/[Biblio-Isis]/trunk/lib/Biblio/Isis.pm
This is repository of my old source code which isn't updated any more. Go to git.rot13.org for current projects!
ViewVC logotype

Diff of /trunk/lib/Biblio/Isis.pm

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 15 by dpavlin, Wed Dec 29 22:46:40 2004 UTC revision 25 by dpavlin, Fri Dec 31 05:43:20 2004 UTC
# Line 2  package IsisDB; Line 2  package IsisDB;
2  use strict;  use strict;
3    
4  use Carp;  use Carp;
5    use File::Glob qw(:globally :nocase);
6    
7  use Data::Dumper;  use Data::Dumper;
8    
9  BEGIN {  BEGIN {
10          use Exporter ();          use Exporter ();
11          use vars qw ($VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS);          use vars qw ($VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS);
12          $VERSION     = 0.05;          $VERSION     = 0.08;
13          @ISA         = qw (Exporter);          @ISA         = qw (Exporter);
14          #Give a hoot don't pollute, do not export more than needed by default          #Give a hoot don't pollute, do not export more than needed by default
15          @EXPORT      = qw ();          @EXPORT      = qw ();
# Line 96  Options are described below: Line 98  Options are described below:
98  =item isisdb  =item isisdb
99    
100  This is full or relative path to ISIS database files which include  This is full or relative path to ISIS database files which include
101  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
102    C<read_fdt> option) files.
103    
104  In this example it uses C<./cds/cds.MST> and related files.  In this example it uses C<./cds/cds.MST> and related files.
105    
# Line 134  sub new { Line 137  sub new {
137                  $self->{$v} = {@_}->{$v};                  $self->{$v} = {@_}->{$v};
138          }          }
139    
140            my @isis_files = grep(/\.(FDT|MST|XRF|CNT)$/i,glob($self->{isisdb}."*"));
141    
142            foreach my $f (@isis_files) {
143                    my $ext = $1 if ($f =~ m/\.(\w\w\w)$/);
144                    $self->{lc($ext)."_file"} = $f;
145            }
146    
147            my @must_exist = qw(mst xrf);
148            push @must_exist, "fdt" if ($self->{read_fdt});
149    
150            foreach my $ext (@must_exist) {
151                    croak "missing ",uc($ext)," file in ",$self->{isisdb} unless ($self->{$ext."_file"});
152            }
153    
154            print STDERR "## using files: ",join(" ",@isis_files),"\n" if ($self->{debug});
155    
156          # if you want to read .FDT file use read_fdt argument when creating class!          # if you want to read .FDT file use read_fdt argument when creating class!
157          if ({@_}->{read_fdt} && -e $self->{isisdb}.".FDT") {          if ($self->{read_fdt} && -e $self->{fdt_file}) {
158    
159                  # read the $db.FDT file for tags                  # read the $db.FDT file for tags
160                  my $fieldzone=0;                  my $fieldzone=0;
161    
162                  open(fileFDT, $self->{isisdb}.".FDT") || croak "can't read '$self->{isisdb}.FDT': $!";                  open(fileFDT, $self->{fdt_file}) || croak "can't read '$self->{fdt_file}': $!";
163    
164                  while (<fileFDT>) {                  while (<fileFDT>) {
165                          chomp;                          chomp;
# Line 164  sub new { Line 183  sub new {
183    
184          # Get the Maximum MFN from $db.MST          # Get the Maximum MFN from $db.MST
185    
186          open(fileMST,$self->{isisdb}.".MST") || croak "can't read '$self->{isisdb}.MST': $!";          open($self->{'fileMST'}, $self->{mst_file}) || croak "can't open '$self->{mst_file}': $!";
