--- trunk/IsisDB.pm 2004/12/29 22:46:40 15 +++ trunk/IsisDB.pm 2005/01/05 21:23:04 33 @@ -2,12 +2,14 @@ use strict; use Carp; +use File::Glob qw(:globally :nocase); + use Data::Dumper; BEGIN { use Exporter (); use vars qw ($VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS); - $VERSION = 0.05; + $VERSION = 0.09; @ISA = qw (Exporter); #Give a hoot don't pollute, do not export more than needed by default @EXPORT = qw (); @@ -28,14 +30,15 @@ isisdb => './cds/cds', ); - for(my $mfn = 1; $mfn <= $isis->{'maxmfn'}; $mfn++) { + for(my $mfn = 1; $mfn <= $isis->count; $mfn++) { print $isis->to_ascii($mfn),"\n"; } =head1 DESCRIPTION This module will read ISIS databases created by DOS CDS/ISIS, WinIsis or -IsisMarc. It can be used as perl-only alternative to OpenIsis module. +IsisMarc. It can be used as perl-only alternative to OpenIsis module which +seems to depriciate it's old C bindings for perl. It can create hash values from data in ISIS database (using C), ASCII dump (using C) or just hash with field names and packed @@ -48,7 +51,7 @@ It also has support for identifiers (only if ISIS database is created by IsisMarc), see C. -This will module will always be slower than OpenIsis module which use C +This module will always be slower than OpenIsis module which use C library. However, since it's written in perl, it's platform independent (so you don't need C compiler), and can be easily modified. I hope that it creates data structures which are easier to use than ones created by @@ -96,7 +99,8 @@ =item isisdb This is full or relative path to ISIS database files which include -common prefix of C<.FDT>, C<.MST>, C<.CNT>, C<.XRF> and C<.MST> files. +common prefix of C<.MST>, and C<.XRF> and optionally C<.FDT> (if using +C option) files. In this example it uses C<./cds/cds.MST> and related files. @@ -119,8 +123,6 @@ =back -It will also set C<$isis-E{'maxmfn'}> which is maximum MFN stored in database. - =cut sub new { @@ -134,15 +136,32 @@ $self->{$v} = {@_}->{$v}; } + my @isis_files = grep(/\.(FDT|MST|XRF|CNT)$/i,glob($self->{isisdb}."*")); + + foreach my $f (@isis_files) { + my $ext = $1 if ($f =~ m/\.(\w\w\w)$/); + $self->{lc($ext)."_file"} = $f; + } + + my @must_exist = qw(mst xrf); + push @must_exist, "fdt" if ($self->{read_fdt}); + + foreach my $ext (@must_exist) { + croak "missing ",uc($ext)," file in ",$self->{isisdb} unless ($self->{$ext."_file"}); + } + + print STDERR "## using files: ",join(" ",@isis_files),"\n" if ($self->{debug}); + # if you want to read .FDT file use read_fdt argument when creating class! - if ({@_}->{read_fdt} && -e $self->{isisdb}.".FDT") { + if ($self->{read_fdt} && -e $self->{fdt_file}) { # read the $db.FDT file for tags my $fieldzone=0; - open(fileFDT, $self->{isisdb}.".FDT") || croak "can't read '$self->{isisdb}.FDT': $!"; + open(my $fileFDT, $self->{fdt_file}) || croak "can't read '$self->{fdt_file}': $!"; + binmode($fileFDT); - while () { + while (<$fileFDT>) { chomp; if ($fieldzone) { my $name=substr($_,0,30); @@ -159,12 +178,13 @@ } } - close(fileFDT); + close($fileFDT); } # Get the Maximum MFN from $db.MST - open(fileMST,$self->{isisdb}.".MST") || croak "can't read '$self->{isisdb}.MST': $!"; + open($self->{'fileMST'}, $self->{mst_file}) || croak "can't open '$self->{mst_file}': $!"; + binmode($self->{'fileMST'}); # MST format: (* = 32 bit signed) # CTLMFN* always 0 @@ -172,69 +192,107 @@ # NXTMFB* last block allocated to master file # NXTMFP offset to next available position in last block # MFTYPE always 0 for user db file (1 for system) - seek(fileMST,4,0); + seek($self->{'fileMST'},4,0) || carp "can't seek to offset 0 in MST: $!"; my $buff; - read(fileMST, $buff, 4); - $self->{'NXTMFN'}=unpack("l",$buff) || carp "NXTNFN is zero"; + read($self->{'fileMST'}, $buff, 4) || carp "can't read NXTMFN from MST: $!"; + $self->{'NXTMFN'}=unpack("V",$buff) || carp "NXTNFN is zero"; - # save maximum MFN - $self->{'maxmfn'} = $self->{'NXTMFN'} - 1; + print STDERR Dumper($self),"\n" if ($self->{debug}); - close(fileMST); + # open files for later + open($self->{'fileXRF'}, $self->{xrf_file}) || croak "can't open '$self->{xrf_file}': $!"; + binmode($self->{'fileXRF'}); + + $self ? return $self : return undef; +} + +=head2 count + +Return number of records in database + + print $isis->count; + +=cut + +sub count { + my $self = shift; + return $self->{'NXTMFN'} - 1; +} + +=head2 read_cnt + +Read content of C<.CNT> file and return hash containing it. + + print Dumper($isis->read_cnt); + +This function is not used by module (C<.CNT> files are not required for this +module to work), but it can be useful to examine your index (while debugging +for example). + +=cut + +sub read_cnt { + my $self = shift; + + croak "missing CNT file in ",$self->{isisdb} unless ($self->{cnt_file}); # Get the index information from $db.CNT - open(fileCNT, $self->{isisdb}.".CNT") || croak "can't read '$self->{isisdb}.CNT': $!"; - - # There is two 26 Bytes fixed lenght records + open(my $fileCNT, $self->{cnt_file}) || carp "can't read '$self->{cnt_file}': $!"; + binmode($fileCNT); - # 0: IDTYPE BTree type 16 - # 2: ORDN Nodes Order 16 - # 4: ORDF Leafs Order 16 - # 6: N Number of Memory buffers for nodes 16 - # 8: K Number of buffers for first level index 16 - # 10: LIV Current number of Index Levels 16 - # 12: POSRX* Pointer to Root Record in N0x 32 - # 16: NMAXPOS* Next Available position in N0x 32 - # 20: FMAXPOS* Next available position in L0x 32 - # 24: ABNORMAL Formal BTree normality indicator 16 - # length: 26 bytes - - sub unpack_cnt { - my $self = shift; - - my @flds = qw(ORDN ORDF N K LIV POSRX NMAXPOS FMAXPOS ABNORMAL); - - my $buff = shift || return; - my @arr = unpack("ssssssllls", $buff); - - print "unpack_cnt: ",join(" ",@arr),"\n" if ($self->{'debug'}); - - my $IDTYPE = shift @arr; - foreach (@flds) { - $self->{$IDTYPE}->{$_} = abs(shift @arr); - } - } + my $buff; - read(fileCNT, $buff, 26); + read($fileCNT, $buff, 26) || carp "can't read first table from CNT: $!"; $self->unpack_cnt($buff); - read(fileCNT, $buff, 26); + read($fileCNT, $buff, 26) || carp "can't read second table from CNT: $!"; $self->unpack_cnt($buff); + close($fileCNT); - close(fileCNT); + return $self->{cnt}; +} - print Dumper($self),"\n" if ($self->{debug}); +=head2 unpack_cnt - # open files for later - open($self->{'fileXRF'}, $self->{isisdb}.".XRF") || croak "can't open '$self->{isisdb}.XRF': $!"; +Unpack one of two 26 bytes fixed length record in C<.CNT> file. - open($self->{'fileMST'}, $self->{isisdb}.".MST") || croak "can't open '$self->{isisdb}.MST': $!"; +Here is definition of record: - $self ? return $self : return undef; + off key description size + 0: IDTYPE BTree type s + 2: ORDN Nodes Order s + 4: ORDF Leafs Order s + 6: N Number of Memory buffers for nodes s + 8: K Number of buffers for first level index s + 10: LIV Current number of Index Levels s + 12: POSRX Pointer to Root Record in N0x l + 16: NMAXPOS Next Available position in N0x l + 20: FMAXPOS Next available position in L0x l + 24: ABNORMAL Formal BTree normality