/[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 3 by dpavlin, Tue Dec 28 01:48:44 2004 UTC revision 19 by dpavlin, Thu Dec 30 23:16:20 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     = 0.01;          $VERSION     = 0.07;
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 18  BEGIN { Line 20  BEGIN {
20    
21  =head1 NAME  =head1 NAME
22    
23  IsisDB - Read CDS/ISIS database  IsisDB - Read CDS/ISIS, WinISIS and IsisMarc database
24    
25  =head1 SYNOPSIS  =head1 SYNOPSIS
26    
27    use IsisDB    use IsisDB;
28    
29    my $isis = new IsisDB(    my $isis = new IsisDB(
30          isisdb => './cds/cds',          isisdb => './cds/cds',
31    );    );
32    
33      for(my $mfn = 1; $mfn <= $isis->{'maxmfn'}; $mfn++) {
34            print $isis->to_ascii($mfn),"\n";
35      }
36    
37  =head1 DESCRIPTION  =head1 DESCRIPTION
38    
39  This module will read CDS/ISIS databases and create hash values out of it.  This module will read ISIS databases created by DOS CDS/ISIS, WinIsis or
40  It can be used as perl-only alternative to OpenIsis module.  IsisMarc. It can be used as perl-only alternative to OpenIsis module.
41    
42    It can create hash values from data in ISIS database (using C<to_hash>),
43    ASCII dump (using C<to_ascii>) or just hash with field names and packed
44    values (like C<^asomething^belse>).
45    
46    Unique feature of this module is ability to C<include_deleted> records.
47    It will also skip zero sized fields (OpenIsis has a bug in XS bindings, so
48    fields which are zero sized will be filled with random junk from memory).
49    
50    It also has support for identifiers (only if ISIS database is created by
51    IsisMarc), see C<to_hash>.
52    
53    This will module will always be slower than OpenIsis module which use C
54    library. However, since it's written in perl, it's platform independent (so
55    you don't need C compiler), and can be easily modified. I hope that it
56    creates data structures which are easier to use than ones created by
57    OpenIsis, so reduced time in other parts of the code should compensate for
58    slower performance of this module (speed of reading ISIS database is
59    rarely an issue).
60    
61  =head1 METHODS  =head1 METHODS
62    
# Line 50  It can be used as perl-only alternative Line 76  It can be used as perl-only alternative
76  # some binary reads  # some binary reads
77  #  #
78    
 sub Read32 {  
         my $self = shift;  
   
         my $f = shift || die "Read32 needs file handle";  
         read($$f,$b,4) || die "can't read 4 bytes from $$f from position ".tell($f);  
         return unpack("l",$b);  
 }  
   
79  =head2 new  =head2 new
80    
81  Open CDS/ISIS database  Open ISIS database
82    
83   my $isis = new IsisDB(   my $isis = new IsisDB(
84          isisdb => './cds/cds',          isisdb => './cds/cds',
85          read_fdt => 1,          read_fdt => 1,
86            include_deleted => 1,
87            hash_filter => sub {
88                    my $v = shift;
89                    $v =~ s#foo#bar#g;
90            },
91          debug => 1,          debug => 1,
92   );   );
93    
# Line 74  Options are described below: Line 97  Options are described below:
97    
98  =item isisdb  =item isisdb
99    
100  Prefix path to CDS/ISIS. It should contain full or relative path to database  This is full or relative path to ISIS database files which include
101  and 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.
105    
106  =item read_fdt  =item read_fdt
107    
108  Boolean flag to specify if field definition table should be read. It's off  Boolean flag to specify if field definition table should be read. It's off
109  by default.  by default.
110    
111    =item include_deleted
112    
113    Don't skip logically deleted records in ISIS.
114    
115    =item hash_filter
116    
117    Filter code ref which will be used before data is converted to hash.
118    
119  =item debug  =item debug
120    
121  Dump a C<lot> of debugging output.  Dump a B<lot> of debugging output.
122    
123  =back  =back
124    
# Line 97  sub new { Line 131  sub new {
131          my $self = {};          my $self = {};
132          bless($self, $class);          bless($self, $class);
133    
134          $self->{isisdb} = {@_}->{isisdb} || croak "new needs database name as argument!";          croak "new needs database name (isisdb) as argument!" unless ({@_}->{isisdb});
135    
136            foreach my $v (qw{isisdb debug include_deleted hash_filter}) {
137                    $self->{$v} = {@_}->{$v};
138            }
139    
140          $self->{debug} = {@_}->{debug};          my @isis_files = grep(/\.(FDT|MST|XRF|CNT)$/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 131  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}': $!";
187    
188          # MST format:   (* = 32 bit signed)          # MST format:   (* = 32 bit signed)
189          # CTLMFN*       always 0          # CTLMFN*       always 0
# Line 139  sub new { Line 191  sub new {
191          # NXTMFB*       last block allocated to master file          # NXTMFB*       last block allocated to master file
192          # NXTMFP        offset to next available position in last block          # NXTMFP        offset to next available position in last block
193          # MFTYPE        always 0 for user db file (1 for system)          # MFTYPE        always 0 for user db file (1 for system)
194          seek(fileMST,4,0);          seek($self->{'fileMST'},4,0);
195          $self->{'NXTMFN'}=$self->Read32(\*fileMST) || carp "NXTNFN is zero";  
196            my $buff;
197    
198            read($self->{'fileMST'}, $buff, 4);
199            $self->{'NXTMFN'}=unpack("l",$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 173  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", $buff);
254    
255                    print STDERR "unpack_cnt: ",join(" ",@arr),"\n" if ($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;          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($self) if ($self->{debug});          return $self->{cnt};
   