187    
188          # MST format:   (* = 32 bit signed)          # MST format:   (* = 32 bit signed)
189          # CTLMFN*       always 0          # CTLMFN*       always 0
# Line 172  sub new { Line 191  sub new {
191          # NXTMFB*       last block allocated to master file          # NXTMFB*       last block allocated to master file
192          # NXTMFP        offset to next available position in last block          # NXTMFP        offset to next available position in last block
193          # MFTYPE        always 0 for user db file (1 for system)          # MFTYPE        always 0 for user db file (1 for system)
194          seek(fileMST,4,0);          seek($self->{'fileMST'},4,0);
195    
196          my $buff;          my $buff;
197    
198          read(fileMST, $buff, 4);          read($self->{'fileMST'}, $buff, 4);
199          $self->{'NXTMFN'}=unpack("l",$buff) || carp "NXTNFN is zero";          $self->{'NXTMFN'}=unpack("l",$buff) || carp "NXTNFN is zero";
200    
201          # save maximum MFN          # save maximum MFN
202          $self->{'maxmfn'} = $self->{'NXTMFN'} - 1;          $self->{'maxmfn'} = $self->{'NXTMFN'} - 1;
203    
204          close(fileMST);  
205    
206    
207            print STDERR Dumper($self),"\n" if ($self->{debug});
208    
209            # open files for later
210            open($self->{'fileXRF'}, $self->{xrf_file}) || croak "can't open '$self->{xrf_file}': $!";
211    
212            $self ? return $self : return undef;
213    }
214    
215    =head2 read_cnt
216    
217    This function is not really used by module, but can be useful to find info
218    about your index (if debugging it for example).
219    
220      print Dumper($isis->read_cnt);
221    
222    =cut
223    
224    sub read_cnt  {
225            my $self = shift;
226    
227            croak "missing CNT file in ",$self->{isisdb} unless ($self->{cnt_file});
228    
229          # Get the index information from $db.CNT          # Get the index information from $db.CNT
230        
231          open(fileCNT, $self->{isisdb}.".CNT") || croak "can't read '$self->{isisdb}.CNT': $!";          open(fileCNT, $self->{cnt_file}) || croak "can't read '$self->{cnt_file}': $!";
232    
233          # There is two 26 Bytes fixed lenght records          # There is two 26 Bytes fixed lenght records
234    
# Line 210  sub new { Line 252  sub new {
252                  my $buff = shift || return;                  my $buff = shift || return;
253                  my @arr = unpack("ssssssllls", $buff);                  my @arr = unpack("ssssssllls", $buff);
254    
255                  print "unpack_cnt: ",join(" ",@arr),"\n" if ($self->{'debug'});                  print STDERR "unpack_cnt: ",join(" ",@arr),"\n" if ($self->{'debug'});
256    
257                  my $IDTYPE = shift @arr;                  my $IDTYPE = shift @arr;
258                  foreach (@flds) {                  foreach (@flds) {
259                          $self->{$IDTYPE}->{$_} = abs(shift @arr);                          $self->{cnt}->{$IDTYPE}->{$_} = abs(shift @arr);
260                  }                  }
261          }          }
262    
263            my $buff;
264    
265          read(fileCNT, $buff, 26);          read(fileCNT, $buff, 26);
266          $self->unpack_cnt($buff);          $self->unpack_cnt($buff);
267    
268          read(fileCNT, $buff, 26);          read(fileCNT, $buff, 26);
269          $self->unpack_cnt($buff);          $self->unpack_cnt($buff);
270    
   
