1 |
package IsisDB; |
package Biblio::Isis; |
2 |
use strict; |
use strict; |
3 |
|
|
4 |
use Carp; |
use Carp; |
5 |
use File::Glob qw(:globally :nocase); |
use File::Glob qw(:globally :nocase); |
6 |
|
|
|
use Data::Dumper; |
|
|
|
|
7 |
BEGIN { |
BEGIN { |
8 |
use Exporter (); |
use Exporter (); |
9 |
use vars qw ($VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS); |
use vars qw ($VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS); |
10 |
$VERSION = 0.08; |
$VERSION = 0.23; |
11 |
@ISA = qw (Exporter); |
@ISA = qw (Exporter); |
12 |
#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 |
13 |
@EXPORT = qw (); |
@EXPORT = qw (); |
18 |
|
|
19 |
=head1 NAME |
=head1 NAME |
20 |
|
|
21 |
IsisDB - Read CDS/ISIS, WinISIS and IsisMarc database |
Biblio::Isis - Read CDS/ISIS, WinISIS and IsisMarc database |
22 |
|
|
23 |
=head1 SYNOPSIS |
=head1 SYNOPSIS |
24 |
|
|
25 |
use IsisDB; |
use Biblio::Isis; |
26 |
|
|
27 |
my $isis = new IsisDB( |
my $isis = new Biblio::Isis( |
28 |
isisdb => './cds/cds', |
isisdb => './cds/cds', |
29 |
); |
); |
30 |
|
|
31 |
for(my $mfn = 1; $mfn <= $isis->{'maxmfn'}; $mfn++) { |
for(my $mfn = 1; $mfn <= $isis->count; $mfn++) { |
32 |
print $isis->to_ascii($mfn),"\n"; |
print $isis->to_ascii($mfn),"\n"; |
33 |
} |
} |
34 |
|
|
35 |
=head1 DESCRIPTION |
=head1 DESCRIPTION |
36 |
|
|
37 |
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 |
38 |
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 |
39 |
|
seems to depriciate it's old C<XS> bindings for perl. |
40 |
|
|
41 |
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>), |
42 |
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 |
49 |
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 |
50 |
IsisMarc), see C<to_hash>. |
IsisMarc), see C<to_hash>. |
51 |
|
|
52 |
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 |
53 |
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 |
54 |
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 |
55 |
creates data structures which are easier to use than ones created by |
creates data structures which are easier to use than ones created by |
79 |
|
|
80 |
Open ISIS database |
Open ISIS database |
81 |
|
|
82 |
my $isis = new IsisDB( |
my $isis = new Biblio::Isis( |
83 |
isisdb => './cds/cds', |
isisdb => './cds/cds', |
84 |
read_fdt => 1, |
read_fdt => 1, |
85 |
include_deleted => 1, |
include_deleted => 1, |
86 |
hash_filter => sub { |
hash_filter => sub { |
87 |
my $v = shift; |
my ($v,$field_number) = @_; |
88 |
$v =~ s#foo#bar#g; |
$v =~ s#foo#bar#g; |
89 |
}, |
}, |
90 |
debug => 1, |
debug => 1, |
91 |
|
join_subfields_with => ' ; ', |
92 |
); |
); |
93 |
|
|
94 |
Options are described below: |
Options are described below: |
114 |
|
|
115 |
=item hash_filter |
=item hash_filter |
116 |
|
|
117 |
Filter code ref which will be used before data is converted to hash. |
Filter code ref which will be used before data is converted to hash. It will |
118 |
|
receive two arguments, whole line from current field (in C<< $_[0] >>) and |
