1 |
package IsisDB; |
package Biblio::Isis; |
2 |
use strict; |
use strict; |
3 |
|
|
4 |
use Carp; |
use Carp; |
5 |
use Data::Dumper; |
use File::Glob qw(:globally :nocase); |
6 |
|
|
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 = 0.02; |
$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 database |
Biblio::Isis - Read CDS/ISIS, WinISIS and IsisMarc database |
22 |
|
|
23 |
=head1 SYNOPSIS |
=head1 SYNOPSIS |
24 |
|
|
25 |
use IsisDB |
use Biblio::Isis; |
26 |
my $isis = new IsisDB( |
|
27 |
|
my $isis = new Biblio::Isis( |
28 |
isisdb => './cds/cds', |
isisdb => './cds/cds', |
29 |
); |
); |
30 |
|
|
31 |
|
for(my $mfn = 1; $mfn <= $isis->count; $mfn++) { |
32 |
|
print $isis->to_ascii($mfn),"\n"; |
33 |
|
} |
34 |
|
|
35 |
=head1 DESCRIPTION |
=head1 DESCRIPTION |
36 |
|
|
37 |
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 |
38 |
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>), |
42 |
|
ASCII dump (using C<to_ascii>) or just hash with field names and packed |
43 |
|
values (like C<^asomething^belse>). |
44 |
|
|
45 |
|
Unique feature of this module is ability to C<include_deleted> records. |
46 |
|
It will also skip zero sized fields (OpenIsis has a bug in XS bindings, so |
47 |
|
fields which are zero sized will be filled with random junk from memory). |
48 |
|
|
49 |
|
It also has support for identifiers (only if ISIS database is created by |
50 |
|
IsisMarc), see C<to_hash>. |
51 |
|
|
52 |
|
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 |
54 |
|
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 |
56 |
|
OpenIsis, so reduced time in other parts of the code should compensate for |
57 |
|
slower performance of this module (speed of reading ISIS database is |
58 |
|
rarely an issue). |
59 |
|
|
60 |
=head1 METHODS |
=head1 METHODS |
61 |
|
|
75 |
# some binary reads |
# some binary reads |
76 |
# |
# |
77 |
|
|
|
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); |
|
|
} |
|
|
|
|
78 |
=head2 new |
=head2 new |
79 |
|
|
80 |
Open CDS/ISIS database |
Open ISIS database |
81 |
|
|
82 |
my $isis = new IsisDB( |
my $isis = new Biblio::Isis( |
83 |
isisdb => './cds/cds', |
isisdb => './cds/cds', |
84 |
read_fdt => 1, |
read_fdt => 1, |
85 |
|
include_deleted => 1, |
86 |
|
hash_filter => sub { |
87 |
|
my ($v,$field_number) = @_; |
88 |
|
$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: |
97 |
|
|
98 |
=item isisdb |
=item isisdb |
99 |
|
|
100 |
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 |
101 |
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 |
102 |
|
C<read_fdt> option) files. |
103 |
|
|
104 |
|
In this example it uses C<./cds/cds.MST> and related files. |
105 |
|
|
106 |
=item read_fdt |
=item read_fdt |
107 |
|
|
108 |
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 |
109 |
by default. |
by default. |
110 |
|
|
111 |
|
=item include_deleted |
112 |
|
|
113 |
|
Don't skip logically deleted records in ISIS. |
114 |
|
|
115 |
|
=item hash_filter |