271          close(fileCNT);          close(fileCNT);
272    
273          print Dumper($self),"\n" if ($self->{debug});          return $self->{cnt};
   
         # open files for later  
         open($self->{'fileXRF'}, $self->{isisdb}.".XRF") || croak "can't open '$self->{isisdb}.XRF': $!";  
   
         open($self->{'fileMST'}, $self->{isisdb}.".MST") || croak "can't open '$self->{isisdb}.MST': $!";  
   
         $self ? return $self : return undef;  
274  }  }
275    
276  =head2 fetch  =head2 fetch
# Line 258  sub fetch { Line 294  sub fetch {
294    
295          my $mfn = shift || croak "fetch needs MFN as argument!";          my $mfn = shift || croak "fetch needs MFN as argument!";
296    
297          print "fetch: $mfn\n" if ($self->{debug});          # is mfn allready in memory?
298            my $old_mfn = $self->{'current_mfn'} || -1;
299            return $self->{record} if ($mfn == $old_mfn);
300    
301            print STDERR "## fetch: $mfn\n" if ($self->{debug});
302    
303          # XXX check this?          # XXX check this?
304          my $mfnpos=($mfn+int(($mfn-1)/127))*4;          my $mfnpos=($mfn+int(($mfn-1)/127))*4;
305    
306          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 ($self->{debug});
307          seek($self->{'fileXRF'},$mfnpos,0);          seek($self->{'fileXRF'},$mfnpos,0);
308    
309          my $buff;          my $buff;
310    
311            # delete old record
312            delete $self->{record};
313    
314          # read XRFMFB abd XRFMFP          # read XRFMFB abd XRFMFP
315          read($self->{'fileXRF'}, $buff, 4);          read($self->{'fileXRF'}, $buff, 4);
316          my $pointer=unpack("l",$buff) || carp "pointer is null";          my $pointer=unpack("l",$buff) || carp "pointer is null";
317    
318            # check for logically deleted record
319            if ($pointer < 0) {
320                    print STDERR "## record $mfn is logically deleted\n" if ($self->{debug});
321                    $self->{deleted} = $mfn;
322    
323                    return unless $self->{include_deleted};
324    
325                    $pointer = abs($pointer);
326            }
327    
328          my $XRFMFB = int($pointer/2048);          my $XRFMFB = int($pointer/2048);
329          my $XRFMFP = $pointer - ($XRFMFB*2048);          my $XRFMFP = $pointer - ($XRFMFB*2048);
330    
331          print "XRFMFB: $XRFMFB XRFMFP: $XRFMFP\n" if ($self->{debug});          # (XRFMFB - 1) * 512 + XRFMFP
332            # why do i have to do XRFMFP % 1024 ?
         # XXX fix this to be more readable!!  
         # e.g. (XRFMFB - 1) * 512 + XRFMFP  
333    
334          my $offset = $pointer;          my $blk_off = (($XRFMFB - 1) * 512) + ($XRFMFP % 1024);
         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)+$offset3;  
335    
336          print "$offset - $offset2 - $offset3 - $offset4\n" if ($self->{debug});          print STDERR "## pointer: $pointer XRFMFB: $XRFMFB XRFMFP: $XRFMFP offset: $blk_off\n" if ($self->{'debug'});
337    
338          # Get Record Information          # Get Record Information
339    
340          seek($self->{'fileMST'},$offset4,0);          seek($self->{'fileMST'},$blk_off,0);
341    
342          read($self->{'fileMST'}, $buff, 4);          read($self->{'fileMST'}, $buff, 4);
343          my $value=unpack("l",$buff);          my $value=unpack("l",$buff);
344    
345            print STDERR "## offset for rowid $value is $blk_off (blk $XRFMFB off $XRFMFP)\n" if ($self->{debug});
346    
347          if ($value!=$mfn) {          if ($value!=$mfn) {
348  print ("Error: The MFN:".$mfn." is not found in MST(".$value.")");                      carp "Error: MFN ".$mfn." not found in MST(".$value.")";    
349                  return -1;      # XXX deleted record?                  #return;
350          }          }
351    
 #       $MFRL=$self->Read16($fileMST);  
 #       $MFBWB=$self->Read32($fileMST);  
 #       $MFBWP=$self->Read16($fileMST);  
 #       $BASE=$self->Read16($fileMST);  
 #       $NVF=$self->Read16($fileMST);  
 #       $STATUS=$self->Read16($fileMST);  
   
352          read($self->{'fileMST'}, $buff, 14);          read($self->{'fileMST'}, $buff, 14);
353    
354          my ($MFRL,$MFBWB,$MFBWP,$BASE,$NVF,$STATUS) = unpack("slssss", $buff);          my ($MFRL,$MFBWB,$MFBWP,$BASE,$NVF,$STATUS) = unpack("slssss", $buff);
355    
356          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});
357    
358          # delete old record          warn "MFRL $MFRL is not even number" unless ($MFRL % 2 == 0);
         delete $self->{record};  
359    
360          if (! $self->{'include_deleted'} && $MFRL < 0) {          warn "BASE is not 18+6*NVF" unless ($BASE == 18 + 6 * $NVF);
                 print "## logically deleted record $mfn, skipping...\n" if ($self->{debug});  
                 return;  
         }  
361    
362          # Get Directory Format          # Get Directory Format
363    
# Line 332  print ("Error: The MFN:".$mfn." is not f Line 367  print ("Error: The MFN:".$mfn." is not f
367    
368          read($self->{'fileMST'}, $buff, 6 * $NVF);          read($self->{'fileMST'}, $buff, 6 * $NVF);
369    
370          my $fld_len = 0;          my $rec_len = 0;
371    
372          for (my $i = 0 ; $i < $NVF ; $i++) {          for (my $i = 0 ; $i < $NVF ; $i++) {
373    
 #               $TAG=$self->Read16($fileMST);  
 #               $POS=$self->Read16($fileMST);  
 #               $LEN=$self->Read16($fileMST);  
   
