/[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 27 by dpavlin, Sat Jan 1 22:29:35 2005 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     = 0.08;
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 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 which
41    seems to depriciate it's old C<XS> bindings for perl.
42    
43    It can create hash values from data in ISIS database (using C<to_hash>),
44    ASCII dump (using C<to_ascii>) or just hash with field names and packed
45    values (like C<^asomething^belse>).
46    
47    Unique feature of this module is ability to C<include_deleted> records.
48    It will also skip zero sized fields (OpenIsis has a bug in XS bindings, so
49    fields which are zero sized will be filled with random junk from memory).
50    
51    It also has support for identifiers (only if ISIS database is created by
52    IsisMarc), see C<to_hash>.
53    
54    This module will always be slower than OpenIsis module which use C
55    library. However, since it's written in perl, it's platform independent (so
56    you don't need C compiler), and can be easily modified. I hope that it
57    creates data structures which are easier to use than ones created by
58    OpenIsis, so reduced time in other parts of the code should compensate for
59    slower performance of this module (speed of reading ISIS database is
60    rarely an issue).
61    
62  =head1 METHODS  =head1 METHODS
63    
# Line 50  It can be used as perl-only alternative Line 77  It can be used as perl-only alternative
77  # some binary reads  # some binary reads
78  #  #
79    
 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);  
 }  
   
