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 = 0.07; |
$VERSION = 0.22_2; |
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, |
88 |
$v =~ s#foo#bar#g; |
$v =~ s#foo#bar#g; |
89 |
}, |
}, |
90 |
debug => 1, |
debug => 1, |
91 |
|
join_subfields_with => ' ; ', |
92 |
|
regexps => [ |
93 |
|
's/something/else/g', |
94 |
|
], |
95 |
); |
); |
96 |
|
|
97 |
Options are described below: |
Options are described below: |
121 |
|
|
122 |
=item debug |
=item debug |
123 |
|
|
124 |
Dump a B<lot> of debugging output. |
Dump a B<lot> of debugging output even at level 1. For even more increase level. |
125 |
|
|
126 |
=back |
=item join_subfields_with |
127 |
|
|
128 |
|
Define delimiter which will be used to join repeatable subfields. This |
129 |
|
option is included to support lagacy application written against version |
130 |
|
older than 0.21 of this module. By default, it disabled. See L</to_hash>. |
131 |
|
|
132 |
It will also set C<$isis-E<gt>{'maxmfn'}> which is maximum MFN stored in database. |
=item regexpes |
133 |
|
|
134 |
|
Define (any number) of regexpes to apply at field values before they are |
135 |
|
splitted into subfield. This is great place to split subfields in input to |
136 |
|
mulitple subfields if needed or rename subfields. |
137 |
|
|
138 |
|
=back |
139 |
|
|
140 |
=cut |
=cut |
141 |
|
|
161 |
push @must_exist, "fdt" if ($self->{read_fdt}); |
push @must_exist, "fdt" if ($self->{read_fdt}); |
162 |
|
|
163 |
foreach my $ext (@must_exist) { |
foreach my $ext (@must_exist) { |
164 |
croak "missing ",uc($ext)," file in ",$self->{isisdb} unless ($self->{$ext."_file"}); |
unless ($self->{$ext."_file"}) { |
165 |
|
carp "missing ",uc($ext)," file in ",$self->{isisdb}; |
166 |
|
return; |
167 |
|
} |
168 |
} |
} |
169 |
|
|
170 |
print STDERR "## using files: ",join(" ",@isis_files),"\n" if ($self->{debug}); |
if ($self->{debug}) { |
171 |
|
print STDERR "## using files: ",join(" ",@isis_files),"\n"; |
172 |
|
eval "use Data::Dump"; |
173 |
|
|
174 |
|
if (! $@) { |
175 |
|
*Dumper = *Data::Dump::dump; |
176 |
|
} else { |
177 |
|
use Data::Dumper; |
178 |
|
} |
179 |
|
} |
180 |
|
|
181 |
# if you want to read .FDT file use read_fdt argument when creating class! |
# if you want to read .FDT file use read_fdt argument when creating class! |
182 |
if ($self->{read_fdt} && -e $self->{fdt_file}) { |
if ($self->{read_fdt} && -e $self->{fdt_file}) { |
184 |
# read the $db.FDT file for tags |
# read the $db.FDT file for tags |
185 |
my $fieldzone=0; |
my $fieldzone=0; |
186 |
|
|
187 |
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}': $!"; |
188 |
|
binmode($fileFDT); |
189 |
|
|
190 |
while (<fileFDT>) { |
while (<$fileFDT>) { |
191 |
chomp; |
chomp; |
192 |
if ($fieldzone) { |
if ($fieldzone) { |
193 |
my $name=substr($_,0,30); |
my $name=substr($_,0,30); |
204 |
} |
} |
205 |
} |
} |
206 |
|
|
207 |
close(fileFDT); |
close($fileFDT); |
208 |
} |
} |
209 |
|
|
210 |
# Get the Maximum MFN from $db.MST |
# Get the Maximum MFN from $db.MST |
211 |
|
|
212 |
open($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}': $!"; |