116 |
|
|
117 |
|
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<< $_[1] >>). |
120 |
|
|
121 |
=item debug |
=item debug |
122 |
|
|
123 |
Dump a C<lot> of debugging output. |
Dump a B<lot> of debugging output even at level 1. For even more increase level. |
124 |
|
|
125 |
=back |
=item join_subfields_with |
126 |
|
|
127 |
|
Define delimiter which will be used to join repeatable subfields. This |
128 |
|
option is included to support lagacy application written against version |
129 |
|
older than 0.21 of this module. By default, it disabled. See L</to_hash>. |
130 |
|
|
131 |
It will also set C<$isis-E<gt>{'maxmfn'}> which is maximum MFN stored in database. |
=back |
132 |
|
|
133 |
=cut |
=cut |
134 |
|
|
137 |
my $self = {}; |
my $self = {}; |
138 |
bless($self, $class); |
bless($self, $class); |
139 |
|
|
140 |
$self->{isisdb} = {@_}->{isisdb} || croak "new needs database name as argument!"; |
croak "new needs database name (isisdb) as argument!" unless ({@_}->{isisdb}); |
141 |
|
|
142 |
$self->{debug} = {@_}->{debug}; |
foreach my $v (qw{isisdb debug include_deleted hash_filter}) { |
143 |
|
$self->{$v} = {@_}->{$v}; |
144 |
|
} |
145 |
|
|
146 |
|
my @isis_files = grep(/\.(FDT|MST|XRF|CNT)$/i,glob($self->{isisdb}."*")); |
147 |
|
|
148 |
|
foreach my $f (@isis_files) { |
149 |
|
my $ext = $1 if ($f =~ m/\.(\w\w\w)$/); |
150 |
|
$self->{lc($ext)."_file"} = $f; |
151 |
|
} |
152 |
|
|
153 |
|
my @must_exist = qw(mst xrf); |
154 |
|
push @must_exist, "fdt" if ($self->{read_fdt}); |
155 |
|
|
156 |
|
foreach my $ext (@must_exist) { |
157 |
|
unless ($self->{$ext."_file"}) { |
158 |
|
carp "missing ",uc($ext)," file in ",$self->{isisdb}; |
159 |
|
return; |
160 |
|
} |
161 |
|
} |
162 |
|
|
163 |
|
if ($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 ({@_}->{read_fdt} && -e $self->{isisdb}.".FDT") { |
if ($self->{read_fdt} && -e $self->{fdt_file}) { |
176 |
|
|
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, $self->{isisdb}.".FDT") || croak "can't read '$self->{isisdb}.FDT': $!"; |
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 $db.MST |
204 |
|
|
205 |
open(fileMST,$self->{isisdb}.".MST") || croak "can't read '$self->{isisdb}.MST': $!"; |
open($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(fileMST,4,0); |
seek($self->{'fileMST'},4,0) || croak "can't seek to offset 0 in MST: $!"; |
|
$self->{'NXTMFN'}=$self->Read32(\*fileMST) || carp "NXTNFN is zero"; |
|
|
|
|
|
# save maximum MFN |
|
|
$self->{'maxmfn'} = $self->{'NXTMFN'} - 1; |
|
|
|
|
|
close(fileMST); |
|
215 |
|
|
216 |
# Get the index information from $db.CNT |
my $buff; |
|
|
|
|
open(fileCNT, $self->{isisdb}.".CNT") || croak "can't read '$self->{isisdb}.CNT': $!"; |
|
217 |
|
|
218 |
# There is two 26 Bytes fixed lenght records |
read($self->{'fileMST'}, $buff, 4) || croak "can't read NXTMFN from MST: $!"; |
219 |
|
$self->{'NXTMFN'}=unpack("V",$buff) || croak "NXTNFN is zero"; |
220 |
|
|
221 |