         $self ? return $self : return undef;  
274  }  }
275    
276  =head2 GetMFN  =head2 fetch
277    
278  Read record with selected MFN  Read record with selected MFN
279    
280    my $rec = $isis->GetMFN(55);    my $rec = $isis->fetch(55);
281    
282  Returns hash with keys which are field names and values are unpacked values  Returns hash with keys which are field names and values are unpacked values
283  for that field.  for that field like this:
284    
285      $rec = {
286        '210' => [ '^aNew York^cNew York University press^dcop. 1988' ],
287        '990' => [ '2140', '88', 'HAY' ],
288      };
289    
290  =cut  =cut
291    
292  sub GetMFN {  sub fetch {
293          my $self = shift;          my $self = shift;
294    
295          my $mfn = shift || croak "GetMFN needs MFN as argument!";          my $mfn = shift || croak "fetch needs MFN as argument!";
296    
297          print "GetMFN: $mfn\n" if ($self->{debug});          # is mfn allready in memory?
298            my $old_mfn = $self->{'current_mfn'} || -1;
299            return if ($mfn == $old_mfn);
300    
301          open(fileXRF, $self->{isisdb}.".XRF") || croak "can't open '$self->{isisdb}.XRF': $!";          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 ($self->{debug});
307          seek(fileXRF,$mfnpos,0);          seek($self->{'fileXRF'},$mfnpos,0);
308    
309            my $buff;
310    
311          # read XRFMFB abd XRFMFP          # read XRFMFB abd XRFMFP
312          my $pointer=$self->Read32(\*fileXRF);          read($self->{'fileXRF'}, $buff, 4);
313            my $pointer=unpack("l",$buff) || carp "pointer is null";
314    
315          my $XRFMFB = int($pointer/2048);          my $XRFMFB = int($pointer/2048);
316          my $XRFMFP = $pointer - ($XRFMFB*2048);          my $XRFMFP = $pointer - ($XRFMFB*2048);
317    
         print "XRFMFB: $XRFMFB XRFMFP: $XRFMFP\n" if ($self->{debug});  
   