213 |
|
binmode($self->{'fileMST'}); |
214 |
|
|
215 |
# MST format: (* = 32 bit signed) |
# MST format: (* = 32 bit signed) |
216 |
# CTLMFN* always 0 |
# CTLMFN* always 0 |
218 |
# NXTMFB* last block allocated to master file |
# NXTMFB* last block allocated to master file |
219 |
# NXTMFP offset to next available position in last block |
# NXTMFP offset to next available position in last block |
220 |
# MFTYPE always 0 for user db file (1 for system) |
# MFTYPE always 0 for user db file (1 for system) |
221 |
seek($self->{'fileMST'},4,0); |
seek($self->{'fileMST'},4,0) || croak "can't seek to offset 0 in MST: $!"; |
222 |
|
|
223 |
my $buff; |
my $buff; |
224 |
|
|
225 |
read($self->{'fileMST'}, $buff, 4); |
read($self->{'fileMST'}, $buff, 4) || croak "can't read NXTMFN from MST: $!"; |
226 |
$self->{'NXTMFN'}=unpack("l",$buff) || carp "NXTNFN is zero"; |
$self->{'NXTMFN'}=unpack("V",$buff) || croak "NXTNFN is zero"; |
|
|
|
|
# save maximum MFN |
|
|
$self->{'maxmfn'} = $self->{'NXTMFN'} - 1; |
|
|
|
|
227 |
|
|
228 |
|
print STDERR "## self ",Dumper($self),"\n" if ($self->{debug}); |
|
|
|
|
print STDERR Dumper($self),"\n" if ($self->{debug}); |
|
229 |
|
|
230 |
# open files for later |
# open files for later |
231 |
open($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}': $!"; |
232 |
|
binmode($self->{'fileXRF'}); |
233 |
|
|
234 |
$self ? return $self : return undef; |
$self ? return $self : return undef; |
235 |
} |
} |
236 |
|
|
237 |
=head2 read_cnt |
=head2 count |
238 |
|
|
239 |
This function is not really used by module, but can be useful to find info |
Return number of records in database |
|
about your index (if debugging it for example). |
|
240 |
|
|
241 |
print Dumper($isis->read_cnt); |
print $isis->count; |
242 |
|
|
243 |
=cut |
=cut |
244 |
|
|
245 |
sub read_cnt { |
sub count { |
246 |
my $self = shift; |
my $self = shift; |
247 |
|
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}; |
|
248 |
} |
} |
249 |
|
|
250 |
=head2 fetch |
=head2 fetch |
270 |
|
|
271 |
# is mfn allready in memory? |
# is mfn allready in memory? |
272 |
my $old_mfn = $self->{'current_mfn'} || -1; |
my $old_mfn = $self->{'current_mfn'} || -1; |
273 |
return if ($mfn == $old_mfn); |
return $self->{record} if ($mfn == $old_mfn); |
274 |
|
|
275 |
print STDERR "## fetch: $mfn\n" if ($self->{debug}); |
print STDERR "## fetch: $mfn\n" if ($self->{debug}); |
276 |
|
|
282 |
|
|
283 |
my $buff; |
my $buff; |
284 |
|
|
285 |
|
# delete old record |
286 |
|
delete $self->{record}; |
287 |
|
|
288 |
# read XRFMFB abd XRFMFP |
# read XRFMFB abd XRFMFP |
289 |
read($self->{'fileXRF'}, $buff, 4); |
read($self->{'fileXRF'}, $buff, 4); |
290 |
my $pointer=unpack("l",$buff) || carp "pointer is null"; |
my $pointer=unpack("V",$buff); |
291 |
|
if (! $pointer) { |
292 |
|
if ($self->{include_deleted}) { |
293 |
|
return; |
294 |
|
} else { |
295 |
|
warn "pointer for MFN $mfn is null\n"; |
296 |
|
return; |
297 |
|
} |
298 |
|
} |
299 |
|
|
300 |
|
# check for logically deleted record |
301 |
|
if ($pointer & 0x80000000) { |
