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 = 0.07; |
$VERSION = 0.09; |
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 (); |
30 |
isisdb => './cds/cds', |
isisdb => './cds/cds', |
31 |
); |
); |
32 |
|
|
33 |
for(my $mfn = 1; $mfn <= $isis->{'maxmfn'}; $mfn++) { |
for(my $mfn = 1; $mfn <= $isis->count; $mfn++) { |
34 |
print $isis->to_ascii($mfn),"\n"; |
print $isis->to_ascii($mfn),"\n"; |
35 |
} |
} |
36 |
|
|
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 |
123 |
|
|
124 |
=back |
=back |
125 |
|
|
|
It will also set C<$isis-E<gt>{'maxmfn'}> which is maximum MFN stored in database. |
|
|
|
|
126 |
=cut |
=cut |
127 |
|
|
128 |
sub new { |
sub new { |
197 |
read($self->{'fileMST'}, $buff, 4); |
read($self->{'fileMST'}, $buff, 4); |
198 |
$self->{'NXTMFN'}=unpack("l",$buff) || carp "NXTNFN is zero"; |
$self->{'NXTMFN'}=unpack("l",$buff) || carp "NXTNFN is zero"; |
199 |
|
|
|
# save maximum MFN |
|
|
$self->{'maxmfn'} = $self->{'NXTMFN'} - 1; |
|
|
|
|
200 |
|
|
201 |
|
|
202 |
|
|
208 |
$self ? return $self : return undef; |
$self ? return $self : return undef; |
209 |
} |
} |
210 |
|
|
211 |
|
=head2 count |
212 |
|
|
213 |
|
Return number of records in database |
214 |
|
|
215 |
|
print $isis->count; |
216 |
|
|
217 |
|
=cut |
218 |
|
|
219 |
|
sub count { |
220 |
|
my $self = shift; |
221 |
|
return $self->{'NXTMFN'} - 1; |
222 |
|
} |
223 |
|
|
224 |
=head2 read_cnt |
=head2 read_cnt |
225 |
|
|
226 |
This function is not really used by module, but can be useful to find info |
Read content of C<.CNT> file and return hash containing it. |
|
about your index (if debugging it for example). |
|
227 |
|
|
228 |
print Dumper($isis->read_cnt); |
print Dumper($isis->read_cnt); |
229 |
|
|
230 |
|
This function is not used by module (C<.CNT> files are not required for this |
231 |
|
module to work), but it can be useful to examine your index (while debugging |
232 |
|
for example). |
233 |
|
|
234 |
=cut |
=cut |
235 |
|
|
236 |
sub read_cnt { |
sub read_cnt { |
242 |
|
|
243 |
open(fileCNT, $self->{cnt_file}) || croak "can't read '$self->{cnt_file}': $!"; |
open(fileCNT, $self->{cnt_file}) || croak "can't read '$self->{cnt_file}': $!"; |
244 |
|
|
|
# There is two 26 Bytes fixed lenght records |
|
|
|
|
|
# 0: IDTYPE BTree type 16 |
|
|
# 2: ORDN Nodes Order 16 |
|
|
# 4: ORDF Leafs Order 16 |
|
|
# 6: N Number of Memory buffers for nodes 16 |
|
|
# 8: K Number of buffers for first level index 16 |
|
|
# 10: LIV Current number of Index Levels 16 |
|
|
# 12: POSRX* Pointer to Root Record in N0x 32 |
|
|
# 16: NMAXPOS* Next Available position in N0x 32 |
|
|
# 20: FMAXPOS* Next available position in L0x 32 |
|
|
# 24: ABNORMAL Formal BTree normality indicator 16 |
|
|
# length: 26 bytes |
|
|
|
|
|
sub unpack_cnt { |
|
|
my $self = shift; |
|
|
|
|
|
my @flds = qw(ORDN ORDF N K LIV POSRX NMAXPOS FMAXPOS ABNORMAL); |
|
|
|
|
|
my $buff = shift || return; |
|
|
my @arr = unpack("ssssssllls", $buff); |
|
|
|
|
|
print STDERR "unpack_cnt: ",join(" ",@arr),"\n" if ($self->{'debug'}); |
|
|
|
|
|
my $IDTYPE = shift @arr; |
|
|
foreach (@flds) { |
|
|
$self->{cnt}->{$IDTYPE}->{$_} = abs(shift @arr); |
|
|
} |
|
|
} |
|
|
|
|
245 |
my $buff; |
my $buff; |
246 |
|
|
247 |
read(fileCNT, $buff, 26); |
read(fileCNT, $buff, 26); |
255 |
return $self->{cnt}; |
return $self->{cnt}; |
256 |
} |
} |
257 |
|
|
258 |
|
=head2 unpack_cnt |
259 |
|
|