80  =head2 new  =head2 new
81    
82  Open CDS/ISIS database  Open ISIS database
83    
84   my $isis = new IsisDB(   my $isis = new IsisDB(
85          isisdb => './cds/cds',          isisdb => './cds/cds',
86          read_fdt => 1,          read_fdt => 1,
87            include_deleted => 1,
88            hash_filter => sub {
89                    my $v = shift;
90                    $v =~ s#foo#bar#g;
91            },
92          debug => 1,          debug => 1,
93   );   );
94    
# Line 74  Options are described below: Line 98  Options are described below:
98    
99  =item isisdb  =item isisdb
100    
101  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
102  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
103    C<read_fdt> option) files.
104    
105    In this example it uses C<./cds/cds.MST> and related files.
106    
107  =item read_fdt  =item read_fdt
108    
109  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
110  by default.  by default.
111    
112    =item include_deleted
113    
114    Don't skip logically deleted records in ISIS.
115    
116    =item hash_filter
117    
118    Filter code ref which will be used before data is converted to hash.
119    
120  =item debug  =item debug
121    
122  Dump a C<lot> of debugging output.  Dump a B<lot> of debugging output.
123    
124  =back  =back
125    
# Line 97  sub new { Line 132  sub new {
132          my $self = {};          my $self = {};
133          bless($self, $class);          bless($self, $class);
134    
135          $self->{isisdb} = {@_}->{isisdb} || croak "new needs database name as argument!";          croak "new needs database name (isisdb) as argument!" unless ({@_}->{isisdb});
136    
137            foreach my $v (qw{isisdb debug include_deleted hash_filter}) {
138                    $self->{$v} = {@_}->{$v};
139            }
140    
141            my @isis_files = grep(/\.(FDT|MST|XRF|CNT)$/i,glob($self->{isisdb}."*"));
142    
143            foreach my $f (@isis_files) {
144                    my $ext = $1 if ($f =~ m/\.(\w\w\w)$/);
145                    $self->{lc($ext)."_file"} = $f;
146            }
147    
148            my @must_exist = qw(mst xrf);
149            push @must_exist, "fdt" if ($self->{read_fdt});
150    
151            foreach my $ext (@must_exist) {
152                    croak "missing ",uc($ext)," file in ",$self->{isisdb} unless ($self->{$ext."_file"});
153            }
154    
155          $self->{debug} = {@_}->{debug};          print STDERR "## using files: ",join(" ",@isis_files),"\n" if ($self->{debug});
156    
157          # 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!
158          if ({@_}->{read_fdt} && -e $self->{isisdb}.".FDT") {          if ($self->{read_fdt} && -e $self->{fdt_file}) {
159    
160                  # read the $db.FDT file for tags                  # read the $db.FDT file for tags
161                  my $fieldzone=0;                  my $fieldzone=0;
162    
163                  open(fileFDT, $self->{isisdb}.".FDT") || croak "can't read '$self->{isisdb}.FDT': $!";                  open(fileFDT, $self->{fdt_file}) || croak "can't read '$self->{fdt_file}': $!";
164    
165                  while (<fileFDT>) {                  while (<fileFDT>) {
166                          chomp;                          chomp;
# Line 131  sub new { Line 184  sub new {
184    
185          # Get the Maximum MFN from $db.MST          # Get the Maximum MFN from $db.MST
186    
187          open(fileMST,$self->{isisdb}.".MST") || croak "can't read '$self->{isisdb}.MST': $!";          open($self->{'fileMST'}, $self->{mst_file}) || croak "can't open '$self->{mst_file}': $!";
188    
189          # MST format:   (* = 32 bit signed)          # MST format:   (* = 32 bit signed)
190          # CTLMFN*       always 0          # CTLMFN*       always 0
# Line 139  sub new { Line 192  sub new {
192          # NXTMFB*       last block allocated to master file          # NXTMFB*       last block allocated to master file
193          # NXTMFP        offset to next available position in last block          # NXTMFP        offset to next available position in last block
194          # MFTYPE        always 0 for user db file (1 for system)          # MFTYPE        always 0 for user db file (1 for system)
195          seek(fileMST,4,0);          seek($self->{'fileMST'},4,0);
196          $self->{'NXTMFN'}=$self->Read32(\*fileMST) || carp "NXTNFN is zero";  
197            my $buff;
198    
199            read($self->{'fileMST'}, $buff, 4);
200            $self->{'NXTMFN'}=unpack("l",$buff) || carp "NXTNFN is zero";
201    
202          # save maximum MFN          # save maximum MFN
203          $self->{'maxmfn'} = $self->{'NXTMFN'} - 1;          $self->{'maxmfn'} = $self->{'NXTMFN'} - 1;
204    
205          close(fileMST);  
206    
207    
208            print STDERR Dumper($self),"\n" if ($self->{debug});
209    
210            # open files for later
211            open($self->{'fileXRF'}, $self->{xrf_file}) || croak "can't open '$self->{xrf_file}': $!";
212    
213            $self ? return $self : return undef;
214    }
215    
216    =head2 read_cnt
217    
218    Read content of C<.CNT> file and return hash containing it.
219    
220      print Dumper($isis->read_cnt);
221    
222    This function is not used by module (C<.CNT> files are not required for this
223    module to work), but it can be useful to examine your index (while debugging
224    for example).
225    
226    =cut
227    
228    sub read_cnt  {
229            my $self = shift;
230    
231            croak "missing CNT file in ",$self->{isisdb} unless ($self->{cnt_file});
232    
233          # Get the index information from $db.CNT          # Get the index information from $db.CNT
234        
235          open(fileCNT, $self->{isisdb}.".CNT") || croak "can't read '$self->{isisdb}.CNT': $!";          open(fileCNT, $self->{cnt_file}) || croak "can't read '$self->{cnt_file}': $!";
236    
237          # There is two 26 Bytes fixed lenght records          # There is two 26 Bytes fixed lenght records
238    
# Line 173  sub new { Line 256  sub new {
256                  my $buff = shift || return;                  my $buff = shift || return;
257                  my @arr = unpack("ssssssllls", $buff);                  my @arr = unpack("ssssssllls", $buff);
258    
259                    print STDERR "unpack_cnt: ",join(" ",@arr),"\n" if ($self->{'debug'});
260    
261                  my $IDTYPE = shift @arr;                  my $IDTYPE = shift @arr;
262                  foreach (@flds) {                  foreach (@flds) {
263                          $self->{$IDTYPE}->{$_} = abs(shift @arr);                          $self->{cnt}->{$IDTYPE}->{$_} = abs(shift @arr);
264                  }                  }
265          }          }
266    
267          my $buff;          my $buff;
268    
269          read(fileCNT, $buff, 26);          read(fileCNT, $buff, 26);
270          $self->unpack_cnt($buff);          $self->unpack_cnt($buff);
271    
272          read(fileCNT, $buff, 26);          read(fileCNT, $buff, 26);
273          $self->unpack_cnt($buff);          $self->unpack_cnt($buff);
274    
   