302 |
|
print STDERR "## record $mfn is logically deleted\n" if ($self->{debug}); |
303 |
|
$self->{deleted} = $mfn; |
304 |
|
|
305 |
|
return unless $self->{include_deleted}; |
306 |
|
|
307 |
|
# abs |
308 |
|
$pointer = ($pointer ^ 0xffffffff) + 1; |
309 |
|
} |
310 |
|
|
311 |
my $XRFMFB = int($pointer/2048); |
my $XRFMFB = int($pointer/2048); |
312 |
my $XRFMFP = $pointer - ($XRFMFB*2048); |
my $XRFMFP = $pointer - ($XRFMFB*2048); |
313 |
|
|
|
|
|
314 |
# (XRFMFB - 1) * 512 + XRFMFP |
# (XRFMFB - 1) * 512 + XRFMFP |
315 |
# why do i have to do XRFMFP % 1024 ? |
# why do i have to do XRFMFP % 1024 ? |
316 |
|
|
317 |
my $blk_off = (($XRFMFB - 1) * 512) + ($XRFMFP % 1024); |
my $blk_off = (($XRFMFB - 1) * 512) + ($XRFMFP % 512); |
318 |
|
|
319 |
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'}); |
320 |
|
|
321 |
# Get Record Information |
# Get Record Information |
322 |
|
|
323 |
seek($self->{'fileMST'},$blk_off,0); |
seek($self->{'fileMST'},$blk_off,0) || croak "can't seek to $blk_off: $!"; |
324 |
|
|
325 |
read($self->{'fileMST'}, $buff, 4); |
read($self->{'fileMST'}, $buff, 4) || croak "can't read 4 bytes at offset $blk_off from MST file: $!"; |
326 |
my $value=unpack("l",$buff); |
my $value=unpack("V",$buff); |
327 |
|
|
328 |
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}); |
329 |
|
|
330 |
if ($value!=$mfn) { |
if ($value!=$mfn) { |
331 |
carp "Error: MFN ".$mfn." not found in MST(".$value.")"; |
if ($value == 0) { |
332 |
#return; # XXX deleted record? |
print STDERR "## record $mfn is physically deleted\n" if ($self->{debug}); |
333 |
} |
$self->{deleted} = $mfn; |
334 |
|
return; |
335 |
|
} |
336 |
|
|
337 |
# $MFRL=$self->Read16($fileMST); |
carp "Error: MFN ".$mfn." not found in MST file, found $value"; |
338 |
# $MFBWB=$self->Read32($fileMST); |
return; |
339 |
# $MFBWP=$self->Read16($fileMST); |
} |
|
# $BASE=$self->Read16($fileMST); |
|
|
# $NVF=$self->Read16($fileMST); |
|
|
# $STATUS=$self->Read16($fileMST); |
|
340 |
|
|
341 |
read($self->{'fileMST'}, $buff, 14); |
read($self->{'fileMST'}, $buff, 14); |
342 |
|
|
343 |
my ($MFRL,$MFBWB,$MFBWP,$BASE,$NVF,$STATUS) = unpack("slssss", $buff); |
my ($MFRL,$MFBWB,$MFBWP,$BASE,$NVF,$STATUS) = unpack("vVvvvv", $buff); |
344 |
|
|
345 |
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}); |
346 |
|
|
347 |
# 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; |
|
|
} |
|
348 |
|
|
349 |
warn "BASE is not 18+6*NVF" unless ($BASE == 18 + 6 * $NVF); |
warn "BASE is not 18+6*NVF" unless ($BASE == 18 + 6 * $NVF); |
350 |
|
|
360 |
|
|
361 |
for (my $i = 0 ; $i < $NVF ; $i++) { |
for (my $i = 0 ; $i < $NVF ; $i++) { |
362 |
|
|
363 |
# $TAG=$self->Read16($fileMST); |
my ($TAG,$POS,$LEN) = unpack("vvv", substr($buff,$i * 6, 6)); |
|
# $POS=$self->Read16($fileMST); |
|
|
# $LEN=$self->Read16($fileMST); |
|
|
|
|
|
my ($TAG,$POS,$LEN) = unpack("sss", substr($buff,$i * 6, 6)); |
|
364 |
|
|
365 |
print STDERR "## TAG: $TAG POS: $POS LEN: $LEN\n" if ($self->{debug}); |
print STDERR "## TAG: $TAG POS: $POS LEN: $LEN\n" if ($self->{debug}); |
366 |