119 |
|
field number (in C<< $_[1] >>). |
120 |
|
|
121 |
=item debug |
=item debug |
122 |
|
|
123 |
Dump a B<lot> of debugging output. |
Dump a B<lot> of debugging output even at level 1. For even more increase level. |
124 |
|
|
125 |
=back |
=item join_subfields_with |
126 |
|
|
127 |
It will also set C<$isis-E<gt>{'maxmfn'}> which is maximum MFN stored in database. |
Define delimiter which will be used to join repeatable subfields. This |
128 |
|
option is included to support lagacy application written against version |
129 |
|
older than 0.21 of this module. By default, it disabled. See L</to_hash>. |
130 |
|
|
131 |
|
=back |
132 |
|
|
133 |
=cut |
=cut |
134 |
|
|
139 |
|
|
140 |
croak "new needs database name (isisdb) as argument!" unless ({@_}->{isisdb}); |
croak "new needs database name (isisdb) as argument!" unless ({@_}->{isisdb}); |
141 |
|
|
142 |
foreach my $v (qw{isisdb debug include_deleted hash_filter}) { |
foreach my $v (qw{isisdb debug include_deleted hash_filter join_subfields_with}) { |
143 |
$self->{$v} = {@_}->{$v}; |
$self->{$v} = {@_}->{$v} if defined({@_}->{$v}); |
144 |
} |
} |
145 |
|
|
146 |
my @isis_files = grep(/\.(FDT|MST|XRF|CNT)$/i,glob($self->{isisdb}."*")); |
my @isis_files = grep(/\.(FDT|MST|XRF|CNT)$/i,glob($self->{isisdb}."*")); |
154 |
push @must_exist, "fdt" if ($self->{read_fdt}); |
push @must_exist, "fdt" if ($self->{read_fdt}); |
155 |
|
|
156 |
foreach my $ext (@must_exist) { |
foreach my $ext (@must_exist) { |
157 |
croak "missing ",uc($ext)," file in ",$self->{isisdb} unless ($self->{$ext."_file"}); |
unless ($self->{$ext."_file"}) { |
158 |
|
carp "missing ",uc($ext)," file in ",$self->{isisdb}; |
159 |
|
return; |
160 |
|
} |
161 |
} |
} |
162 |
|
|
163 |
print STDERR "## using files: ",join(" ",@isis_files),"\n" if ($self->{debug}); |
if ($self->{debug}) { |
164 |
|
print STDERR "## using files: ",join(" ",@isis_files),"\n"; |
165 |
|
eval "use Data::Dump"; |
166 |
|
|
167 |
|
if (! $@) { |
168 |
|
*Dumper = *Data::Dump::dump; |
169 |
|
} else { |
170 |
|
use Data::Dumper; |
171 |
|
} |
172 |
|
} |
173 |
|
|
174 |
# if you want to read .FDT file use read_fdt argument when creating class! |
# if you want to read .FDT file use read_fdt argument when creating class! |
175 |
if ($self->{read_fdt} && -e $self->{fdt_file}) { |
if ($self->{read_fdt} && -e $self->{fdt_file}) { |
177 |
# read the $db.FDT file for tags |
# read the $db.FDT file for tags |
178 |
my $fieldzone=0; |
my $fieldzone=0; |
179 |
|
|
180 |
open(fileFDT, $self->{fdt_file}) || croak "can't read '$self->{fdt_file}': $!"; |
open(my $fileFDT, $self->{fdt_file}) || croak "can't read '$self->{fdt_file}': $!"; |
181 |
|
binmode($fileFDT); |
182 |
|
|
183 |
while (<fileFDT>) { |
while (<$fileFDT>) { |
184 |
chomp; |
chomp; |
185 |
if ($fieldzone) { |
if ($fieldzone) { |
186 |
my $name=substr($_,0,30); |
my $name=substr($_,0,30); |
197 |
} |
} |
198 |
} |
} |
199 |
|
|
200 |
close(fileFDT); |
close($fileFDT); |
201 |