         # XXX fix this to be more readable!!  
         # e.g. (XRFMFB - 1) * 512 + XRFMFP  
318    
319          my $offset = $pointer;          # (XRFMFB - 1) * 512 + XRFMFP
320          my $offset2=int($offset/2048)-1;          # why do i have to do XRFMFP % 1024 ?
         my $offset22=int($offset/4096);  
         my $offset3=$offset-($offset22*4096);  
         if ($offset3>512) {  
                 $offset3=$offset3-2048;  
         }  
         my $offset4=($offset2*512)+$offset3;  
321    
322          print "$offset - $offset2 - $offset3 - $offset4\n" if ($self->{debug});          my $blk_off = (($XRFMFB - 1) * 512) + ($XRFMFP % 1024);
323    
324          close(fileXRF);          print STDERR "## pointer: $pointer XRFMFB: $XRFMFB XRFMFP: $XRFMFP offset: $blk_off\n" if ($self->{'debug'});
325    
326          # Get Record Information          # Get Record Information
327    
328          open(fileMST, $self->{isisdb}.".MST") || croak "can't open '$self->{isisdb}.MST': $!";          seek($self->{'fileMST'},$blk_off,0);
329    
330          seek(fileMST,$offset4,0);          read($self->{'fileMST'}, $buff, 4);
331            my $value=unpack("l",$buff);
332    
333          my $value=$self->Read32(\*fileMST);          print STDERR "## offset for rowid $value is $blk_off (blk $XRFMFB off $XRFMFP)\n" if ($self->{debug});
334    
335          if ($value!=$mfn) {          if ($value!=$mfn) {
336  print ("Error: The MFN:".$mfn." is not found in MST(".$value.")");                      carp "Error: MFN ".$mfn." not found in MST(".$value.")";    
337                  return -1;      # XXX deleted record?                  #return;                # XXX deleted record?
338          }          }
339    
340  #       $MFRL=$self->Read16($fileMST);  #       $MFRL=$self->Read16($fileMST);
# Line 264  print ("Error: The MFN:".$mfn." is not f Line 344  print ("Error: The MFN:".$mfn." is not f
344  #       $NVF=$self->Read16($fileMST);  #       $NVF=$self->Read16($fileMST);
345  #       $STATUS=$self->Read16($fileMST);  #       $STATUS=$self->Read16($fileMST);
346    
347          my $buff;          read($self->{'fileMST'}, $buff, 14);
         read(fileMST, $buff, 14);  
348    
349          my ($MFRL,$MFBWB,$MFBWP,$BASE,$NVF,$STATUS) = unpack("slssss", $buff);          my ($MFRL,$MFBWB,$MFBWP,$BASE,$NVF,$STATUS) = unpack("slssss", $buff);
350    
351          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});
352    
353            # delete old record
354            delete $self->{record};
355    
356            ## FIXME this is a bug
357            if (! $self->{'include_deleted'} && $MFRL < 0) {
358                    print "## logically deleted record $mfn, skipping...\n" if ($self->{debug});
359                    return;
360            }
361    
362            warn "BASE is not 18+6*NVF" unless ($BASE == 18 + 6 * $NVF);
363    
364          # Get Directory Format          # Get Directory Format
365    
# Line 277  print ("Error: The MFN:".$mfn." is not f Line 367  print ("Error: The MFN:".$mfn." is not f
367          my @FieldLEN;          my @FieldLEN;
368          my @FieldTAG;          my @FieldTAG;
369    
370            read($self->{'fileMST'}, $buff, 6 * $NVF);
371    
372            my $rec_len = 0;
373    
374          for (my $i = 0 ; $i < $NVF ; $i++) {          for (my $i = 0 ; $i < $NVF ; $i++) {
375    
376  #               $TAG=$self->Read16($fileMST);  #               $TAG=$self->Read16($fileMST);
377  #               $POS=$self->Read16($fileMST);  #               $POS=$self->Read16($fileMST);
378  #               $LEN=$self->Read16($fileMST);  #               $LEN=$self->Read16($fileMST);
379    
380                  read(fileMST, $buff, 6);                  my ($TAG,$POS,$LEN) = unpack("sss", substr($buff,$i * 6, 6));
                 my ($TAG,$POS,$LEN) = unpack("sss", $buff);  
381    
382                  print "TAG: $TAG POS: $POS LEN: $LEN\n" if ($self->{debug});                  print STDERR "## TAG: $TAG POS: $POS LEN: $LEN\n" if ($self->{debug});
383    
384                  # 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.
385                  #                  #
# Line 301  print ("Error: The MFN:".$mfn." is not f Line 394  print ("Error: The MFN:".$mfn." is not f
394                  push @FieldTAG,$TAG;                  push @FieldTAG,$TAG;
395                  push @FieldPOS,$POS;                  push @FieldPOS,$POS;
396                  push @FieldLEN,$LEN;                  push @FieldLEN,$LEN;
397    
398                    $rec_len += $LEN;
399          }          }
400    
401          # Get Variable Fields          # Get Variable Fields
402    
403          delete $self->{record};          read($self->{'fileMST'},$buff,$rec_len);
404    
405            print STDERR "## rec_len: $rec_len poc: ",tell($self->{'fileMST'})."\n" if ($self->{debug});
406    
407          for (my $i = 0 ; $i < $NVF ; $i++) {          for (my $i = 0 ; $i < $NVF ; $i++) {
408                  my $rec;                  # skip zero-sized fields
409                  read(fileMST,$rec,$FieldLEN[$i]);                  next if ($FieldLEN[$i] == 0);
                 push @{$self->{record}->{$FieldTAG[$i]}}, $rec;  
         }  
         close(fileMST);  
410    
411          # The record is marked for deletion                  push @{$self->{record}->{$FieldTAG[$i]}}, substr($buff,$FieldPOS[$i],$FieldLEN[$i]);
         if ($STATUS==1) {  
                 return -1;  
412          }          }
413    
414          print Dumper($self) if ($self->{debug});          $self->{'current_mfn'} = $mfn;
415    
416            print Dumper($self),"\n" if ($self->{debug});
417    
418          return $self->{'record'};          return $self->{'record'};
419  }  }
420    
421  =head2 to_ascii  =head2 to_ascii
422    
423  Dump ascii output of selected MFN  Dump ASCII output of record with specified MFN
424    
425      print $isis->to_ascii(42);
426    
427    It outputs something like this:
428    
429    print $isis->to_ascii(55);    210   ^aNew York^cNew York University press^dcop. 1988
430      990   2140
431      990   88
432      990   HAY
433    
434    If C<read_fdt> is specified when calling C<new> it will display field names
435    from C<.FDT> file instead of numeric tags.
436    
437  =cut  =cut
438    
# Line 337  sub to_ascii { Line 441  sub to_ascii {
441    
442          my $mfn = shift || croak "need MFN";          my $mfn = shift || croak "need MFN";
443    
444          my $rec = $self->GetMFN($mfn);          my $rec = $self->fetch($mfn);
445    
446          my $out = "0\t$mfn";          my $out = "0\t$mfn";
447    
448          foreach my $f (sort keys %{$rec}) {          foreach my $f (sort keys %{$rec}) {
449                  $out .= "\n$f\t".join("\n$f\t",@{$self->{record}->{$f}});                  my $fn = $self->tag_name($f);
450                    $out .= "\n$fn\t".join("\n$fn\t",@{$self->{record}->{$f}});
451          }          }
452    
453          $out .= "\n";          $out .= "\n";
# Line 350  sub to_ascii { Line 455  sub to_ascii {
455          return $out;          return $out;
456  }  }
457    
458  ################# old cruft which is not ported from php to perl  =head2 to_hash
   