260 |
|
Unpack one of two 26 bytes fixed length record in C<.CNT> file. |
261 |
|
|
262 |
|
Here is definition of record: |
263 |
|
|
264 |
|
off key description size |
265 |
|
0: IDTYPE BTree type s |
266 |
|
2: ORDN Nodes Order s |
267 |
|
4: ORDF Leafs Order s |
268 |
|
6: N Number of Memory buffers for nodes s |
269 |
|
8: K Number of buffers for first level index s |
270 |
|
10: LIV Current number of Index Levels s |
271 |
|
12: POSRX Pointer to Root Record in N0x l |
272 |
|
16: NMAXPOS Next Available position in N0x l |
273 |
|
20: FMAXPOS Next available position in L0x l |
274 |
|
24: ABNORMAL Formal BTree normality indicator s |
275 |
|
length: 26 bytes |
276 |
|
|
277 |
|
This will fill C<$self> object under C<cnt> with hash. It's used by C<read_cnt>. |
278 |
|
|
279 |
|
=cut |
280 |
|
|
281 |
|
sub unpack_cnt { |
282 |
|
my $self = shift; |
283 |
|
|
284 |
|
my @flds = qw(ORDN ORDF N K LIV POSRX NMAXPOS FMAXPOS ABNORMAL); |
285 |
|
|
286 |
|
my $buff = shift || return; |
287 |
|
my @arr = unpack("ssssssllls", $buff); |
288 |
|
|
289 |
|
print STDERR "unpack_cnt: ",join(" ",@arr),"\n" if ($self->{'debug'}); |
290 |
|
|
291 |
|
my $IDTYPE = shift @arr; |
292 |
|
foreach (@flds) { |
293 |
|
$self->{cnt}->{$IDTYPE}->{$_} = abs(shift @arr); |
294 |
|
} |
295 |
|
} |
296 |
|
|
297 |
=head2 fetch |
=head2 fetch |
298 |
|
|
299 |
Read record with selected MFN |
Read record with selected MFN |
317 |
|
|
318 |
# is mfn allready in memory? |
# is mfn allready in memory? |
319 |
my $old_mfn = $self->{'current_mfn'} || -1; |
my $old_mfn = $self->{'current_mfn'} || -1; |
320 |
return if ($mfn == $old_mfn); |
return $self->{record} if ($mfn == $old_mfn); |
321 |
|
|
322 |
print STDERR "## fetch: $mfn\n" if ($self->{debug}); |
print STDERR "## fetch: $mfn\n" if ($self->{debug}); |
323 |
|
|
329 |
|
|
330 |
my $buff; |
my $buff; |
331 |
|
|
332 |
|
# delete old record |
333 |
|
delete $self->{record}; |
334 |
|
|
335 |
# read XRFMFB abd XRFMFP |
# read XRFMFB abd XRFMFP |
336 |
read($self->{'fileXRF'}, $buff, 4); |
read($self->{'fileXRF'}, $buff, 4); |
337 |
my $pointer=unpack("l",$buff) || carp "pointer is null"; |
my $pointer=unpack("l",$buff) || carp "pointer is null"; |
338 |
|
|
339 |
|
# check for logically deleted record |
340 |
|
if ($pointer < 0) { |
341 |
|
print STDERR "## record $mfn is logically deleted\n" if ($self->{debug}); |
342 |
|
$self->{deleted} = $mfn; |
343 |
|
|
344 |
|
return unless $self->{include_deleted}; |
345 |
|
|
346 |
|
$pointer = abs($pointer); |
347 |
|
} |
348 |
|
|
349 |
my $XRFMFB = int($pointer/2048); |
my $XRFMFB = int($pointer/2048); |
350 |
my $XRFMFP = $pointer - ($XRFMFB*2048); |
my $XRFMFP = $pointer - ($XRFMFB*2048); |
351 |
|
|
|
|
|
352 |
# (XRFMFB - 1) * 512 + XRFMFP |
# (XRFMFB - 1) * 512 + XRFMFP |
353 |
# why do i have to do XRFMFP % 1024 ? |
# why do i have to do XRFMFP % 1024 ? |
354 |
|
|
355 |
my $blk_off = (($XRFMFB - 1) * 512) + ($XRFMFP % 1024); |
my $blk_off = (($XRFMFB - 1) * 512) + ($XRFMFP % 512); |
356 |
|
|
357 |
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'}); |
358 |
|
|
366 |
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}); |
367 |
|
|
368 |
if ($value!=$mfn) { |
if ($value!=$mfn) { |
369 |
carp "Error: MFN ".$mfn." not found in MST(".$value.")"; |
if ($value == 0) { |
370 |
#return; # XXX deleted record? |
print STDERR "## record $mfn is physically deleted\n" if ($self->{debug}); |