|
|
396 |
|
|
397 |
$self->{'current_mfn'} = $mfn; |
$self->{'current_mfn'} = $mfn; |
398 |
|
|
399 |
print Dumper($self),"\n" if ($self->{debug}); |
print STDERR Dumper($self),"\n" if ($self->{debug}); |
400 |
|
|
401 |
return $self->{'record'}; |
return $self->{'record'}; |
402 |
} |
} |
403 |
|
|
404 |
|
=head2 mfn |
405 |
|
|
406 |
|
Returns current MFN position |
407 |
|
|
408 |
|
my $mfn = $isis->mfn; |
409 |
|
|
410 |
|
=cut |
411 |
|
|
412 |
|
# This function should be simple return $self->{current_mfn}, |
413 |
|
# but if new is called with _hack_mfn it becomes setter. |
414 |
|
# It's useful in tests when setting $isis->{record} directly |
415 |
|
|
416 |
|
sub mfn { |
417 |
|
my $self = shift; |
418 |
|
return $self->{current_mfn}; |
419 |
|
}; |
420 |
|
|
421 |
|
|
422 |
=head2 to_ascii |
=head2 to_ascii |
423 |
|
|
424 |
Dump ASCII output of record with specified MFN |
Returns ASCII output of record with specified MFN |
425 |
|
|
426 |
print $isis->to_ascii(42); |
print $isis->to_ascii(42); |
427 |
|
|
428 |
It outputs something like this: |
This outputs something like this: |
429 |
|
|
430 |
210 ^aNew York^cNew York University press^dcop. 1988 |
210 ^aNew York^cNew York University press^dcop. 1988 |
431 |
990 2140 |
990 2140 |
442 |
|
|
443 |
my $mfn = shift || croak "need MFN"; |
my $mfn = shift || croak "need MFN"; |
444 |
|
|
445 |
my $rec = $self->fetch($mfn); |
my $rec = $self->fetch($mfn) || return; |
446 |
|
|
447 |
my $out = "0\t$mfn"; |
my $out = "0\t$mfn"; |
448 |
|
|
462 |
|
|
463 |
my $hash = $isis->to_hash($mfn); |
my $hash = $isis->to_hash($mfn); |
464 |
|
|
465 |
It has ability to convert characters (using C<hash_filter> from ISIS |
It has ability to convert characters (using C<hash_filter>) from ISIS |
466 |
database before creating structures enabling character re-mapping or quick |
database before creating structures enabling character re-mapping or quick |
467 |
fix-up of data. |
fix-up of data. |
468 |
|
|
498 |
} |
} |
499 |
], |
], |
500 |
|
|
501 |
|
In case there are repeatable subfields in record, this will create |
502 |
|
following structure: |
503 |
|
|
504 |
|
'900' => [ { |
505 |
|
'a' => [ 'foo', 'bar', 'baz' ], |
506 |
|
}] |
507 |
|
|
508 |
|
Or in more complex example of |
509 |
|
|
510 |
|
902 ^aa1^aa2^aa3^bb1^aa4^bb2^cc1^aa5 |
511 |
|
|
512 |
|
it will create |
513 |
|
|
514 |
|
902 => [ |
515 |
|
{ a => ["a1", "a2", "a3", "a4", "a5"], b => ["b1", "b2"], c => "c1" }, |
516 |
|
], |
517 |
|
|
518 |
|
This behaviour can be changed using C<join_subfields_with> option to L</new>, |
519 |
|
in which case C<to_hash> will always create single value for each subfield. |
520 |
|
This will change result to: |
521 |
|
|
522 |
|
|
523 |
|
|
524 |
This method will also create additional field C<000> with MFN. |
This method will also create additional field C<000> with MFN. |
525 |
|
|
526 |
|
There is also more elaborative way to call C<to_hash> like this: |
527 |
|
|
528 |
|
my $hash = $isis->to_hash({ |
529 |
|
mfn => 42, |
530 |
|
include_subfields => 1, |
531 |
|
regexps => [ |
532 |
|