374                  my ($TAG,$POS,$LEN) = unpack("sss", substr($buff,$i * 6, 6));                  my ($TAG,$POS,$LEN) = unpack("sss", substr($buff,$i * 6, 6));
375    
376                  print "TAG: $TAG POS: $POS LEN: $LEN\n" if ($self->{debug});                  print STDERR "## TAG: $TAG POS: $POS LEN: $LEN\n" if ($self->{debug});
377    
378                  # The TAG does not exists in .FDT so we set it to 0.                  # The TAG does not exists in .FDT so we set it to 0.
379                  #                  #
# Line 358  print ("Error: The MFN:".$mfn." is not f Line 389  print ("Error: The MFN:".$mfn." is not f
389                  push @FieldPOS,$POS;                  push @FieldPOS,$POS;
390                  push @FieldLEN,$LEN;                  push @FieldLEN,$LEN;
391    
392                  $fld_len += $LEN;                  $rec_len += $LEN;
393          }          }
394    
395          # Get Variable Fields          # Get Variable Fields
396    
397          read($self->{'fileMST'},$buff,$fld_len);          read($self->{'fileMST'},$buff,$rec_len);
398    
399            print STDERR "## rec_len: $rec_len poc: ",tell($self->{'fileMST'})."\n" if ($self->{debug});
400    
401          for (my $i = 0 ; $i < $NVF ; $i++) {          for (my $i = 0 ; $i < $NVF ; $i++) {
402                  # skip zero-sized fields                  # skip zero-sized fields
# Line 371  print ("Error: The MFN:".$mfn." is not f Line 404  print ("Error: The MFN:".$mfn." is not f
404    
405                  push @{$self->{record}->{$FieldTAG[$i]}}, substr($buff,$FieldPOS[$i],$FieldLEN[$i]);                  push @{$self->{record}->{$FieldTAG[$i]}}, substr($buff,$FieldPOS[$i],$FieldLEN[$i]);
406          }          }
         close(fileMST);  
407    
408          print Dumper($self),"\n" if ($self->{debug});          $self->{'current_mfn'} = $mfn;
409    
410            print STDERR Dumper($self),"\n" if ($self->{debug});
411    
412          return $self->{'record'};          return $self->{'record'};
413  }  }
# Line 467  sub to_hash { Line 501  sub to_hash {
501          my $mfn = shift || confess "need mfn!";          my $mfn = shift || confess "need mfn!";
502    
503          # init record to include MFN as field 000          # init record to include MFN as field 000
504          my $rec = { '000' => $mfn };          my $rec = { '000' => [ $mfn ] };
505    
506          my $row = $self->fetch($mfn);          my $row = $self->fetch($mfn);
507    
# Line 480  sub to_hash { Line 514  sub to_hash {
514                          my $val;                          my $val;
515    
516                          # has identifiers?                          # has identifiers?
517                          ($val->{'i1'},$val->{'i2'}) = ($1,$2) if ($l =~ s/^([01 #])([01 #])//);                          ($val->{'i1'},$val->{'i2'}) = ($1,$2) if ($l =~ s/^([01 #])([01 #])\^/\^/);
518    
519                          # has subfields?                          # has subfields?
520                          if ($l =~ m/\^/) {                          if ($l =~ m/\^/) {
Legend:
Removed from v.15  
changed lines
  Added in v.25
  ViewVC Help
Powered by ViewVC 1.1.26