275          close(fileCNT);          close(fileCNT);
276    
277          print Dumper($self) if ($self->{debug});          return $self->{cnt};
   
         $self ? return $self : return undef;  
278  }  }
279    
280  =head2 GetMFN  =head2 fetch
281    
282  Read record with selected MFN  Read record with selected MFN
283    
284    my $rec = $isis->GetMFN(55);    my $rec = $isis->fetch(55);
285    
286  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
287  for that field.  for that field like this:
288    
289      $rec = {
290        '210' => [ '^aNew York^cNew York University press^dcop. 1988' ],
291        '990' => [ '2140', '88', 'HAY' ],
292      };
293    
294  =cut  =cut
295    
296  sub GetMFN {  sub fetch {
297          my $self = shift;          my $self = shift;
298    
299          my $mfn = shift || croak "GetMFN needs MFN as argument!";          my $mfn = shift || croak "fetch needs MFN as argument!";
300    
301          print "GetMFN: $mfn\n" if ($self->{debug});          # is mfn allready in memory?
302            my $old_mfn = $self->{'current_mfn'} || -1;
303            return $self->{record} if ($mfn == $old_mfn);
304    
305          open(fileXRF, $self->{isisdb}.".XRF") || croak "can't open '$self->{isisdb}.XRF': $!";          print STDERR "## fetch: $mfn\n" if ($self->{debug});
306    
307          # XXX check this?          # XXX check this?
308          my $mfnpos=($mfn+int(($mfn-1)/127))*4;          my $mfnpos=($mfn+int(($mfn-1)/127))*4;
309    
310          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 ($self->{debug});
311          seek(fileXRF,$mfnpos,0);          seek($self->{'fileXRF'},$mfnpos,0);
312    
313          # read XRFMFB abd XRFMFP          my $buff;
         my $pointer=$self->Read32(\*fileXRF);  
314    
315          my $XRFMFB = int($pointer/2048);          # delete old record
316          my $XRFMFP = $pointer - ($XRFMFB*2048);          delete $self->{record};
317    
318            # read XRFMFB abd XRFMFP
319            read($self->{'fileXRF'}, $buff, 4);
320            my $pointer=unpack("l",$buff) || carp "pointer is null";
321    
322          print "XRFMFB: $XRFMFB XRFMFP: $XRFMFP\n" if ($self->{debug});          # check for logically deleted record
323            if ($pointer < 0) {
324                    print STDERR "## record $mfn is logically deleted\n" if ($self->{debug});
325                    $self->{deleted} = $mfn;
326    
327          # XXX fix this to be more readable!!                  return unless $self->{include_deleted};
         # e.g. (XRFMFB - 1) * 512 + XRFMFP  
328    
329          my $offset = $pointer;                  $pointer = abs($pointer);
         my $offset2=int($offset/2048)-1;  
         my $offset22=int($offset/4096);  
         my $offset3=$offset-($offset22*4096);  
         if ($offset3>512) {  
                 $offset3=$offset3-2048;  
330          }          }
         my $offset4=($offset2*512)+$offset3;  
331    
332          print "$offset - $offset2 - $offset3 - $offset4\n" if ($self->{debug});          my $XRFMFB = int($pointer/2048);
333            my $XRFMFP = $pointer - ($XRFMFB*2048);
334    
335            # (XRFMFB - 1) * 512 + XRFMFP
336            # why do i have to do XRFMFP % 1024 ?
337    
338          close(fileXRF);          my $blk_off = (($XRFMFB - 1) * 512) + ($XRFMFP % 512);
339    
340            print STDERR "## pointer: $pointer XRFMFB: $XRFMFB XRFMFP: $XRFMFP offset: $blk_off\n" if ($self->{'debug'});
341    
342          # Get Record Information          # Get Record Information
343    
344          open(fileMST, $self->{isisdb}.".MST") || croak "can't open '$self->{isisdb}.MST': $!";          seek($self->{'fileMST'},$blk_off,0);
345    
346          seek(fileMST,$offset4,0);          read($self->{'fileMST'}, $buff, 4);
347            my $value=unpack("l",$buff);
348    
349          my $value=$self->Read32(\*fileMST);          print STDERR "## offset for rowid $value is $blk_off (blk $XRFMFB off $XRFMFP)\n" if ($self->{debug});
350    
351          if ($value!=$mfn) {          if ($value!=$mfn) {
352  print ("Error: The MFN:".$mfn." is not found in MST(".$value.")");                      if ($value == 0) {
353                  return -1;      # XXX deleted record?                          print STDERR "## record $mfn is physically deleted\n" if ($self->{debug});
354                            $self->{deleted} = $mfn;
355                            return;
356                    }
357    
358                    carp "Error: MFN ".$mfn." not found in MST file, found $value";    
359                    return;
360          }          }
361    
362  #       $MFRL=$self->Read16($fileMST);          read($self->{'fileMST'}, $buff, 14);
 #       $MFBWB=$self->Read32($fileMST);  
 #       $MFBWP=$self->Read16($fileMST);  
 #       $BASE=$self->Read16($fileMST);  
 #       $NVF=$self->Read16($fileMST);  
 #       $STATUS=$self->Read16($fileMST);  
   