# 0: IDTYPE BTree type 16 |
print STDERR "## self ",Dumper($self),"\n" if ($self->{debug}); |
|
# 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); |
|
|
|
|
|
my $IDTYPE = shift @arr; |
|
|
foreach (@flds) { |
|
|
$self->{$IDTYPE}->{$_} = abs(shift @arr); |
|
|
} |
|
|
} |
|
222 |
|
|
223 |
my $buff; |
# open files for later |
224 |
read(fileCNT, $buff, 26); |
open($self->{'fileXRF'}, $self->{xrf_file}) || croak "can't open '$self->{xrf_file}': $!"; |
225 |
$self->unpack_cnt($buff); |
binmode($self->{'fileXRF'}); |
|
|
|
|
read(fileCNT, $buff, 26); |
|
|
$self->unpack_cnt($buff); |
|
226 |
|
|
227 |
|
$self ? return $self : return undef; |
228 |
|
} |
229 |
|
|
230 |
close(fileCNT); |
=head2 count |
231 |
|
|
232 |
print Dumper($self) if ($self->{debug}); |
Return number of records in database |
233 |
|
|
234 |
# open files for later |
print $isis->count; |
|
open($self->{'fileXRF'}, $self->{isisdb}.".XRF") || croak "can't open '$self->{isisdb}.XRF': $!"; |
|
235 |
|
|
236 |
open($self->{'fileMST'}, $self->{isisdb}.".MST") || croak "can't open '$self->{isisdb}.MST': $!"; |
=cut |
237 |
|
|
238 |
$self ? return $self : return undef; |
sub count { |
239 |
|
my $self = shift; |
240 |
|
return $self->{'NXTMFN'} - 1; |
241 |
} |
} |
242 |
|
|
243 |
=head2 fetch |
=head2 fetch |
247 |
my $rec = $isis->fetch(55); |
my $rec = $isis->fetch(55); |
248 |
|
|
249 |
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 |
250 |
for that field. |
for that field like this: |
251 |
|
|
252 |
|
$rec = { |
253 |
|
'210' => [ '^aNew York^cNew York University press^dcop. 1988' ], |
254 |
|
'990' => [ '2140', '88', 'HAY' ], |
255 |
|
}; |
256 |
|
|
257 |
=cut |
=cut |
258 |
|
|
261 |
|
|
262 |
my $mfn = shift || croak "fetch needs MFN as argument!"; |
my $mfn = shift || croak "fetch needs MFN as argument!"; |
263 |
|
|
264 |
print "fetch: $mfn\n" if ($self->{debug}); |
# is mfn allready in memory? |
265 |
|
my $old_mfn = $self->{'current_mfn'} || -1; |
266 |
|
return $self->{record} if ($mfn == $old_mfn); |
267 |
|
|
268 |
|
print STDERR "## fetch: $mfn\n" if ($self->{debug}); |
269 |
|
|
270 |
# XXX check this? |
# XXX check this? |
271 |
my $mfnpos=($mfn+int(($mfn-1)/127))*4; |
my $mfnpos=($mfn+int(($mfn-1)/127))*4; |
272 |
|
|
273 |
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 ($self->{debug}); |
274 |
seek($self->{'fileXRF'},$mfnpos,0); |
seek($self->{'fileXRF'},$mfnpos,0); |
275 |
|
|
276 |
|
my $buff; |
277 |
|
|
278 |
|
# delete old record |
279 |
|
delete $self->{record}; |
280 |
|
|
281 |
# read XRFMFB abd XRFMFP |
# read XRFMFB abd XRFMFP |
282 |
my $pointer=$self->Read32(\*{$self->{'fileXRF'}}); |
read($self->{'fileXRF'}, $buff, 4); |
283 |
|
my $pointer=unpack("V",$buff); |
284 |
|
if (! $pointer) { |
285 |
|
if ($self->{include_deleted}) { |
286 |
|
return; |
287 |
|
} else { |
288 |
|
warn "pointer for MFN $mfn is null\n"; |
289 |
|
return; |
290 |
|
} |
291 |
|
} |
292 |
|
|
293 |
|
# check for logically deleted record |
294 |
|
