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 = 0.06; |
$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 (); |
37 |
=head1 DESCRIPTION |
=head1 DESCRIPTION |
38 |
|
|
39 |
This module will read ISIS databases created by DOS CDS/ISIS, WinIsis or |
This module will read ISIS databases created by DOS CDS/ISIS, WinIsis or |
40 |
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 |
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>), |
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 |
ASCII dump (using C<to_ascii>) or just hash with field names and packed |
51 |
It also has support for identifiers (only if ISIS database is created by |
It also has support for identifiers (only if ISIS database is created by |
52 |
IsisMarc), see C<to_hash>. |
IsisMarc), see C<to_hash>. |
53 |
|
|
54 |
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 |
55 |
library. However, since it's written in perl, it's platform independent (so |
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 |
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 |
creates data structures which are easier to use than ones created by |
99 |
=item isisdb |
=item isisdb |
100 |
|
|
101 |
This is full or relative path to ISIS database files which include |
This is full or relative path to ISIS database files which include |
102 |
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. |
In this example it uses C<./cds/cds.MST> and related files. |
106 |
|
|
138 |
$self->{$v} = {@_}->{$v}; |
$self->{$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->{$ext."_file"}); |
153 |
|
} |
154 |
|
|
155 |
|
print STDERR "## using files: ",join(" ",@isis_files),"\n" if ($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; |
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}': $!"; |
188 |
|
|
189 |
# MST format: (* = 32 bit signed) |
# MST format: (* = 32 bit signed) |
190 |
# CTLMFN* always 0 |
# CTLMFN* always 0 |
192 |
# NXTMFB* last block allocated to master file |
# NXTMFB* last block allocated to master file |
193 |
# NXTMFP offset to next available position in last block |
# NXTMFP offset to next available position in last block |
194 |
# MFTYPE always 0 for user db file (1 for system) |
# MFTYPE always 0 for user db file (1 for system) |
195 |
seek(fileMST,4,0); |
seek($self->{'fileMST'},4,0); |
196 |
|
|
197 |
my $buff; |
my $buff; |
198 |
|
|
199 |
read(fileMST, $buff, 4); |
read($self->{'fileMST'}, $buff, 4); |
200 |
$self->{'NXTMFN'}=unpack("l",$buff) || carp "NXTNFN is zero"; |
$self->{'NXTMFN'}=unpack("l",$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 |
|
|
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; |
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 STDERR 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; |
|
278 |
} |
} |
279 |
|
|
280 |
=head2 fetch |
=head2 fetch |
300 |
|
|
301 |
# is mfn allready in memory? |
# is mfn allready in memory? |
302 |
my $old_mfn = $self->{'current_mfn'} || -1; |
my $old_mfn = $self->{'current_mfn'} || -1; |
303 |
return if ($mfn == $old_mfn); |
return $self->{record} if ($mfn == $old_mfn); |
304 |
|
|
305 |
print STDERR "## fetch: $mfn\n" if ($self->{debug}); |
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 STDERR "## seeking to $mfnpos in file '$self->{isisdb}.XRF'\n" if ($self->{debug}); |
print STDERR "## seeking to $mfnpos in file '$self->{xrf_file}'\n" if ($self->{debug}); |
311 |
seek($self->{'fileXRF'},$mfnpos,0); |
seek($self->{'fileXRF'},$mfnpos,0); |
312 |
|
|
313 |
my $buff; |
my $buff; |
314 |
|
|
315 |
|
# delete old record |
316 |
|
delete $self->{record}; |
317 |
|
|
318 |
# read XRFMFB abd XRFMFP |
# read XRFMFB abd XRFMFP |
319 |
read($self->{'fileXRF'}, $buff, 4); |
read($self->{'fileXRF'}, $buff, 4); |
320 |
my $pointer=unpack("l",$buff) || carp "pointer is null"; |
my $pointer=unpack("l",$buff) || carp "pointer is null"; |
321 |
|
|
322 |
|
# 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 |
|
return unless $self->{include_deleted}; |
328 |
|
|
329 |
|
$pointer = abs($pointer); |
330 |
|
} |
331 |
|
|
332 |
my $XRFMFB = int($pointer/2048); |
my $XRFMFB = int($pointer/2048); |
333 |