371 |
} |
$self->{deleted} = $mfn; |
372 |
|
return; |
373 |
|
} |
374 |
|
|
375 |
# $MFRL=$self->Read16($fileMST); |
carp "Error: MFN ".$mfn." not found in MST file, found $value"; |
376 |
# $MFBWB=$self->Read32($fileMST); |
return; |
377 |
# $MFBWP=$self->Read16($fileMST); |
} |
|
# $BASE=$self->Read16($fileMST); |
|
|
# $NVF=$self->Read16($fileMST); |
|
|
# $STATUS=$self->Read16($fileMST); |
|
378 |
|
|
379 |
read($self->{'fileMST'}, $buff, 14); |
read($self->{'fileMST'}, $buff, 14); |
380 |
|
|
382 |
|
|
383 |
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}); |
384 |
|
|
385 |
# 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; |
|
|
} |
|
386 |
|
|
387 |
warn "BASE is not 18+6*NVF" unless ($BASE == 18 + 6 * $NVF); |
warn "BASE is not 18+6*NVF" unless ($BASE == 18 + 6 * $NVF); |
388 |
|
|
398 |
|
|
399 |
for (my $i = 0 ; $i < $NVF ; $i++) { |
for (my $i = 0 ; $i < $NVF ; $i++) { |
400 |
|
|
|
# $TAG=$self->Read16($fileMST); |
|
|
# $POS=$self->Read16($fileMST); |
|
|
# $LEN=$self->Read16($fileMST); |
|
|
|
|
401 |
my ($TAG,$POS,$LEN) = unpack("sss", substr($buff,$i * 6, 6)); |
my ($TAG,$POS,$LEN) = unpack("sss", substr($buff,$i * 6, 6)); |
402 |
|
|
403 |
print STDERR "## TAG: $TAG POS: $POS LEN: $LEN\n" if ($self->{debug}); |
print STDERR "## TAG: $TAG POS: $POS LEN: $LEN\n" if ($self->{debug}); |
434 |
|
|
435 |
$self->{'current_mfn'} = $mfn; |
$self->{'current_mfn'} = $mfn; |
436 |
|
|
437 |
print Dumper($self),"\n" if ($self->{debug}); |
print STDERR Dumper($self),"\n" if ($self->{debug}); |
438 |
|
|
439 |
return $self->{'record'}; |
return $self->{'record'}; |
440 |
} |
} |
441 |
|
|
442 |
=head2 to_ascii |
=head2 to_ascii |
443 |
|
|
444 |
Dump ASCII output of record with specified MFN |
Returns ASCII output of record with specified MFN |
445 |
|
|
446 |
print $isis->to_ascii(42); |
print $isis->to_ascii(42); |
447 |
|
|
448 |
It outputs something like this: |
This outputs something like this: |
449 |
|
|
450 |
210 ^aNew York^cNew York University press^dcop. 1988 |
210 ^aNew York^cNew York University press^dcop. 1988 |
451 |
990 2140 |
990 2140 |
482 |
|
|
483 |
my $hash = $isis->to_hash($mfn); |
my $hash = $isis->to_hash($mfn); |
484 |
|
|
485 |
It has ability to convert characters (using C<hash_filter> from ISIS |
It has ability to convert characters (using C<hash_filter>) from ISIS |
486 |
database before creating structures enabling character re-mapping or quick |
database before creating structures enabling character re-mapping or quick |
487 |
fix-up of data. |
fix-up of data. |
488 |
|
|
578 |
|
|
579 |
=head1 BUGS |
=head1 BUGS |
580 |
|
|
581 |
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 |
582 |
|
some variations in input databases which has been tested with this module. |
583 |
|
When I was in doubt, I assumed that OpenIsis's implementation was right |
584 |
|
(except for obvious bugs). |
585 |
|
|
586 |
|
However, every effort has been made to test this module with as much |
587 |
|
databases (and programs that create them) as possible. |
588 |
|
|
589 |
|
I would be very greatful for success or failure reports about usage of this |
590 |
|
module with databases from programs other than WinIsis and IsisMarc. I had |
591 |
|
tested this against ouput of one C<isis.dll>-based application, but I don't |
592 |
|
know any details about it's version. |
593 |
|
|
594 |
=head1 AUTHOR |
=head1 AUTHOR |
595 |
|
|