indicator s + length: 26 bytes + +This will fill C<$self> object under C with hash. It's used by C. + +=cut + +sub unpack_cnt { + my $self = shift; + + my @flds = qw(ORDN ORDF N K LIV POSRX NMAXPOS FMAXPOS ABNORMAL); + + my $buff = shift || return; + my @arr = unpack("vvvvvvVVVv", $buff); + + print STDERR "unpack_cnt: ",join(" ",@arr),"\n" if ($self->{'debug'}); + + my $IDTYPE = shift @arr; + foreach (@flds) { + $self->{cnt}->{$IDTYPE}->{$_} = abs(shift @arr); + } } =head2 fetch @@ -258,71 +316,77 @@ my $mfn = shift || croak "fetch needs MFN as argument!"; - print "fetch: $mfn\n" if ($self->{debug}); + # is mfn allready in memory? + my $old_mfn = $self->{'current_mfn'} || -1; + return $self->{record} if ($mfn == $old_mfn); + + print STDERR "## fetch: $mfn\n" if ($self->{debug}); # XXX check this? my $mfnpos=($mfn+int(($mfn-1)/127))*4; - print "seeking to $mfnpos in file '$self->{isisdb}.XRF'\n" if ($self->{debug}); + print STDERR "## seeking to $mfnpos in file '$self->{xrf_file}'\n" if ($self->{debug}); seek($self->{'fileXRF'},$mfnpos,0); my $buff; + # delete old record + delete $self->{record}; + # read XRFMFB abd XRFMFP read($self->{'fileXRF'}, $buff, 4); - my $pointer=unpack("l",$buff) || carp "pointer is null"; + my $pointer=unpack("V",$buff) || carp "pointer is null"; + + # check for logically deleted record + if ($pointer & 0x80000000) { + print STDERR "## record $mfn is logically deleted\n" if ($self->{debug}); + $self->{deleted} = $mfn; + + return unless $self->{include_deleted}; + + # abs + $pointer = ($pointer ^ 0xffffffff) + 1; + } my $XRFMFB = int($pointer/2048); my $XRFMFP = $pointer - ($XRFMFB*2048); - print "XRFMFB: $XRFMFB XRFMFP: $XRFMFP\n" if ($self->{debug}); + # (XRFMFB - 1) * 512 + XRFMFP + # why do i have to do XRFMFP % 1024 ? - # XXX fix this to be more readable!! - # e.g. (XRFMFB - 1) * 512 + XRFMFP - - my $offset = $pointer; - my $offset2=int($offset/2048)-1; - my $offset22=int($offset/4096); - my $offset3=$offset-($offset22*4096); - if ($offset3>512) { - $offset3=$offset3-2048; - } - my $offset4=($offset2*512)+$offset3; + my $blk_off = (($XRFMFB - 1) * 512) + ($XRFMFP % 512); - print "$offset - $offset2 - $offset3 - $offset4\n" if ($self->{debug}); + print STDERR "## pointer: $pointer XRFMFB: $XRFMFB XRFMFP: $XRFMFP offset: $blk_off\n" if ($self->{'debug'}); # Get Record Information - seek($self->{'fileMST'},$offset4,0); + seek($self->{'fileMST'},$blk_off,0) || croak "can't seek to $blk_off: $!"; - read($self->{'fileMST'}, $buff, 4); - my $value=unpack("l",$buff); + read($self->{'fileMST'}, $buff, 4) || croak "can't read 4 bytes at offset $blk_off from MST file: $!"; + my $value=unpack("V",$buff); + + print STDERR "## offset for rowid $value is $blk_off (blk $XRFMFB off $XRFMFP)\n" if ($self->{debug}); if ($value!=$mfn) { -print ("Error: The MFN:".$mfn." is not found in MST(".$value.")"); - return -1; # XXX deleted record? - } + if ($value == 0) { + print STDERR "## record $mfn is physically deleted\n" if ($self->{debug}); + $self->{deleted} = $mfn; + return; + } -# $MFRL=$self->Read16($fileMST); -# $MFBWB=$self->Read32($fileMST); -# $MFBWP=$self->Read16($fileMST); -# $BASE=$self->Read16($fileMST); -# $NVF=$self->Read16($fileMST); -# $STATUS=$self->Read16($fileMST); + carp "Error: MFN ".