} |
} |
202 |
|
|
203 |
# Get the Maximum MFN from $db.MST |
# Get the Maximum MFN from $db.MST |
204 |
|
|
205 |
open($self->{'fileMST'}, $self->{mst_file}) || croak "can't open '$self->{mst_file}': $!"; |
open($self->{'fileMST'}, $self->{mst_file}) || croak "can't open '$self->{mst_file}': $!"; |
206 |
|
binmode($self->{'fileMST'}); |
207 |
|
|
208 |
# MST format: (* = 32 bit signed) |
# MST format: (* = 32 bit signed) |
209 |
# CTLMFN* always 0 |
# CTLMFN* always 0 |
211 |
# NXTMFB* last block allocated to master file |
# NXTMFB* last block allocated to master file |
212 |
# NXTMFP offset to next available position in last block |
# NXTMFP offset to next available position in last block |
213 |
# MFTYPE always 0 for user db file (1 for system) |
# MFTYPE always 0 for user db file (1 for system) |
214 |
seek($self->{'fileMST'},4,0); |
seek($self->{'fileMST'},4,0) || croak "can't seek to offset 0 in MST: $!"; |
215 |
|
|
216 |
my $buff; |
my $buff; |
217 |
|
|
218 |
read($self->{'fileMST'}, $buff, 4); |
read($self->{'fileMST'}, $buff, 4) || croak "can't read NXTMFN from MST: $!"; |
219 |
$self->{'NXTMFN'}=unpack("l",$buff) || carp "NXTNFN is zero"; |
$self->{'NXTMFN'}=unpack("V",$buff) || croak "NXTNFN is zero"; |
|
|
|
|
# save maximum MFN |
|
|
$self->{'maxmfn'} = $self->{'NXTMFN'} - 1; |
|
|
|
|
|
|
|
220 |
|
|
221 |
|
print STDERR "## self ",Dumper($self),"\n" if ($self->{debug}); |
|
print STDERR Dumper($self),"\n" if ($self->{debug}); |
|
222 |
|
|
223 |
# open files for later |
# open files for later |
224 |
open($self->{'fileXRF'}, $self->{xrf_file}) || croak "can't open '$self->{xrf_file}': $!"; |
open($self->{'fileXRF'}, $self->{xrf_file}) || croak "can't open '$self->{xrf_file}': $!"; |
225 |
|
binmode($self->{'fileXRF'}); |
226 |
|
|
227 |
$self ? return $self : return undef; |
$self ? return $self : return undef; |
228 |
} |
} |
229 |
|
|
230 |
=head2 read_cnt |
=head2 count |
231 |
|
|
232 |
This function is not really used by module, but can be useful to find info |
Return number of records in database |
|
about your index (if debugging it for example). |
|
233 |
|
|
234 |
print Dumper($isis->read_cnt); |
print $isis->count; |
235 |
|
|
236 |
=cut |
=cut |
237 |
|
|
238 |
sub read_cnt { |
sub count { |
239 |
my $self = shift; |
my $self = shift; |
240 |
|
return $self->{'NXTMFN'} - 1; |
|
croak "missing CNT file in ",$self->{isisdb} unless ($self->{cnt_file}); |
|
|
|
|
|
# Get the index information from $db.CNT |
|
|
|
|
|
open(fileCNT, $self->{cnt_file}) || croak "can't read '$self->{cnt_file}': $!"; |
|
|
|
|
|
# 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); |
|
|
} |
|
|
} |
|
|
|
|
|
my $buff; |
|
|
|
|
|
read(fileCNT, $buff, 26); |
|
|
$self->unpack_cnt($buff); |
|
|
|
|
|
read(fileCNT, $buff, 26); |
|
|
$self->unpack_cnt($buff); |
|
|
|
|
|
close(fileCNT); |
|
|
|
|
|
return $self->{cnt}; |
|
241 |
} |
} |
242 |
|
|
243 |
=head2 fetch |
=head2 fetch |
280 |
|
|
281 |
# read XRFMFB abd XRFMFP |
# read XRFMFB abd XRFMFP |
282 |
