x (@{$config->{indexer}->{$field}->{'config'}}) {
+
+ my $delimiter = x($x->{delimiter}) || ' ';
+ my $val = $cfg->val($database, x($x->{content}));
+
+ my ($s,$d,$i) = (1,1,0); # swish, display default
+ $s = 0 if (lc($x->{type}) eq "display");
+ $d = 0 if (lc($x->{type}) eq "swish");
+ ($s,$d,$i) = (0,0,1) if (lc($x->{type}) eq "index");
+
+ if ($val) {
+ $display_data .= $delimiter.$val if ($d);
+ $swish_data .= $val if ($s);
+ $index->insert($field, $val, $path) if ($i);
+ }
+
+ }
+
if ($display_data) {
# dump formatted output in <html>
if ($html) {
- $xml .= xmlify("html",$html);
+ #$xml .= xmlify("html",$html);
+ $xml .= "<html><![CDATA[ $html ]]></html>";
}
if ($xml) {
$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";
my $type_base = $type;
$type_base =~ s/_.+$//g;
- $config=XMLin("./import_xml/$type.xml", forcearray => [ $type2tag{$type_base} ], forcecontent => 1);
+ $config=XMLin("./import_xml/$type.xml", forcearray => [ $type2tag{$type_base}, 'config', 'format' ], forcecontent => 1);
# output current progress indicator
my $last_p = 0;
sub progress {
- #return if (! $opts{q}); # FIXME
+ return if (! $show_progress);
my $current = shift;
my $total = shift || 1;
my $p = int($current * 100 / $total);
}
}
+ my $fake_dir = 1;
+ sub fakeprogress {
+ return if (! $show_progress);
+ 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";
$import2cp = Text::Iconv->new($config->{isis_codepage},$codepage);
my $db = OpenIsis::open( $isis_db );
+ # check if .txt database for OpenIsis is zero length,
+ # if so, erase it and re-open database
+ sub check_txt_db {
+ my $isis_db = shift || die "need isis database name";
+ my $reopen = 0;
+
+ if (-e $isis_db.".TXT") {
+ print STDERR "WARNING: removing $isis_db.TXT OpenIsis database...\n";
+ unlink $isis_db.".TXT" || warn "FATAL: unlink error on '$isis_db.TXT': $!";
+ $reopen++;
+ }
+ if (-e $isis_db.".PTR") {
+ print STDERR "WARNING: removing $isis_db.PTR OpenIsis database...\n";
+ unlink $isis_db.".PTR" || warn "FATAL: unlink error on '$isis_db.PTR': $!";
+ $reopen++;
+ }
+ return OpenIsis::open( $isis_db ) if ($reopen);
+ }
+
+ # EOF error
+ if ($db == -1) {
+ $db = check_txt_db($isis_db);
+ if ($db == -1) {
+ print STDERR "FATAL: OpenIsis can't open zero size file $isis_db\n";
+ next;
+ }
+ }
+
+ # OpenIsis::ERR_BADF
+ if ($db == -4) {
+ print STDERR "FATAL: OpenIsis can't find file $isis_db\n";
+ next;
+ # OpenIsis::ERR_IO
+ } elsif ($db == -5) {
+ print STDERR "FATAL: OpenIsis can't access file $isis_db\n";
+ next;
+ } elsif ($db < 0) {
+ print STDERR "FATAL: OpenIsis unknown error $db with file $isis_db\n";
+ next;
+ }
+
my $max_rowid = OpenIsis::maxRowid( $db );
+ # if 0 records, try to rease isis .txt database
+ if ($max_rowid == 0) {
+ # force removal of database
+ $db = check_txt_db($isis_db);
+ $max_rowid = OpenIsis::maxRowid( $db );
+ }
+
print STDERR "Reading database: $isis_db [$max_rowid rows]\n";
my $path = $database;
my $swishpath = $path."#".int($row->{mfn});
- if (my $xml = data2xml($type_base,$row,$add_xml)) {
+ if (my $xml = data2xml($type_base,$row,$add_xml,$cfg,$database)) {
$xml = $cp2utf->convert($xml);
use bytes; # as opposed to chars
print "Path-Name: $swishpath\n";
}
}
}
+ # for this to work with current version of OpenIsis (0.9.0)
+ # you might need my patch from
+ # http://www.rot13.org/~dpavlin/projects/openisis-0.9.0-perl_close.diff
+ OpenIsis::close($db);
print STDERR "\n";
} elsif ($type_base eq "excel") {
next if (! $row);
- if (my $xml = data2xml($type,$row,$add_xml)) {
+ 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$xml\n";
+ }
+ }
+ } elsif ($type_base eq "marc") {
+
+ use MARC;
+
+ $import2cp = Text::Iconv->new($config->{marc_codepage},$codepage);
+ my $marc_file = $cfg -> val($database, 'marc_file') || die "$database doesn't have 'marc_file' defined!";
+
+ # optional argument is format
+ my $format = x($config->{format}) || 'usmarc';
+
+ print STDERR "Reading MARC file '$marc_file'\n";
+
+ my $marc = new MARC;
+ my $nr = $marc->openmarc({
+ file=>$marc_file, format=>$format
+ }) || die "Can't open MARC file '$marc_file'";
+
+ my $i=0; # record nr.
+
+ my $rec;
+
+ while ($marc->nextmarc(1)) {
+
+ # XXX
+ fakeprogress($i++);
+
+ my $swishpath = $database."#".$i;
+
+ if (my $xml = data2xml($type_base,$marc,$add_xml,$cfg,$database)) {
$xml = $cp2utf->convert($xml);
use bytes; # as opposed to chars
print "Path-Name: $swishpath\n";
print "Document-Type: XML\n\n$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 $line=1;
+
+ while (<FEED>) {
+ chomp;
+
+ if (/^$/) {
+ 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);
+
+ }
}
}
=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<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.
+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<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.
+
+=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
projects / webpac / blobdiff