 =begin php  
459    
460    # Load the dictionary from the $db.L0x files.  Read record with specified MFN and convert it to hash
   # Not usefull Yet  
     
   sub LoadDictionary()  
   {  
     $fileL01=fopen($self->{isisdb}.".L01","r");  
     rewind($fileL01);    
   
     do  
     {  
   
       $POS=$self->Read32($fileL01);  
       $OCK=$self->Read16($fileL01);  
       $IT=$self->Read16($fileL01);  
       $PS=$self->Read32($fileL01);  
 print "<br>PS:".$PS." ".$self->{ORDF}->{1}." ";  
       for ($i=0;$i<$OCK;$i++)  
       {  
         $KEY=fread($fileL01,10);  
         
         print $KEY." ### ";  
   
         $INFO1=$self->Read32($fileL01);  
         $INFO2=$self->Read32($fileL01);  
   
         #L01Key->{$key}=array($INFO1,$INFO2);  
       }  
       
       rewind($fileL01);  
       $offset=($PS-1)*(12+$self->{ORDF}->{1}*18*2);  
       fseek($fileL01,$offset);  
461    
462      } While (!feof($fileL01));    my $hash = $isis->to_hash($mfn);
   
     fclose($fileL01);  
   }  
   
   # self function search through the tree and returns an array of pointers to IFP  
   # The function must be recursive  
   
   sub SearchTree($search,$fileNB,$PUNT)  
   {        
       $offset=(($PUNT-1)*(8+2*$self->{ORDN}->{1}*14));  
   
         rewind($fileNB1);  
   
         fseek($fileNB,$offset);  
   
         $POS=$self->Read32($fileNB);  
         $OCK=$self->Read16($fileNB);  
         $IT=$self->Read16($fileNB);  
   
 #print "<br>".$POS." - ".$OCK." - ".$IT;  
   
         $OLDPUNT=$POS;  
         $j=0;  
         for ($i=0;$i<$OCK;$i++)  
         {  
           $KEY=fread($fileNB,10);  
         