's/something/else/g', |
533 |
|
], |
534 |
|
}); |
535 |
|
|
536 |
|
Each option controll creation of hash: |
537 |
|
|
538 |
|
=over 4 |
539 |
|
|
540 |
|
=item mfn |
541 |
|
|
542 |
|
Specify MFN number of record |
543 |
|
|
544 |
|
=item include_subfields |
545 |
|
|
546 |
|
This option will create additional key in hash called C<subfields> which will |
547 |
|
have original record subfield order and index to that subfield like this: |
548 |
|
|
549 |
|
902 => [ { |
550 |
|
a => ["a1", "a2", "a3", "a4", "a5"], |
551 |
|
b => ["b1", "b2"], |
552 |
|
c => "c1", |
553 |
|
subfields => ["a", 0, "a", 1, "a", 2, "b", 0, "a", 3, "b", 1, "c", 0, "a", 4], |
554 |
|
} ], |
555 |
|
|
556 |
|
=item join_subfields_with |
557 |
|
|
558 |
|
Define delimiter which will be used to join repeatable subfields. You can |
559 |
|
specify option here instead in L</new> if you want to have per-record control. |
560 |
|
|
561 |
|
=item regexpes |
562 |
|
|
563 |
|
Override C<regexpes> specified in L</new>. |
564 |
|
|
565 |
|
=back |
566 |
|
|
567 |
=cut |
=cut |
568 |
|
|
569 |
sub to_hash { |
sub to_hash { |
570 |
my $self = shift; |
my $self = shift; |
571 |
|
|
572 |
|
|
573 |
my $mfn = shift || confess "need mfn!"; |
my $mfn = shift || confess "need mfn!"; |
574 |
|
my $arg; |
575 |
|
|
576 |
|
if (ref($mfn) eq 'HASH') { |
577 |
|
$arg = $mfn; |
578 |
|
$mfn = $arg->{mfn} || confess "need mfn in arguments"; |
579 |
|
} |
580 |
|
|
581 |
|
$arg->{regexpes} ||= $self->{regexpes}; |
582 |
|
|
583 |
|
confess "regexps must be HASH" if ($arg->{regexps} && ref($arg->{regexps}) ne 'HASH'); |
584 |
|
|
585 |
# init record to include MFN as field 000 |
# init record to include MFN as field 000 |
586 |
my $rec = { '000' => [ $mfn ] }; |
my $rec = { '000' => [ $mfn ] }; |
587 |
|
|
588 |
my $row = $self->fetch($mfn); |
my $row = $self->fetch($mfn) || return; |
589 |
|
|
590 |
foreach my $k (keys %{$row}) { |
my $j_rs = $arg->{join_subfields_with}; |
591 |
foreach my $l (@{$row->{$k}}) { |
$j_rs = $self->{join_subfields_with} unless(defined($j_rs)); |
592 |
|
my $i_sf = $arg->{include_subfields}; |
593 |
|
|
594 |
|
foreach my $f_nr (keys %{$row}) { |
595 |
|
foreach my $l (@{$row->{$f_nr}}) { |
596 |
|
|
597 |
# filter output |
# filter output |
598 |
$l = $self->{'hash_filter'}->($l) if ($self->{'hash_filter'}); |
if ($self->{'hash_filter'}) { |
599 |
|
$l = $self->{'hash_filter'}->($l); |
600 |
|
next unless defined($l); |
601 |
|
} |
602 |
|
|
603 |
|
# apply regexps |
604 |
|
if ($arg->{regexps} && defined($arg->{regexps}->{$f_nr})) { |
605 |
|
confess "regexps->{$f_nr} must be ARRAY" if (ref($arg->{regexps}->{$f_nr}) ne 'ARRAY'); |
606 |
|
my $c = 0; |
607 |
|
foreach my $r (@{ $arg->{regexps}->{$f_nr} }) { |
608 |
|
while ( eval '$l =~ ' . $r ) { $c++ }; |
609 |
|
} |
610 |
|
warn "## field $f_nr triggered $c regexpes\n" if ($c && $self->{debug}); |
611 |
|
} |
612 |
|
|
613 |
my $val; |
my $val; |
614 |
|
my $r_sf; # repeatable subfields in this record |
615 |
|
|
616 |
# has identifiers? |
# has identifiers? |
617 |