read($self->{'fileXRF'}, $buff, 4); |
read($self->{'fileXRF'}, $buff, 4); |
283 |
my $pointer=unpack("l",$buff) || carp "pointer is null"; |
my $pointer=unpack("V",$buff); |
284 |
|
if (! $pointer) { |
285 |
|
if ($self->{include_deleted}) { |
286 |
|
return; |
287 |
|
} else { |
288 |
|
warn "pointer for MFN $mfn is null\n"; |
289 |
|
return; |
290 |
|
} |
291 |
|
} |
292 |
|
|
293 |
# check for logically deleted record |
# check for logically deleted record |
294 |
if ($pointer < 0) { |
if ($pointer & 0x80000000) { |
295 |
print STDERR "## record $mfn is logically deleted\n" if ($self->{debug}); |
print STDERR "## record $mfn is logically deleted\n" if ($self->{debug}); |
296 |
$self->{deleted} = $mfn; |
$self->{deleted} = $mfn; |
297 |
|
|
298 |
return unless $self->{include_deleted}; |
return unless $self->{include_deleted}; |
299 |
|
|
300 |
$pointer = abs($pointer); |
# abs |
301 |
|
$pointer = ($pointer ^ 0xffffffff) + 1; |
302 |
} |
} |
303 |
|
|
304 |
my $XRFMFB = int($pointer/2048); |
my $XRFMFB = int($pointer/2048); |
307 |
# (XRFMFB - 1) * 512 + XRFMFP |
# (XRFMFB - 1) * 512 + XRFMFP |
308 |
# why do i have to do XRFMFP % 1024 ? |
# why do i have to do XRFMFP % 1024 ? |
309 |
|
|
310 |
my $blk_off = (($XRFMFB - 1) * 512) + ($XRFMFP % 1024); |
my $blk_off = (($XRFMFB - 1) * 512) + ($XRFMFP % 512); |
311 |
|
|
312 |
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'}); |
313 |
|
|
314 |
# Get Record Information |
# Get Record Information |
315 |
|
|
316 |
seek($self->{'fileMST'},$blk_off,0); |
seek($self->{'fileMST'},$blk_off,0) || croak "can't seek to $blk_off: $!"; |
317 |
|
|
318 |
read($self->{'fileMST'}, $buff, 4); |
read($self->{'fileMST'}, $buff, 4) || croak "can't read 4 bytes at offset $blk_off from MST file: $!"; |
319 |
my $value=unpack("l",$buff); |
my $value=unpack("V",$buff); |
320 |
|
|
321 |
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}); |
322 |
|
|
323 |
if ($value!=$mfn) { |
if ($value!=$mfn) { |
324 |
carp "Error: MFN ".$mfn." not found in MST(".$value.")"; |
if ($value == 0) { |
325 |
#return; |
print STDERR "## record $mfn is physically deleted\n" if ($self->{debug}); |
326 |
|
$self->{deleted} = $mfn; |
327 |
|
return; |
328 |
|
} |
329 |
|
|
330 |
|
carp "Error: MFN ".$mfn." not found in MST file, found $value"; |
331 |
|
return; |
332 |
} |
} |
333 |
|
|
334 |
read($self->{'fileMST'}, $buff, 14); |
read($self->{'fileMST'}, $buff, 14); |
335 |
|
|
336 |
my ($MFRL,$MFBWB,$MFBWP,$BASE,$NVF,$STATUS) = unpack("slssss", $buff); |
my ($MFRL,$MFBWB,$MFBWP,$BASE,$NVF,$STATUS) = unpack("vVvvvv", $buff); |
337 |
|
|
338 |
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}); |
339 |
|
|
353 |
|
|
354 |
for (my $i = 0 ; $i < $NVF ; $i++) { |
for (my $i = 0 ; $i < $NVF ; $i++) { |
355 |
|
|
356 |
my ($TAG,$POS,$LEN) = unpack("sss", substr($buff,$i * 6, 6)); |
my ($TAG,$POS,$LEN) = unpack("vvv", substr($buff,$i * 6, 6)); |
357 |
|
|
358 |
print STDERR "## TAG: $TAG POS: $POS LEN: $LEN\n" if ($self->{debug}); |