           $PUNT=$self->Read32($fileNB);  
   
 #print " ## ".chop($KEY)."(".$PUNT."-".$OLDPUNT.") ## ";  
   
           If (strcmp($search,chop($KEY))<0)  
           {  
             break;  
           }  
           $OLDPUNT=$PUNT;    
         }          
 #print $OLDPUNT;  
         Return $OLDPUNT;  
   }  
463    
464    # Search ISIS for record containing search  It has ability to convert characters (using C<hash_filter> from ISIS
465    # Return a sorted array of MFN  database before creating structures enabling character re-mapping or quick
466    fix-up of data.
467    
468    This function returns hash which is like this:
469    
470      $hash = {
471        '210' => [
472                   {
473                     'c' => 'New York University press',
474                     'a' => 'New York',
475                     'd' => 'cop. 1988'
476                   }
477                 ],
478        '990' => [
479                   '2140',
480                   '88',
481                   'HAY'
482                 ],
483      };
484    
485    You can later use that hash to produce any output from ISIS data.
486    
487    If database is created using IsisMarc, it will also have to special fields
488    which will be used for identifiers, C<i1> and C<i2> like this:
489    
490      '200' => [
491                 {
492                   'i1' => '1',
493                   'i2' => ' '
494                   'a' => 'Goa',
495                   'f' => 'Valdo D\'Arienzo',
496                   'e' => 'tipografie e tipografi nel XVI secolo',
497                 }
498               ],
499    
500    sub Search($search)  This method will also create additional field C<000> with MFN.
   {  
   
   $search=strtoupper($search);  
 #print "Searching....".$search." - ".$self->{POSRX}->{1}."<br>";  
     # first search .x01  
       
   
     # Search in .N01    
   
   
     $fileN01=fopen($self->{isisdb}.".N01","r");  
     $offset=(($self->{POSRX}->{1}-1)*(8+2*$self->{ORDN}->{1}*14));  
   
       do  
       {  
         rewind($fileN01);  
   
         fseek($fileN01,$offset);  
   
         $POS=$self->Read32($fileN01);  
         $OCK=$self->Read16($fileN01);  
         $IT=$self->Read16($fileN01);  
   
 #print "<br>".$POS." - ".$OCK." - ".$IT;  
   
         $OLDPUNT=$POS;  
         for ($i=0;$i<$OCK;$i++)  
         {  
           $KEY=fread($fileN01,10);  
         
           $PUNT=$self->Read32($fileN01);  
   
 #print " ## ".chop($KEY)."(".$PUNT."-".$OLDPUNT.") ## ";  
   
           If (strcmp($search,chop($KEY))<0)  
           {  
             break;  
           }  
           $OLDPUNT=$PUNT;    
         }  
         $offset=(($OLDPUNT-1)*(8+2*$self->{ORDN}->{1}*14));        
       } while ($OLDPUNT>0);  
 #print $OLDPUNT;  
   
   
     fclose($fileN01);  
   
     # Now look for records in .L01 file  
     $fileL01=fopen($self->{isisdb}.".L01","r");  
     rewind($fileL01);  
   
     $offset=(-$OLDPUNT-1)*(12+$self->{ORDF}->{1}*18*2);  
     fseek($fileL01,$offset);  
   
     $POS=$self->Read32($fileL01);  
     $OCK=$self->Read16($fileL01);  
     $IT=$self->Read16($fileL01);  
     $PS=$self->Read32($fileL01);  
 #print "<br>POS:".$POS." ".$self->{ORDF}->{1}." ";  
     for ($i=0;$i<$OCK;$i++)  
     {  
       $KEY=fread($fileL01,10);  
         
 #print $KEY." ### ";  
   
       $INFO1=$self->Read32($fileL01);  
       $INFO2=$self->Read32($fileL01);  
   
       If (strcmp($search,chop($KEY))==0)  
       {  
         break;  
       }  
     }      
   
     fclose($fileL01);  
   
 #print $INFO1."--".$INFO2;  
   
     # Now look in .IFP for the MFN  
     $fileIFP=fopen($self->{isisdb}.".IFP","r");  
     rewind($fileIFP);  
     $offset=($INFO1-1)*512+($INFO2*4);  
     fseek($fileIFP,$offset);    
   
     $IFPBLK=$self->Read32($fileIFP);  
   