if ($pointer & 0x80000000) { |
295 |
|
print STDERR "## record $mfn is logically deleted\n" if ($self->{debug}); |
296 |
|
$self->{deleted} = $mfn; |
297 |
|
|
298 |
|
return unless $self->{include_deleted}; |
299 |
|
|
300 |
|
# abs |
301 |
|
$pointer = ($pointer ^ 0xffffffff) + 1; |
302 |
|
} |
303 |
|
|
304 |
my $XRFMFB = int($pointer/2048); |
my $XRFMFB = int($pointer/2048); |
305 |
my $XRFMFP = $pointer - ($XRFMFB*2048); |
my $XRFMFP = $pointer - ($XRFMFB*2048); |
306 |
|
|
307 |
print "XRFMFB: $XRFMFB XRFMFP: $XRFMFP\n" if ($self->{debug}); |
# (XRFMFB - 1) * 512 + XRFMFP |
308 |
|
# why do i have to do XRFMFP % 1024 ? |
309 |
|
|
310 |
# XXX fix this to be more readable!! |
my $blk_off = (($XRFMFB - 1) * 512) + ($XRFMFP % 512); |
|
# e.g. (XRFMFB - 1) * 512 + XRFMFP |
|
|
|
|
|
my $offset = $pointer; |
|
|
my $offset2=int($offset/2048)-1; |
|
|
my $offset22=int($offset/4096); |
|
|
my $offset3=$offset-($offset22*4096); |
|
|
if ($offset3>512) { |
|
|
$offset3=$offset3-2048; |
|
|
} |
|
|
my $offset4=($offset2*512)+$offset3; |
|
311 |
|
|
312 |
print "$offset - $offset2 - $offset3 - $offset4\n" if ($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($self->{'fileMST'},$offset4,0); |
seek($self->{'fileMST'},$blk_off,0) || croak "can't seek to $blk_off: $!"; |
317 |
|
|
318 |
my $value=$self->Read32(\*{$self->{'fileMST'}}); |
read($self->{'fileMST'}, $buff, 4) || croak "can't read 4 bytes at offset $blk_off from MST file: $!"; |
319 |
|
my $value=unpack("V",$buff); |
320 |
|
|
321 |
|
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 |
print ("Error: The MFN:".$mfn." is not found in MST(".$value.")"); |
if ($value == 0) { |
325 |
return -1; # XXX deleted record? |
print STDERR "## record $mfn is physically deleted\n" if ($self->{debug}); |
326 |
} |
$self->{deleted} = $mfn; |
327 |
|
return; |
328 |
|
} |
329 |
|
|
330 |
# $MFRL=$self->Read16($fileMST); |
carp "Error: MFN ".$mfn." not found in MST file, found $value"; |
331 |
# $MFBWB=$self->Read32($fileMST); |
return; |
332 |
# $MFBWP=$self->Read16($fileMST); |
} |
|
# $BASE=$self->Read16($fileMST); |
|
|
# $NVF=$self->Read16($fileMST); |
|
|
# $STATUS=$self->Read16($fileMST); |
|
333 |
|
|
|
my $buff; |
|
334 |
read($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}); |
339 |
|
|
340 |
|
warn "MFRL $MFRL is not even number" unless ($MFRL % 2 == 0); |
341 |
|
|
342 |
print "MFRL: $MFRL MFBWB: $MFBWB MFBWP: $MFBWP BASE: $BASE NVF: $NVF STATUS: $STATUS\n" if ($self->{debug}); |
warn "BASE is not 18+6*NVF" unless ($BASE == 18 + 6 * $NVF); |
343 |
|
|
344 |
# Get Directory Format |
# Get Directory Format |
345 |
|
|
347 |
my @FieldLEN; |
my @FieldLEN; |
348 |
my @FieldTAG; |
my @FieldTAG; |
349 |
|
|
350 |
for (my $i = 0 ; $i < $NVF ; $i++) { |
read($self->{'fileMST'}, $buff, 6 * $NVF); |
351 |
|
|
352 |
|
my $rec_len = 0; |
353 |
|
|
354 |
# $TAG=$self->Read16($fileMST); |
for (my $i = 0 ; $i < $NVF ; $i++) { |
|
# $POS=$self->Read16($fileMST); |
|
|
