--- trunk/all2xml.pl 2003/03/15 21:48:48 42 +++ trunk/all2xml.pl 2003/07/13 21:57:12 97 @@ -18,9 +18,9 @@ my $config; -use index_DBI; # there is no other, right now ;-) - -my $index = new index_DBI(); # open index +#use index_DBI; # default DBI module for index +use index_DBI_cache; # faster DBI module using memory cache +my $index; my %opts; @@ -32,9 +32,9 @@ getopts('d:m:qs', \%opts); -my $db_dir; +my $path; # this is name of database -Text::Iconv->raise_error(1); # Conversion errors raise exceptions +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! @@ -46,18 +46,31 @@ return $utf2cp->convert($_[0]); } -# decode isis import codepage -my $isis2cp; +# decode isis/excel or other import codepage +my $import2cp; # outgoing xml must be in UTF-8 my $cp2utf = Text::Iconv->new($codepage,'UTF-8'); -sub isis2xml { +# mapping between data type and tag which specify +# format in XML file +my %type2tag = ( + 'isis' => 'isis', + 'excel' => 'column', + 'marc' => 'marc', + 'feed' => 'feed' +); + +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 $xml; @@ -88,12 +101,13 @@ my ($swish,$display); - foreach my $x (@{$config->{indexer}->{$field}->{isis}}) { + my $tag = $type2tag{$type} || die "can't find which tag to use for type $type"; + foreach my $x (@{$config->{indexer}->{$field}->{$tag}}) { my $format = x($x->{content}); my $delimiter = x($x->{delimiter}) || ' '; - my $isis_i = 0; # isis repeatable offset + my $repeat_off = 0; # repeatable offset my ($s,$d,$i) = (1,1,0); # swish, display default $s = 0 if (lc($x->{type}) eq "display"); @@ -110,9 +124,18 @@ # init vars so that we go into while... ($swish,$display) = (1,1); - while ($swish || $display) { - ($swish,$display) = parse_format($format,$row,$isis_i++,$isis2cp); + # 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 = $x->{filter}; @@ -124,7 +147,6 @@ if ($filter) { no strict 'refs'; $swish_data .= join(" ",&$filter($swish)); -print STDERR "#### $swish_data\n"; } else { $swish_data .= $swish; } @@ -150,19 +172,45 @@ # type="index" ; insert into index if ($i && $display) { - my $index_data = $display; - if ($filter) { - no strict 'refs'; - foreach my $d (&$filter($index_data)) { - $index->insert($field, $d, $db_dir); - } - } else { - $index->insert($field, $index_data, $db_dir); + 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); } } } } + # 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 ($display_data) { @@ -200,7 +248,8 @@ # dump formatted output in if ($html) { - $xml .= xmlify("html",$html); + #$xml .= xmlify("html",$html); + $xml .= ""; } if ($xml) { @@ -213,55 +262,225 @@ ########################################################################## +# 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'); + foreach my $database ($cfg->Sections) { - my $isis_db = $cfg -> val($database, 'isis_db') || die "$database doesn't have 'isis_db' defined!"; - my $type = $cfg -> val($database, 'type') || die "$database doesn't have 'type' defined"; + my $type = lc($cfg -> val($database, 'type')) || die "$database doesn't have 'type' defined"; my $add_xml = $cfg -> val($database, 'xml'); # optional - $config=XMLin("./import_xml/$type.xml", forcearray => [ 'isis' ], forcecontent => 1); +print STDERR "reading ./import_xml/$type.xml\n"; + + # extract just type basic + my $type_base = $type; + $type_base =~ s/_.+$//g; - $isis2cp = Text::Iconv->new($config->{isis_codepage},$codepage); + $config=XMLin("./import_xml/$type.xml", forcearray => [ $type2tag{$type_base}, 'config' ], forcecontent => 1); - my $db = OpenIsis::open( $isis_db ); - if (0) { -# # FIX -# if (! $db ) { - print STDERR "WARNING: can't open '$isis_db'\n"; - next ; + # output current progress indicator + my $last_p = 0; + sub progress { + # XXX return if ($show_progress ne ""); + 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; + } } - my $max_rowid = OpenIsis::maxRowid( $db ); + 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]); + } + } - print STDERR "Reading database: $isis_db [$max_rowid rows]\n"; + # now read database +print STDERR "using: $type...\n"; - my $path = $database; # was $isis_db + if ($type_base eq "isis") { - my $last_p = 0; + 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 = OpenIsis::open( $isis_db ); + + my $max_rowid = OpenIsis::maxRowid( $db ); + + print STDERR "Reading database: $isis_db [$max_rowid rows]\n"; + + my $path = $database; + + for (my $row_id = 1; $row_id <= $max_rowid; $row_id++ ) { + my $row = OpenIsis::read( $db, $row_id ); + if ($row && $row->{mfn}) { + + progress($row->{mfn}, $max_rowid); + + my $swishpath = $path."#".int($row->{mfn}); -# { my $row_id = 4514; -# FIX - for (my $row_id = 1; $row_id <= $max_rowid; $row_id++ ) { - my $row = OpenIsis::read( $db, $row_id ); - if ($row && $row->{mfn}) { - # output current process indicator - my $p = int($row->{mfn} * 100 / $max_rowid); - if ($p != $last_p) { - printf STDERR ("%5d / %5d [%-51s] %-2d %% \r",$row->{mfn},$max_rowid,"=" x ($p/2).">", $p ) if (! $opts{q}); - $last_p = $p; + 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") { + 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++; + } + + 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) { + $row->{int2col($iC)} = $cell->Value; + } + } + + progress($iR, $end_row); + +# print "row[$iR/$end_row] "; +# foreach (keys %{$row}) { +# print "$_: ",$row->{$_},"\t"; +# } +# print "\n"; + + my $swishpath = $database."#".$iR; - if (my $xml = $cp2utf->convert(isis2xml($row,$add_xml))) { + 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: $path#".int($row->{mfn})."\n"; + 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: $swishpath\n"; + print "Content-Length: ".(length($xml)+1)."\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 $line=1; + + while () { + chomp; + + if (/^$/) { + 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); + + } } - print STDERR "\n"; } # call this to commit index @@ -273,15 +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