     $IFPNXTB=$self->Read32($fileIFP);  
     $IFPNXTP=$self->Read32($fileIFP);  
     $IFPTOTP=$self->Read32($fileIFP);  
     $IFPSEGP=$self->Read32($fileIFP);  
     $IFPSEGC=$self->Read32($fileIFP);  
   
   
 #print "<br>IFP:".$IFPBLK." # ".$IFPNXTB." - ".$IFPNXTP." - ".$IFPTOTP." - ".$IFPSEGP." - ".$IFPSEGC;  
   
     rewind($fileIFP);  
     $offset=($INFO1-1)*512+24+($INFO2*4);  
     fseek($fileIFP,$offset);      
       
     $j=24+($INFO2*4);  
     $k=0;  
     $l=1;  
     $OLDPMFN="";  
     for ($i=0;$i<$IFPSEGP;$i++)  
     {  
       $B1=$self->Read8($fileIFP);  
       $B2=$self->Read8($fileIFP);  
       $B3=$self->Read8($fileIFP);  
       $B4=$self->Read8($fileIFP);  
       $B5=$self->Read8($fileIFP);  
       $B6=$self->Read8($fileIFP);  
       $B7=$self->Read8($fileIFP);  
       $B8=$self->Read8($fileIFP);  
   
       $PMFN=$B1*65536+$B2*256+$B3;  
       $PTAG=$B4*256+$B5;  
       $POCC=$B6;  
       $PCNT=$B7*256+$B8;  
   
       if ($OLDPMFN!=$PMFN)  
       {  
         if ($PMFN!=0)  
         {  
           $self->{MFNArray}->{$l}=$PMFN;  
           $OLDPMFN=$PMFN;  
           $l+=1;  
         }  
       }  
   
       $j=$j+8;  
 #print "<br>".$PMFN."-".$PTAG." - ".$POCC." - ".$PCNT;  
 #print "@@".$j."@@@@";  
       if ($j>=504)  
       {  
         if ($IFPNXTB==0 && $IFPNXTP==0)  
         {  
           $k=$k+1;  
           rewind($fileIFP);  
           $offset=($INFO1-1+$k)*512;    
           fseek($fileIFP,$offset);        
           $B=$self->Read32($fileIFP);  
 #print "<br>-".$B."-<br>";  
           $j=0;  
         } else  
         {  
           rewind($fileIFP);  
           $offset=($IFPNXTB-1)*512;    
           fseek($fileIFP,$offset);  
   
           $OLDIFPNXTB=$IFPNXTB;  
           $OLDIFPNXTP=$IFPNXTP;  
   
           $IFPBLK=$self->Read32($fileIFP);  
   
           $IFPNXTB=$self->Read32($fileIFP);  
           $IFPNXTP=$self->Read32($fileIFP);  
           $IFPTOTP=$self->Read32($fileIFP);  
           $IFPSEGP=$self->Read32($fileIFP);  
           $IFPSEGC=$self->Read32($fileIFP);  
   
           rewind($fileIFP);  
           $offset=($OLDIFPNXTB-1)*512+24+($OLDIFPNXTP*4);  
           fseek($fileIFP,$offset);      
       
           $j=24+($OLDIFPNXTP*4);  
           $k=0;  
           $j=0;  
         }  
       }  
   
     }      
     fclose($fileIFP);  
     return $l-1;  
   }  
501    
502  =cut  =cut
503    
504  #  sub to_hash {
 # XXX porting from php left-over:  
 #  
 # do I *REALLY* need those methods, or should I use  
 # $self->{something} directly?  
 #  
 # Probably direct usage is better!  
 #  
   
 sub GetFieldName {  
505          my $self = shift;          my $self = shift;
         return $self->{FieldName};  
 }  
506    
507  sub GetTagName {          my $mfn = shift || confess "need mfn!";
         my $self = shift;  
         return $self->{TagName};  
 }  
508    
509  sub GetFieldTag {          # init record to include MFN as field 000
510          my $self = shift;          my $rec = { '000' => [ $mfn ] };
         return $self->{FieldTAG};  
 }  
511    
512  sub GetNextMFN {          my $row = $self->fetch($mfn);
         my $self = shift;  
         return $self->{NXTMFN};  
 }  
   
 sub GetMFNArray {  
         my $self = shift;  
         return $self->{MFNArray};  
 }  
 =begin php  
   