($val->{'i1'},$val->{'i2'}) = ($1,$2) if ($l =~ s/^([01 #])([01 #])\^/\^/); |
($val->{'i1'},$val->{'i2'}) = ($1,$2) if ($l =~ s/^([01 #])([01 #])\^/\^/); |
620 |
if ($l =~ m/\^/) { |
if ($l =~ m/\^/) { |
621 |
foreach my $t (split(/\^/,$l)) { |
foreach my $t (split(/\^/,$l)) { |
622 |
next if (! $t); |
next if (! $t); |
623 |
$val->{substr($t,0,1)} = substr($t,1); |
my ($sf,$v) = (substr($t,0,1), substr($t,1)); |
624 |
|
# XXX this might be option, but why? |
625 |
|
next unless ($v); |
626 |
|
# warn "### $f_nr^$sf:$v",$/ if ($self->{debug} > 1); |
627 |
|
|
628 |
|
if (ref( $val->{$sf} ) eq 'ARRAY') { |
629 |
|
|
630 |
|
push @{ $val->{$sf} }, $v; |
631 |
|
|
632 |
|
# record repeatable subfield it it's offset |
633 |
|
push @{ $val->{subfields} }, ( $sf, $#{ $val->{$sf} } ) if (! $j_rs && $i_sf); |
634 |
|
$r_sf->{$sf}++; |
635 |
|
|
636 |
|
} elsif (defined( $val->{$sf} )) { |
637 |
|
|
638 |
|
# convert scalar field to array |
639 |
|
$val->{$sf} = [ $val->{$sf}, $v ]; |
640 |
|
|
641 |
|
push @{ $val->{subfields} }, ( $sf, 1 ) if (! $j_rs && $i_sf); |
642 |
|
$r_sf->{$sf}++; |
643 |
|
|
644 |
|
} else { |
645 |
|
$val->{$sf} = $v; |
646 |
|
push @{ $val->{subfields} }, ( $sf, 0 ) if ($i_sf); |
647 |
|
} |
648 |
} |
} |
649 |
} else { |
} else { |
650 |
$val = $l; |
$val = $l; |
651 |
} |
} |
652 |
|
|
653 |
push @{$rec->{$k}}, $val; |
if ($j_rs) { |
654 |
|
map { |
655 |
|
$val->{$_} = join($j_rs, @{ $val->{$_} }); |
656 |
|
} keys %$r_sf |
657 |
|
} |
658 |
|
|
659 |
|
push @{$rec->{$f_nr}}, $val; |
660 |
} |
} |
661 |
} |
} |
662 |
|
|
677 |
return $self->{'TagName'}->{$tag} || $tag; |
return $self->{'TagName'}->{$tag} || $tag; |
678 |
} |
} |
679 |
|
|
680 |
|
|
681 |
|
=head2 read_cnt |
682 |
|
|
683 |
|
Read content of C<.CNT> file and return hash containing it. |
684 |
|
|
685 |
|
print Dumper($isis->read_cnt); |
686 |
|
|
687 |
|
This function is not used by module (C<.CNT> files are not required for this |
688 |
|
module to work), but it can be useful to examine your index (while debugging |
689 |
|
for example). |
690 |
|
|
691 |
|
=cut |
692 |
|
|
693 |
|
sub read_cnt { |
694 |
|
my $self = shift; |
695 |
|
|
696 |
|
croak "missing CNT file in ",$self->{isisdb} unless ($self->{cnt_file}); |
697 |
|
|
698 |
|
# Get the index information from $db.CNT |
699 |
|
|
700 |
|
open(my $fileCNT, $self->{cnt_file}) || croak "can't read '$self->{cnt_file}': $!"; |
701 |
|
binmode($fileCNT); |
702 |
|
|
703 |
|
my $buff; |
704 |
|
|
705 |
|
read($fileCNT, $buff, 26) || croak "can't read first table from CNT: $!"; |
706 |
|
$self->unpack_cnt($buff); |
707 |
|
|
708 |
|
read($fileCNT, $buff, 26) || croak "can't read second table from CNT: $!"; |
709 |
|
$self->unpack_cnt($buff); |
710 |
|
|
711 |
|
close($fileCNT); |
712 |
|
|
713 |
|
return $self->{cnt}; |
714 |
|
} |
715 |
|
|
716 |
|
=head2 unpack_cnt |
717 |
|
|
718 |
|
Unpack one of two 26 bytes fixed length record in C<.CNT> file. |
719 |
|
|
720 |
|
Here is definition of record: |
721 |
|
|
722 |
|
off key description size |
723 |
|
0: IDTYPE BTree type s |
724 |
|
2: ORDN Nodes Order s |
725 |
|
4: ORDF Leafs Order s |