         my $buff;  
         read(fileMST, $buff, 14);  
363    
364          my ($MFRL,$MFBWB,$MFBWP,$BASE,$NVF,$STATUS) = unpack("slssss", $buff);          my ($MFRL,$MFBWB,$MFBWP,$BASE,$NVF,$STATUS) = unpack("slssss", $buff);
365    
366          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});
367    
368            warn "MFRL $MFRL is not even number" unless ($MFRL % 2 == 0);
369    
370            warn "BASE is not 18+6*NVF" unless ($BASE == 18 + 6 * $NVF);
371    
372          # Get Directory Format          # Get Directory Format
373    
# Line 277  print ("Error: The MFN:".$mfn." is not f Line 375  print ("Error: The MFN:".$mfn." is not f
375          my @FieldLEN;          my @FieldLEN;
376          my @FieldTAG;          my @FieldTAG;
377    
378          for (my $i = 0 ; $i < $NVF ; $i++) {          read($self->{'fileMST'}, $buff, 6 * $NVF);
379    
380            my $rec_len = 0;
381    
382  #               $TAG=$self->Read16($fileMST);          for (my $i = 0 ; $i < $NVF ; $i++) {
 #               $POS=$self->Read16($fileMST);  
 #               $LEN=$self->Read16($fileMST);  
383    
384                  read(fileMST, $buff, 6);                  my ($TAG,$POS,$LEN) = unpack("sss", substr($buff,$i * 6, 6));
                 my ($TAG,$POS,$LEN) = unpack("sss", $buff);  
385    
386                  print "TAG: $TAG POS: $POS LEN: $LEN\n" if ($self->{debug});                  print STDERR "## TAG: $TAG POS: $POS LEN: $LEN\n" if ($self->{debug});
387    
388                  # 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.
389                  #                  #
# Line 301  print ("Error: The MFN:".$mfn." is not f Line 398  print ("Error: The MFN:".$mfn." is not f
398                  push @FieldTAG,$TAG;                  push @FieldTAG,$TAG;
399                  push @FieldPOS,$POS;                  push @FieldPOS,$POS;
400                  push @FieldLEN,$LEN;                  push @FieldLEN,$LEN;
401    
402                    $rec_len += $LEN;
403          }          }
404    
405          # Get Variable Fields          # Get Variable Fields
406    
407          delete $self->{record};          read($self->{'fileMST'},$buff,$rec_len);
408    
409            print STDERR "## rec_len: $rec_len poc: ",tell($self->{'fileMST'})."\n" if ($self->{debug});
410    
411          for (my $i = 0 ; $i < $NVF ; $i++) {          for (my $i = 0 ; $i < $NVF ; $i++) {
412                  my $rec;                  # skip zero-sized fields
413                  read(fileMST,$rec,$FieldLEN[$i]);                  next if ($FieldLEN[$i] == 0);
                 push @{$self->{record}->{$FieldTAG[$i]}}, $rec;  
         }  
         close(fileMST);  
414    
415          # The record is marked for deletion                  push @{$self->{record}->{$FieldTAG[$i]}}, substr($buff,$FieldPOS[$i],$FieldLEN[$i]);
         if ($STATUS==1) {  
                 return -1;  
416          }          }
417    
418          print Dumper($self) if ($self->{debug});          $self->{'current_mfn'} = $mfn;
419    
420            print STDERR Dumper($self),"\n" if ($self->{debug});
421    
422          return $self->{'record'};          return $self->{'record'};
423  }  }
424    
425  =head2 to_ascii  =head2 to_ascii
426    
427  Dump ascii output of selected MFN  Returns ASCII output of record with specified MFN
428    
429      print $isis->to_ascii(42);
430    
431    This outputs something like this:
432    
433      210   ^aNew York^cNew York University press^dcop. 1988
434      990   2140
435      990   88
436      990   HAY
437    
438    print $isis->to_ascii(55);  If C<read_fdt> is specified when calling C<new> it will display field names
439    from C<.FDT> file instead of numeric tags.
440    
441  =cut  =cut
442    
# Line 337  sub to_ascii { Line 445  sub to_ascii {
445    
446          my $mfn = shift || croak "need MFN";          my $mfn = shift || croak "need MFN";
447    
448          my $rec = $self->GetMFN($mfn);          my $rec = $self->fetch($mfn);
449    
450          my $out = "0\t$mfn";          my $out = "0\t$mfn";
451    
452          foreach my $f (sort keys %{$rec}) {          foreach my $f (sort keys %{$rec}) {
453                  $out .= "\n$f\t".join("\n$f\t",@{$self->{record}->{$f}});                  my $fn = $self->tag_name($f);
454                    $out .= "\n$fn\t".join("\n$fn\t",@{$self->{record}->{$f}});
455          }          }
456    
457          $out .= "\n";          $out .= "\n";
# Line 350  sub to_ascii { Line 459  sub to_ascii {
459          return $out;          return $out;
460  }  }
461    
462  ################# old cruft which is not ported from php to perl  =head2 to_hash
   