print STDERR "## TAG: $TAG POS: $POS LEN: $LEN\n" if ($self->{debug}); |
359 |
|
|
394 |
return $self->{'record'}; |
return $self->{'record'}; |
395 |
} |
} |
396 |
|
|
397 |
|
=head2 mfn |
398 |
|
|
399 |
|
Returns current MFN position |
400 |
|
|
401 |
|
my $mfn = $isis->mfn; |
402 |
|
|
403 |
|
=cut |
404 |
|
|
405 |
|
# This function should be simple return $self->{current_mfn}, |
406 |
|
# but if new is called with _hack_mfn it becomes setter. |
407 |
|
# It's useful in tests when setting $isis->{record} directly |
408 |
|
|
409 |
|
sub mfn { |
410 |
|
my $self = shift; |
411 |
|
return $self->{current_mfn}; |
412 |
|
}; |
413 |
|
|
414 |
|
|
415 |
=head2 to_ascii |
=head2 to_ascii |
416 |
|
|
417 |
Dump ASCII output of record with specified MFN |
Returns ASCII output of record with specified MFN |
418 |
|
|
419 |
print $isis->to_ascii(42); |
print $isis->to_ascii(42); |
420 |
|
|
421 |
It outputs something like this: |
This outputs something like this: |
422 |
|
|
423 |
210 ^aNew York^cNew York University press^dcop. 1988 |
210 ^aNew York^cNew York University press^dcop. 1988 |
424 |
990 2140 |
990 2140 |
435 |
|
|
436 |
my $mfn = shift || croak "need MFN"; |
my $mfn = shift || croak "need MFN"; |
437 |
|
|
438 |
my $rec = $self->fetch($mfn); |
my $rec = $self->fetch($mfn) || return; |
439 |
|
|
440 |
my $out = "0\t$mfn"; |
my $out = "0\t$mfn"; |
441 |
|
|
455 |
|
|
456 |
my $hash = $isis->to_hash($mfn); |
my $hash = $isis->to_hash($mfn); |
457 |
|
|
458 |
It has ability to convert characters (using C<hash_filter> from ISIS |
It has ability to convert characters (using C<hash_filter>) from ISIS |
459 |
database before creating structures enabling character re-mapping or quick |
database before creating structures enabling character re-mapping or quick |
460 |
fix-up of data. |
fix-up of data. |
461 |
|
|
491 |
} |
} |
492 |
], |
], |
493 |
|
|
494 |
|
In case there are repeatable subfields in record, this will create |
495 |
|
following structure: |
496 |
|
|
497 |
|
'900' => [ { |
498 |
|
'a' => [ 'foo', 'bar', 'baz' ], |
499 |
|
}] |
500 |
|
|
501 |
|
Or in more complex example of |
502 |
|
|
503 |
|
902 ^aa1^aa2^aa3^bb1^aa4^bb2^cc1^aa5 |
504 |
|
|
505 |
|
it will create |
506 |
|
|
507 |
|
902 => [ |
508 |
|
{ a => ["a1", "a2", "a3", "a4", "a5"], b => ["b1", "b2"], c => "c1" }, |
509 |
|
], |
510 |
|
|
511 |
|
This behaviour can be changed using C<join_subfields_with> option to L</new>, |
512 |
|
in which case C<to_hash> will always create single value for each subfield. |
513 |
|
This will change result to: |
514 |
|
|
515 |
|
|
516 |
|
|
517 |
This method will also create additional field C<000> with MFN. |
This method will also create additional field C<000> with MFN. |
518 |
|
|
519 |
|
There is also more elaborative way to call C<to_hash> like this: |
520 |
|
|
521 |
|
my $hash = $isis->to_hash({ |
522 |
|
mfn => 42, |
523 |
|
include_subfields => 1, |
524 |
|
}); |
525 |
|
|
526 |
|
Each option controll creation of hash: |
527 |
|