# $LEN=$self->Read16($fileMST); |
|
355 |
|
|
356 |
read($self->{'fileMST'}, $buff, 6); |
my ($TAG,$POS,$LEN) = unpack("vvv", substr($buff,$i * 6, 6)); |
|
my ($TAG,$POS,$LEN) = unpack("sss", $buff); |
|
357 |
|
|
358 |
print "TAG: $TAG POS: $POS LEN: $LEN\n" if ($self->{debug}); |
print STDERR "## TAG: $TAG POS: $POS LEN: $LEN\n" if ($self->{debug}); |
359 |
|
|
360 |
# 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. |
361 |
# |
# |
370 |
push @FieldTAG,$TAG; |
push @FieldTAG,$TAG; |
371 |
push @FieldPOS,$POS; |
push @FieldPOS,$POS; |
372 |
push @FieldLEN,$LEN; |
push @FieldLEN,$LEN; |
373 |
|
|
374 |
|
$rec_len += $LEN; |
375 |
} |
} |
376 |
|
|
377 |
# Get Variable Fields |
# Get Variable Fields |
378 |
|
|
379 |
delete $self->{record}; |
read($self->{'fileMST'},$buff,$rec_len); |
380 |
|
|
381 |
|
print STDERR "## rec_len: $rec_len poc: ",tell($self->{'fileMST'})."\n" if ($self->{debug}); |
382 |
|
|
383 |
for (my $i = 0 ; $i < $NVF ; $i++) { |
for (my $i = 0 ; $i < $NVF ; $i++) { |
384 |
my $rec; |
# skip zero-sized fields |
385 |
read($self->{'fileMST'},$rec,$FieldLEN[$i]); |
next if ($FieldLEN[$i] == 0); |
|
push @{$self->{record}->{$FieldTAG[$i]}}, $rec; |
|
|
} |
|
|
close(fileMST); |
|
386 |
|
|
387 |
# The record is marked for deletion |
push @{$self->{record}->{$FieldTAG[$i]}}, substr($buff,$FieldPOS[$i],$FieldLEN[$i]); |
|
if ($STATUS==1) { |
|
|
return -1; |
|
388 |
} |
} |
389 |
|
|
390 |
print Dumper($self) if ($self->{debug}); |
$self->{'current_mfn'} = $mfn; |
391 |
|
|
392 |
|
print STDERR Dumper($self),"\n" if ($self->{debug}); |
393 |
|
|
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 selected MFN |
Returns ASCII output of record with specified MFN |
418 |
|
|
419 |
|
print $isis->to_ascii(42); |
420 |
|
|
421 |
|
This outputs something like this: |
422 |
|
|
423 |
|
210 ^aNew York^cNew York University press^dcop. 1988 |
424 |
|
990 2140 |
425 |
|
990 88 |
426 |
|
990 HAY |
427 |
|
|
428 |
print $isis->to_ascii(55); |
If C<read_fdt> is specified when calling C<new> it will display field names |
429 |
|
from C<.FDT> file instead of numeric tags. |
430 |
|
|
431 |
=cut |
=cut |
432 |
|
|
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 |
|
|
442 |
foreach my $f (sort keys %{$rec}) { |
foreach my $f (sort keys %{$rec}) { |
443 |
$out .= "\n$f\t".join("\n$f\t",@{$self->{record}->{$f}}); |
my $fn = $self->tag_name($f); |
444 |
|
$out .= "\n$fn\t".join("\n$fn\t",@{$self->{record}->{$f}}); |
445 |
} |
} |
446 |
|
|
447 |
$out .= "\n"; |
$out .= "\n"; |
449 |
return $out; |
return $out; |
450 |
} |
} |
451 |
|
|
452 |
# |
=head2 to_hash |
453 |
# XXX porting from php left-over: |
|
454 |
# |
Read record with specified MFN and convert it to hash |
455 |
# do I *REALLY* need those methods, or should I use |
|
456 |
# $self->{something} directly? |
my $hash = $isis->to_hash($mfn); |
457 |
# |
|
458 |
# Probably direct usage is better! |
It has ability to convert characters (using C<hash_filter>) from ISIS |