 =begin php  
   
   # Load the dictionary from the $db.L0x files.  
   # 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);  
   
     } While (!feof($fileL01));  
463    
464      fclose($fileL01);  Read record with specified MFN and convert it to hash
   }  
   
   # self function search through the tree and returns an array of pointers to IFP  
   # The function must be recursive  
465    
466    sub SearchTree($search,$fileNB,$PUNT)    my $hash = $isis->to_hash($mfn);
   {        
       $offset=(($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;  
   }  
467    
468    # Search ISIS for record containing search  It has ability to convert characters (using C<hash_filter>) from ISIS
469    # Return a sorted array of MFN  database before creating structures enabling character re-mapping or quick
470    fix-up of data.
471    
472    This function returns hash which is like this:
473    
474      $hash = {
475        '210' => [
476                   {
477                     'c' => 'New York University press',
478                     'a' => 'New York',
479                     'd' => 'cop. 1988'
480                   }
481                 ],
482        '990' => [
483                   '2140',
484                   '88',
485                   'HAY'
486                 ],
487      };
488    
489    You can later use that hash to produce any output from ISIS data.
490    
491    If database is created using IsisMarc, it will also have to special fields
492    which will be used for identifiers, C<i1> and C<i2> like this:
493    
494      '200' => [
495                 {
496                   'i1' => '1',
497                   'i2' => ' '
498                   'a' => 'Goa',
499                   'f' => 'Valdo D\'Arienzo',
500                   'e' => 'tipografie e tipografi nel XVI secolo',
501                 }
502               ],
503    
504    sub Search($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;  
   }  
505    
506  =cut  =cut
507    
508  #  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 {  
         my $self = shift;  
         return $self->{FieldName};  
 }  
   
 sub GetTagName {  
         my $self = shift;  
         return $self->{TagName};  
 }  
   