|
528 |
|
=over 4 |
529 |
|
|
530 |
|
=item mfn |
531 |
|
|
532 |
|
Specify MFN number of record |
533 |
|
|
534 |
|
=item include_subfields |
535 |
|
|
536 |
|
This option will create additional key in hash called C<subfields> which will |
537 |
|
have original record subfield order and index to that subfield like this: |
538 |
|
|
539 |
|
902 => [ { |
540 |
|
a => ["a1", "a2", "a3", "a4", "a5"], |
541 |
|
b => ["b1", "b2"], |
542 |
|
c => "c1", |
543 |
|
subfields => ["a", 0, "a", 1, "a", 2, "b", 0, "a", 3, "b", 1, "c", 0, "a", 4], |
544 |
|
} ], |
545 |
|
|
546 |
|
=item join_subfields_with |
547 |
|
|
548 |
|
Define delimiter which will be used to join repeatable subfields. You can |
549 |
|
specify option here instead in L</new> if you want to have per-record control. |
550 |
|
|
551 |
|
=item hash_filter |
552 |
|
|
553 |
|
You can override C<hash_filter> defined in L</new> using this option. |
554 |
|
|
555 |
|
=back |
556 |
|
|
557 |
=cut |
=cut |
558 |
|
|
559 |
sub to_hash { |
sub to_hash { |
560 |
my $self = shift; |
my $self = shift; |
561 |
|
|
562 |
|
|
563 |
my $mfn = shift || confess "need mfn!"; |
my $mfn = shift || confess "need mfn!"; |
564 |
|
my $arg; |
565 |
|
|
566 |
|
my $hash_filter = $self->{hash_filter}; |
567 |
|
|
568 |
|
if (ref($mfn) eq 'HASH') { |
569 |
|
$arg = $mfn; |
570 |
|
$mfn = $arg->{mfn} || confess "need mfn in arguments"; |
571 |
|
$hash_filter = $arg->{hash_filter} if ($arg->{hash_filter}); |
572 |
|
} |
573 |
|
|
574 |
# init record to include MFN as field 000 |
# init record to include MFN as field 000 |
575 |
my $rec = { '000' => [ $mfn ] }; |
my $rec = { '000' => [ $mfn ] }; |
576 |
|
|
577 |
my $row = $self->fetch($mfn); |
my $row = $self->fetch($mfn) || return; |
578 |
|
|
579 |
|
my $j_rs = $arg->{join_subfields_with} || $self->{join_subfields_with}; |
580 |
|
$j_rs = $self->{join_subfields_with} unless(defined($j_rs)); |
581 |
|
my $i_sf = $arg->{include_subfields}; |
582 |
|
|
583 |
foreach my $k (keys %{$row}) { |
foreach my $f_nr (keys %{$row}) { |
584 |
foreach my $l (@{$row->{$k}}) { |
foreach my $l (@{$row->{$f_nr}}) { |
585 |
|
|
586 |
# filter output |
# filter output |
587 |
$l = $self->{'hash_filter'}->($l) if ($self->{'hash_filter'}); |
$l = $hash_filter->($l, $f_nr) if ($hash_filter); |
588 |
|
next unless defined($l); |
589 |
|
|
590 |
my $val; |
my $val; |
591 |
|
my $r_sf; # repeatable subfields in this record |
592 |
|
|
593 |
# has identifiers? |
# has identifiers? |
594 |
($val->{'i1'},$val->{'i2'}) = ($1,$2) if ($l =~ s/^([01 #])([01 #])\^/\^/); |
($val->{'i1'},$val->{'i2'}) = ($1,$2) if ($l =~ s/^([01 #])([01 #])\^/\^/); |
597 |
if ($l =~ m/\^/) { |
if ($l =~ m/\^/) { |
598 |
foreach my $t (split(/\^/,$l)) { |
foreach my $t (split(/\^/,$l)) { |
599 |
next if (! $t); |
next if (! $t); |
600 |
$val->{substr($t,0,1)} = substr($t,1); |
my ($sf,$v) = (substr($t,0,1), substr($t,1)); |
601 |
|
# XXX this might be option, but why? |
602 |
|
next unless (defined($v) && $v ne ''); |
603 |
|