$mfn." not found in MST file, found $value"; + return; + } read($self->{'fileMST'}, $buff, 14); - my ($MFRL,$MFBWB,$MFBWP,$BASE,$NVF,$STATUS) = unpack("slssss", $buff); + my ($MFRL,$MFBWB,$MFBWP,$BASE,$NVF,$STATUS) = unpack("vVvvvv", $buff); - print "MFRL: $MFRL MFBWB: $MFBWB MFBWP: $MFBWP BASE: $BASE NVF: $NVF STATUS: $STATUS\n" if ($self->{debug}); + print STDERR "## MFRL: $MFRL MFBWB: $MFBWB MFBWP: $MFBWP BASE: $BASE NVF: $NVF STATUS: $STATUS\n" if ($self->{debug}); - # delete old record - delete $self->{record}; + warn "MFRL $MFRL is not even number" unless ($MFRL % 2 == 0); - if (! $self->{'include_deleted'} && $MFRL < 0) { - print "## logically deleted record $mfn, skipping...\n" if ($self->{debug}); - return; - } + warn "BASE is not 18+6*NVF" unless ($BASE == 18 + 6 * $NVF); # Get Directory Format @@ -332,17 +396,13 @@ read($self->{'fileMST'}, $buff, 6 * $NVF); - my $fld_len = 0; + my $rec_len = 0; for (my $i = 0 ; $i < $NVF ; $i++) { -# $TAG=$self->Read16($fileMST); -# $POS=$self->Read16($fileMST); -# $LEN=$self->Read16($fileMST); - - my ($TAG,$POS,$LEN) = unpack("sss", substr($buff,$i * 6, 6)); + my ($TAG,$POS,$LEN) = unpack("vvv", substr($buff,$i * 6, 6)); - print "TAG: $TAG POS: $POS LEN: $LEN\n" if ($self->{debug}); + print STDERR "## TAG: $TAG POS: $POS LEN: $LEN\n" if ($self->{debug}); # The TAG does not exists in .FDT so we set it to 0. # @@ -358,12 +418,14 @@ push @FieldPOS,$POS; push @FieldLEN,$LEN; - $fld_len += $LEN; + $rec_len += $LEN; } # Get Variable Fields - read($self->{'fileMST'},$buff,$fld_len); + read($self->{'fileMST'},$buff,$rec_len); + + print STDERR "## rec_len: $rec_len poc: ",tell($self->{'fileMST'})."\n" if ($self->{debug}); for (my $i = 0 ; $i < $NVF ; $i++) { # skip zero-sized fields @@ -371,20 +433,21 @@ push @{$self->{record}->{$FieldTAG[$i]}}, substr($buff,$FieldPOS[$i],$FieldLEN[$i]); } - close(fileMST); - print Dumper($self),"\n" if ($self->{debug}); + $self->{'current_mfn'} = $mfn; + + print STDERR Dumper($self),"\n" if ($self->{debug}); return $self->{'record'}; } =head2 to_ascii -Dump ASCII output of record with specified MFN +Returns ASCII output of record with specified MFN print $isis->to_ascii(42); -It outputs something like this: +This outputs something like this: 210 ^aNew York^cNew York University press^dcop. 1988 990 2140 @@ -421,7 +484,7 @@ my $hash = $isis->to_hash($mfn); -It has ability to convert characters (using C from ISIS +It has ability to convert characters (using C) from ISIS database before creating structures enabling character re-mapping or quick fix-up of data. @@ -467,7 +530,7 @@ my $mfn = shift || confess "need mfn!"; # init record to include MFN as field 000 - my $rec = { '000' => $mfn }; + my $rec = { '000' => [ $mfn ] }; my $row = $self->fetch($mfn); @@ -480,7 +543,7 @@ my $val; # has identifiers? - ($val->{'i1'},$val->{'i2'}) = ($1,$2) if ($l =~ s/^([01 #])([01 #])//); + ($val->{'i1'},$val->{'i2'}) = ($1,$2) if ($l =~ s/^([01 #])([01 #])\^/\^/); # has subfields? if ($l =~ m/\^/) { @@ -517,7 +580,18 @@ =head1 BUGS -This module has been very lightly tested. Use with caution and report bugs. +Some parts of CDS/ISIS documentation are not detailed enough to exmplain +some variations in input databases which has been tested with this module. +When I was in doubt, I assumed that OpenIsis's implementation was right +(except for obvious bugs). + +However, every effort has been made to test this module with as much +databases (and programs that create them) as possible. + +I would be very greatful for success or failure reports about usage of this +module with databases from programs other than WinIsis and IsisMarc. I had +tested this against ouput of one C-based application, but I don't +know any details about it's version. =head1 AUTHOR