726 |
|
6: N Number of Memory buffers for nodes s |
727 |
|
8: K Number of buffers for first level index s |
728 |
|
10: LIV Current number of Index Levels s |
729 |
|
12: POSRX Pointer to Root Record in N0x l |
730 |
|
16: NMAXPOS Next Available position in N0x l |
731 |
|
20: FMAXPOS Next available position in L0x l |
732 |
|
24: ABNORMAL Formal BTree normality indicator s |
733 |
|
length: 26 bytes |
734 |
|
|
735 |
|
This will fill C<$self> object under C<cnt> with hash. It's used by C<read_cnt>. |
736 |
|
|
737 |
|
=cut |
738 |
|
|
739 |
|
sub unpack_cnt { |
740 |
|
my $self = shift; |
741 |
|
|
742 |
|
my @flds = qw(ORDN ORDF N K LIV POSRX NMAXPOS FMAXPOS ABNORMAL); |
743 |
|
|
744 |
|
my $buff = shift || return; |
745 |
|
my @arr = unpack("vvvvvvVVVv", $buff); |
746 |
|
|
747 |
|
print STDERR "unpack_cnt: ",join(" ",@arr),"\n" if ($self->{'debug'}); |
748 |
|
|
749 |
|
my $IDTYPE = shift @arr; |
750 |
|
foreach (@flds) { |
751 |
|
$self->{cnt}->{$IDTYPE}->{$_} = abs(shift @arr); |
752 |
|
} |
753 |
|
} |
754 |
|
|
755 |
1; |
1; |
756 |
|
|
757 |
=head1 BUGS |
=head1 BUGS |
758 |
|
|
759 |
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 |
760 |
|
some variations in input databases which has been tested with this module. |
761 |
|
When I was in doubt, I assumed that OpenIsis's implementation was right |
762 |
|
(except for obvious bugs). |
763 |
|
|
764 |
|
However, every effort has been made to test this module with as much |
765 |
|
databases (and programs that create them) as possible. |
766 |
|
|
767 |
|
I would be very greatful for success or failure reports about usage of this |
768 |
|
module with databases from programs other than WinIsis and IsisMarc. I had |
769 |
|
tested this against ouput of one C<isis.dll>-based application, but I don't |
770 |
|
know any details about it's version. |
771 |
|
|
772 |
|
=head1 VERSIONS |
773 |
|
|
774 |
|
As this is young module, new features are added in subsequent version. It's |
775 |
|
a good idea to specify version when using this module like this: |
776 |
|
|
777 |
|
use Biblio::Isis 0.21 |
778 |
|
|
779 |
|
Below is list of changes in specific version of module (so you can target |
780 |
|
older versions if you really have to): |
781 |
|
|
782 |
|
=over 8 |
783 |
|
|
784 |
|
=item 0.21 |
785 |
|
|
786 |
|
Added C<join_subfields_with> to L</new> and L</to_hash>. |
787 |
|
|
788 |
|
Added C<include_subfields> to L</to_hash>. |
789 |
|
|
790 |
|
=item 0.20 |
791 |
|
|
792 |
|
Added C<< $isis->mfn >>, support for repeatable subfields and |
793 |
|
C<< $isis->to_hash({ mfn => 42, ... }) >> calling convention |
794 |
|
|
795 |
|
=back |
796 |
|
|
797 |
=head1 AUTHOR |
=head1 AUTHOR |
798 |
|
|
815 |
|
|
816 |
=head1 SEE ALSO |
=head1 SEE ALSO |
817 |
|
|
818 |
|
L<Biblio::Isis::Manual> for CDS/ISIS manual appendix F, G and H which describe file format |
819 |
|
|
820 |
OpenIsis web site L<http://www.openisis.org> |
OpenIsis web site L<http://www.openisis.org> |
821 |
|
|
822 |
perl4lib site L<http://perl4lib.perl.org> |
perl4lib site L<http://perl4lib.perl.org> |