# warn "### $f_nr^$sf:$v",$/ if ($self->{debug} > 1); |
604 |
|
|
605 |
|
if (ref( $val->{$sf} ) eq 'ARRAY') { |
606 |
|
|
607 |
|
push @{ $val->{$sf} }, $v; |
608 |
|
|
609 |
|
# record repeatable subfield it it's offset |
610 |
|
push @{ $val->{subfields} }, ( $sf, $#{ $val->{$sf} } ) if (! $j_rs && $i_sf); |
611 |
|
$r_sf->{$sf}++; |
612 |
|
|
613 |
|
} elsif (defined( $val->{$sf} )) { |
614 |
|
|
615 |
|
# convert scalar field to array |
616 |
|
$val->{$sf} = [ $val->{$sf}, $v ]; |
617 |
|
|
618 |
|
push @{ $val->{subfields} }, ( $sf, 1 ) if (! $j_rs && $i_sf); |
619 |
|
$r_sf->{$sf}++; |
620 |
|
|
621 |
|
} else { |
622 |
|
$val->{$sf} = $v; |
623 |
|
push @{ $val->{subfields} }, ( $sf, 0 ) if ($i_sf); |
624 |
|
} |
625 |
} |
} |
626 |
} else { |
} else { |
627 |
$val = $l; |
$val = $l; |
628 |
} |
} |
629 |
|
|
630 |
push @{$rec->{$k}}, $val; |
if ($j_rs) { |
631 |
|
map { |
632 |
|
$val->{$_} = join($j_rs, @{ $val->{$_} }); |
633 |
|
} keys %$r_sf |
634 |
|
} |
635 |
|
|
636 |
|
push @{$rec->{$f_nr}}, $val; |
637 |
} |
} |
638 |
} |
} |
639 |
|
|
654 |
return $self->{'TagName'}->{$tag} || $tag; |
return $self->{'TagName'}->{$tag} || $tag; |
655 |
} |
} |
656 |
|
|
657 |
|
|
658 |
|
=head2 read_cnt |
659 |
|
|
660 |
|
Read content of C<.CNT> file and return hash containing it. |
661 |
|
|
662 |
|
print Dumper($isis->read_cnt); |
663 |
|
|
664 |
|
This function is not used by module (C<.CNT> files are not required for this |
665 |
|
module to work), but it can be useful to examine your index (while debugging |
666 |
|
for example). |
667 |
|
|
668 |
|
=cut |
669 |
|
|
670 |
|
sub read_cnt { |
671 |
|
my $self = shift; |
672 |
|
|
673 |
|
croak "missing CNT file in ",$self->{isisdb} unless ($self->{cnt_file}); |
674 |
|
|
675 |
|
# Get the index information from $db.CNT |
676 |
|
|
677 |
|
open(my $fileCNT, $self->{cnt_file}) || croak "can't read '$self->{cnt_file}': $!"; |
678 |
|
binmode($fileCNT); |
679 |
|
|
680 |
|
my $buff; |
681 |
|
|
682 |
|
read($fileCNT, $buff, 26) || croak "can't read first table from CNT: $!"; |
683 |
|
$self->unpack_cnt($buff); |
684 |
|
|
685 |
|
read($fileCNT, $buff, 26) || croak "can't read second table from CNT: $!"; |
686 |
|
$self->unpack_cnt($buff); |
687 |
|
|
688 |
|
close($fileCNT); |
689 |
|
|
690 |
|
return $self->{cnt}; |
691 |
|
} |
692 |
|
|
693 |
|
=head2 unpack_cnt |
694 |
|
|
695 |
|
Unpack one of two 26 bytes fixed length record in C<.CNT> file. |
696 |
|
|
697 |
|
Here is definition of record: |
698 |
|
|
699 |
|
off key description size |
700 |
|
0: IDTYPE BTree type s |
701 |
|
2: ORDN Nodes Order s |
702 |
|
4: ORDF Leafs Order s |
703 |
|
6: N Number of Memory buffers for nodes s |
704 |
|
8: K Number of buffers for first level index s |
705 |
|
10: LIV Current number of Index Levels s |
706 |
|
12: POSRX Pointer to Root Record in N0x l |
707 |
|
16: NMAXPOS Next Available position in N0x l |
708 |
|
20: FMAXPOS Next available position in L0x l |
709 |
|
24: ABNORMAL Formal BTree normality indicator s |