 sub GetFieldTag {  
509          my $self = shift;          my $self = shift;
         return $self->{FieldTAG};  
 }  
510    
511  sub GetNextMFN {          my $mfn = shift || confess "need mfn!";
         my $self = shift;  
         return $self->{NXTMFN};  
 }  
512    
513  sub GetMFNArray {          # init record to include MFN as field 000
514          my $self = shift;          my $rec = { '000' => [ $mfn ] };
         return $self->{MFNArray};  
 }  
 =begin php  
515    
516    sub Read32($fileNB)          my $row = $self->fetch($mfn);
   {  
     $B1=ord(fread($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.")";  
517    
518      return $value;            foreach my $k (keys %{$row}) {
519    }                  foreach my $l (@{$row->{$k}}) {
520    
521    sub Read24($fileNB)                          # filter output
522    {                          $l = $self->{'hash_filter'}->($l) if ($self->{'hash_filter'});
     $B1=ord(fread($fileNB,1));  
     $B2=ord(fread($fileNB,1));  
     $B3=ord(fread($fileNB,1));  
523    
524      $value=$B1+$B2*256+$B3*65536;                          my $val;
525    
526  #    print "(".$B1.",".$B2.",".$B3.":".$value.")";                          # has identifiers?
527                            ($val->{'i1'},$val->{'i2'}) = ($1,$2) if ($l =~ s/^([01 #])([01 #])\^/\^/);
528    
529      return $value;                            # has subfields?
530    }                          if ($l =~ m/\^/) {
531                                    foreach my $t (split(/\^/,$l)) {
532                                            next if (! $t);
533                                            $val->{substr($t,0,1)} = substr($t,1);
534                                    }
535                            } else {
536                                    $val = $l;
537                            }
538    
539    sub Read16($fileNB)                          push @{$rec->{$k}}, $val;
540    {                  }
541      $B1=ord(fread($fileNB,1));          }
     $B2=ord(fread($fileNB,1));  
542    
543      $value=$B1+$B2*256;          return $rec;
544  #    print "(".$B1.",".$B2.":".$value.")";  }
545    
546      return $value;    =head2 tag_name
   }  
547    
548    sub Read8($fileNB)  Return name of selected tag
   {  
     $B1=ord(fread($fileNB,1));  
549    
550      $value=$B1;   print $isis->tag_name('200');
 #    print "(".$value.")";  
551    
552      return $value;    =cut
   }  
553    
554    sub Not8($value)  sub tag_name {
555    {          my $self = shift;
556      $value=decbin($value);          my $tag = shift || return;
557      if (strlen($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;  
   }  
558  }  }
559    
 =cut  
   
560  1;  1;
 __END__  
561    
562  =head1 BUGS  =head1 BUGS
563    
564  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
565    some variations in input databases which has been tested with this module.
566    When I was in doubt, I assumed that OpenIsis's implementation was right
567    (except for obvious bugs).
568    
569    However, every effort has been made to test this module with as much
570    databases (and programs that create them) as possible.
571    
572    I would be very greatful for success or failure reports about usage of this
573    module with databases from programs other than WinIsis and IsisMarc. I had
574    tested this against ouput of one C<isis.dll>-based application, but I don't
575    know any details about it's version.
576    
577  =head1 AUTHOR  =head1 AUTHOR
578    
# Line 734  This module has been very lightly tested Line 581  This module has been very lightly tested
581          dpavlin@rot13.org          dpavlin@rot13.org
582          http://www.rot13.org/~dpavlin/          http://www.rot13.org/~dpavlin/
583    
584  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
585  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.
586    
587  =head1 COPYRIGHT  =head1 COPYRIGHT
588    
# Line 748  LICENSE file included with this module. Line 595  LICENSE file included with this module.
595    
596  =head1 SEE ALSO  =head1 SEE ALSO
597    
598  L<http://www.openisis.org|OpenIsis>, perl(1).  OpenIsis web site L<http://www.openisis.org>
599    
600    perl4lib site L<http://perl4lib.perl.org>
601    
Legend:
Removed from v.3  
changed lines
  Added in v.27
  ViewVC Help
Powered by ViewVC 1.1.26