my $XRFMFP = $pointer - ($XRFMFB*2048); |
my $XRFMFP = $pointer - ($XRFMFB*2048); |
334 |
|
|
|
|
|
335 |
# (XRFMFB - 1) * 512 + XRFMFP |
# (XRFMFB - 1) * 512 + XRFMFP |
336 |
# why do i have to do XRFMFP % 1024 ? |
# why do i have to do XRFMFP % 1024 ? |
337 |
|
|
338 |
my $blk_off = (($XRFMFB - 1) * 512) + ($XRFMFP % 1024); |
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'}); |
print STDERR "## pointer: $pointer XRFMFB: $XRFMFB XRFMFP: $XRFMFP offset: $blk_off\n" if ($self->{'debug'}); |
341 |
|
|
349 |
print STDERR "## offset for rowid $value is $blk_off (blk $XRFMFB off $XRFMFP)\n" if ($self->{debug}); |
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 |
carp "Error: MFN ".$mfn." not found in MST(".$value.")"; |
if ($value == 0) { |
353 |
#return; # XXX deleted record? |
print STDERR "## record $mfn is physically deleted\n" if ($self->{debug}); |
354 |
} |
$self->{deleted} = $mfn; |
355 |
|
return; |
356 |
|
} |
357 |
|
|
358 |
# $MFRL=$self->Read16($fileMST); |
carp "Error: MFN ".$mfn." not found in MST file, found $value"; |
359 |
# $MFBWB=$self->Read32($fileMST); |
return; |
360 |
# $MFBWP=$self->Read16($fileMST); |
} |
|
# $BASE=$self->Read16($fileMST); |
|
|
# $NVF=$self->Read16($fileMST); |
|
|
# $STATUS=$self->Read16($fileMST); |
|
361 |
|
|
362 |
read($self->{'fileMST'}, $buff, 14); |
read($self->{'fileMST'}, $buff, 14); |
363 |
|
|
365 |
|
|
366 |
print STDERR "## 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 |
# delete old record |
warn "MFRL $MFRL is not even number" unless ($MFRL % 2 == 0); |
|
delete $self->{record}; |
|
|
|
|
|
## FIXME this is a bug |
|
|
if (! $self->{'include_deleted'} && $MFRL < 0) { |
|
|
print "## logically deleted record $mfn, skipping...\n" if ($self->{debug}); |
|
|
return; |
|
|
} |
|
369 |
|
|
370 |
warn "BASE is not 18+6*NVF" unless ($BASE == 18 + 6 * $NVF); |
warn "BASE is not 18+6*NVF" unless ($BASE == 18 + 6 * $NVF); |
371 |
|
|
381 |
|
|
382 |
for (my $i = 0 ; $i < $NVF ; $i++) { |
for (my $i = 0 ; $i < $NVF ; $i++) { |
383 |
|
|
|
# $TAG=$self->Read16($fileMST); |
|
|
# $POS=$self->Read16($fileMST); |
|
|
# $LEN=$self->Read16($fileMST); |
|
|
|
|
384 |
my ($TAG,$POS,$LEN) = unpack("sss", substr($buff,$i * 6, 6)); |
my ($TAG,$POS,$LEN) = unpack("sss", substr($buff,$i * 6, 6)); |
385 |
|
|
386 |
print STDERR "## TAG: $TAG POS: $POS LEN: $LEN\n" if ($self->{debug}); |
print STDERR "## TAG: $TAG POS: $POS LEN: $LEN\n" if ($self->{debug}); |
414 |
|
|
415 |
push @{$self->{record}->{$FieldTAG[$i]}}, substr($buff,$FieldPOS[$i],$FieldLEN[$i]); |
push @{$self->{record}->{$FieldTAG[$i]}}, substr($buff,$FieldPOS[$i],$FieldLEN[$i]); |
416 |
} |
} |
|
close(fileMST); |
|
417 |
|
|
418 |
$self->{'current_mfn'} = $mfn; |
$self->{'current_mfn'} = $mfn; |
419 |
|
|
420 |
print Dumper($self),"\n" if ($self->{debug}); |
print STDERR Dumper($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 record with specified MFN |
Returns ASCII output of record with specified MFN |
428 |
|
|
429 |
print $isis->to_ascii(42); |
print $isis->to_ascii(42); |
430 |
|
|
431 |
It outputs something like this: |
This outputs something like this: |
432 |
|
|
433 |
210 ^aNew York^cNew York University press^dcop. 1988 |
210 ^aNew York^cNew York University press^dcop. 1988 |
434 |
990 2140 |
990 2140 |
465 |
|
|
466 |
my $hash = $isis->to_hash($mfn); |
my $hash = $isis->to_hash($mfn); |
467 |
|
|
468 |
It has ability to convert characters (using C<hash_filter> from ISIS |
It has ability to convert characters (using C<hash_filter>) from ISIS |
469 |
database before creating structures enabling character re-mapping or quick |
database before creating structures enabling character re-mapping or quick |
470 |
fix-up of data. |
fix-up of data. |
471 |
|
|
524 |
my $val; |
my $val; |
525 |
|
|
526 |
# has identifiers? |
# has identifiers? |
527 |
($val->{'i1'},$val->{'i2'}) = ($1,$2) if ($l =~ s/^([01 #])([01 #])//); |
($val->{'i1'},$val->{'i2'}) = ($1,$2) if ($l =~ s/^([01 #])([01 #])\^/\^/); |
528 |
|
|
529 |
# has subfields? |
# has subfields? |
530 |
if ($l =~ m/\^/) { |
if ($l =~ m/\^/) { |
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 |
|
|