710 |
|
length: 26 bytes |
711 |
|
|
712 |
|
This will fill C<$self> object under C<cnt> with hash. It's used by C<read_cnt>. |
713 |
|
|
714 |
|
=cut |
715 |
|
|
716 |
|
sub unpack_cnt { |
717 |
|
my $self = shift; |
718 |
|
|
719 |
|
my @flds = qw(ORDN ORDF N K LIV POSRX NMAXPOS FMAXPOS ABNORMAL); |
720 |
|
|
721 |
|
my $buff = shift || return; |
722 |
|
my @arr = unpack("vvvvvvVVVv", $buff); |
723 |
|
|
724 |
|
print STDERR "unpack_cnt: ",join(" ",@arr),"\n" if ($self->{'debug'}); |
725 |
|
|
726 |
|
my $IDTYPE = shift @arr; |
727 |
|
foreach (@flds) { |
728 |
|
$self->{cnt}->{$IDTYPE}->{$_} = abs(shift @arr); |
729 |
|
} |
730 |
|
} |
731 |
|
|
732 |
1; |
1; |
733 |
|
|
734 |
=head1 BUGS |
=head1 BUGS |
735 |
|
|
736 |
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 |
737 |
|
some variations in input databases which has been tested with this module. |
738 |
|
When I was in doubt, I assumed that OpenIsis's implementation was right |
739 |
|
(except for obvious bugs). |
740 |
|
|
741 |
|
However, every effort has been made to test this module with as much |
742 |
|
databases (and programs that create them) as possible. |
743 |
|
|
744 |
|
I would be very greatful for success or failure reports about usage of this |
745 |
|
module with databases from programs other than WinIsis and IsisMarc. I had |
746 |
|
tested this against ouput of one C<isis.dll>-based application, but I don't |
747 |
|
know any details about it's version. |
748 |
|
|
749 |
|
=head1 VERSIONS |
750 |
|
|
751 |
|
As this is young module, new features are added in subsequent version. It's |
752 |
|
a good idea to specify version when using this module like this: |
753 |
|
|
754 |
|
use Biblio::Isis 0.23 |
755 |
|
|
756 |
|
Below is list of changes in specific version of module (so you can target |
757 |
|
older versions if you really have to): |
758 |
|
|
759 |
|
=over 8 |
760 |
|
|
761 |
|
=item 0.23 |
762 |
|
|
763 |
|
Added C<hash_filter> to L</to_hash> |
764 |
|
|
765 |
|
Fixed bug with documented C<join_subfields_with> in L</new> which wasn't |
766 |
|
implemented |
767 |
|
|
768 |
|
=item 0.22 |
769 |
|
|
770 |
|
Added field number when calling C<hash_filter> |
771 |
|
|
772 |
|
=item 0.21 |
773 |
|
|
774 |
|
Added C<join_subfields_with> to L</new> and L</to_hash>. |
775 |
|
|
776 |
|
Added C<include_subfields> to L</to_hash>. |
777 |
|
|
778 |
|
=item 0.20 |
779 |
|
|
780 |
|
Added C<< $isis->mfn >>, support for repeatable subfields and |
781 |
|
C<< $isis->to_hash({ mfn => 42, ... }) >> calling convention |
782 |
|
|
783 |
|
=back |
784 |
|
|
785 |
=head1 AUTHOR |
=head1 AUTHOR |
786 |
|
|
803 |
|
|
804 |
=head1 SEE ALSO |
=head1 SEE ALSO |
805 |
|
|
806 |
|
L<Biblio::Isis::Manual> for CDS/ISIS manual appendix F, G and H which describe file format |
807 |
|
|
808 |
OpenIsis web site L<http://www.openisis.org> |
OpenIsis web site L<http://www.openisis.org> |
809 |
|
|
810 |
perl4lib site L<http://perl4lib.perl.org> |
perl4lib site L<http://perl4lib.perl.org> |