   sub Read32($fileNB)  
   {  
     $B1=ord(fread($fileNB,1));  
     $B2=ord(fread($fileNB,1));  
     $B3=ord(fread($fileNB,1));  
     $B4=ord(fread($fileNB,1));  
   
     if ($B4<=128)  
     {  
       $value=$B1+$B2*256+$B3*65536+$B4*16777216;  
     } else  
     {  
       $value=$self->Not8($B1)+$self->Not8($B2)*256+$self->Not8($B3)*65536+$self->Not8($B4)*16777216;  
       $value=-($value+1);  
     }  
 #    print "(".$B1.",".$B2.",".$B3.",".$B4.":".$value.")";  
513    
514      return $value;            foreach my $k (keys %{$row}) {
515    }                  foreach my $l (@{$row->{$k}}) {
516    
517    sub Read24($fileNB)                          # filter output
518    {                          $l = $self->{'hash_filter'}->($l) if ($self->{'hash_filter'});
     $B1=ord(fread($fileNB,1));  
     $B2=ord(fread($fileNB,1));  
     $B3=ord(fread($fileNB,1));  
519    
520      $value=$B1+$B2*256+$B3*65536;                          my $val;
521    
522  #    print "(".$B1.",".$B2.",".$B3.":".$value.")";                          # has identifiers?
523                            ($val->{'i1'},$val->{'i2'}) = ($1,$2) if ($l =~ s/^([01 #])([01 #])//);
524    
525      return $value;                            # has subfields?
526    }                          if ($l =~ m/\^/) {
527                                    foreach my $t (split(/\^/,$l)) {
528                                            next if (! $t);
529                                            $val->{substr($t,0,1)} = substr($t,1);
530                                    }
531                            } else {
532                                    $val = $l;
533                            }
534    
535    sub Read16($fileNB)                          push @{$rec->{$k}}, $val;
536    {                  }
537      $B1=ord(fread($fileNB,1));          }
     $B2=ord(fread($fileNB,1));  
538    
539      $value=$B1+$B2*256;          return $rec;
540  #    print "(".$B1.",".$B2.":".$value.")";  }
541    
542      return $value;    =head2 tag_name
   }  
543    
544    sub Read8($fileNB)  Return name of selected tag
   {  
     $B1=ord(fread($fileNB,1));  
545    
546      $value=$B1;   print $isis->tag_name('200');
 #    print "(".$value.")";  
547    
548      return $value;    =cut
   }  
549    
550    sub Not8($value)  sub tag_name {
551    {          my $self = shift;
552      $value=decbin($value);          my $tag = shift || return;
553      if (strlen($value)<8)          return $self->{'TagName'}->{$tag} || $tag;
     {  
       $buffer="";  
       for($i=0;$i<(8-strlen($value));$i++)  
       {  
         $buffer.="0";  
       }  
       $value=$buffer.$value;  
     }  
     $value=ereg_replace("0","3",$value);  
     $value=ereg_replace("1","0",$value);  
     $value=ereg_replace("3","1",$value);  
     $value=bindec($value);  
     return $value;  
   }  
554  }  }
555    
 =cut  
   
556  1;  1;
 __END__  
557    
558  =head1 BUGS  =head1 BUGS
559    
# Line 734  This module has been very lightly tested Line 566  This module has been very lightly tested
566          dpavlin@rot13.org          dpavlin@rot13.org
567          http://www.rot13.org/~dpavlin/          http://www.rot13.org/~dpavlin/
568    
569  This module is based heavily on code from LIBISIS.PHP - Library to read ISIS files V0.1.1  This module is based heavily on code from C<LIBISIS.PHP> library to read ISIS files V0.1.1
570  written in php and (c) 2000 Franck Martin - <franck@sopac.org> released under LGPL.  written in php and (c) 2000 Franck Martin <franck@sopac.org> and released under LGPL.
571    
572  =head1 COPYRIGHT  =head1 COPYRIGHT
573    
# Line 748  LICENSE file included with this module. Line 580  LICENSE file included with this module.
580    
581  =head1 SEE ALSO  =head1 SEE ALSO
582    
583  L<http://www.openisis.org|OpenIsis>, perl(1).  OpenIsis web site L<http://www.openisis.org>
584    
585    perl4lib site L<http://perl4lib.perl.org>
586    
Legend:
Removed from v.3  
changed lines
  Added in v.19
  ViewVC Help
Powered by ViewVC 1.1.26