459 |
# |
database before creating structures enabling character re-mapping or quick |
460 |
|
fix-up of data. |
461 |
|
|
462 |
|
This function returns hash which is like this: |
463 |
|
|
464 |
|
$hash = { |
465 |
|
'210' => [ |
466 |
|
{ |
467 |
|
'c' => 'New York University press', |
468 |
|
'a' => 'New York', |
469 |
|
'd' => 'cop. 1988' |
470 |
|
} |
471 |
|
], |
472 |
|
'990' => [ |
473 |
|
'2140', |
474 |
|
'88', |
475 |
|
'HAY' |
476 |
|
], |
477 |
|
}; |
478 |
|
|
479 |
|
You can later use that hash to produce any output from ISIS data. |
480 |
|
|
481 |
|
If database is created using IsisMarc, it will also have to special fields |
482 |
|
which will be used for identifiers, C<i1> and C<i2> like this: |
483 |
|
|
484 |
|
'200' => [ |
485 |
|
{ |
486 |
|
'i1' => '1', |
487 |
|
'i2' => ' ' |
488 |
|
'a' => 'Goa', |
489 |
|
'f' => 'Valdo D\'Arienzo', |
490 |
|
'e' => 'tipografie e tipografi nel XVI secolo', |
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. |
518 |
|
|
519 |
sub TagName { |
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 |
558 |
|
|
559 |
|
sub to_hash { |
560 |
|
my $self = shift; |
561 |
|
|
562 |
|
|
563 |
|
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 |
575 |
|
my $rec = { '000' => [ $mfn ] }; |
576 |
|
|
577 |
|
my $row = $self->fetch($mfn) || return; |
578 |
|
|
579 |
|
my $j_rs = $arg->{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 $f_nr (keys %{$row}) { |
584 |
|
foreach my $l (@{$row->{$f_nr}}) { |
585 |
|
|
586 |
|
# filter output |
587 |
|
$l = $hash_filter->($l, $f_nr) if ($hash_filter); |
588 |
|
next unless defined($l); |
589 |
|
|
590 |
|
my $val; |
591 |
|
my $r_sf; # repeatable subfields in this record |
592 |
|
|
593 |
|
# has identifiers? |
594 |
|
($val->{'i1'},$val->{'i2'}) = ($1,$2) if ($l =~ s/^([01 #])([01 #])\^/\^/); |
595 |
|
|
596 |
|
# has subfields? |
597 |
|
if ($l =~ m/\^/) { |
598 |
|
foreach my $t (split(/\^/,$l)) { |
599 |
|
next if (! $t); |
600 |
|
my ($sf,$v) = (substr($t,0,1), substr($t,1)); |
601 |
|
# XXX this might be option, but why? |
602 |
|
next unless ($v); |
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 { |
627 |
|
$val = $l; |
628 |
|
} |
629 |
|
|
630 |
|
if ($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 |
|
|
640 |
|
return $rec; |
641 |
|
} |
642 |
|
|
643 |
|
=head2 tag_name |
644 |
|
|
645 |
|
Return name of selected tag |
646 |
|
|
647 |
|
print $isis->tag_name('200'); |
648 |
|
|
649 |
|
=cut |
650 |
|
|
651 |
|
sub tag_name { |
652 |
|
my $self = shift; |
653 |
|
my $tag = shift || return; |
654 |
|
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; |
my $self = shift; |
672 |
return $self->{TagName}; |
|
673 |
|
croak "missing CNT file in ",$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 $self->{cnt}; |
691 |
} |
} |
692 |
|
|
693 |
sub NextMFN { |
=head2 unpack_cnt |
694 |
|
|
695 |
|
Unpack one of two 26 bytes fixed length record in C<.CNT> file. |
696 |
|
|
697 |
|
Here is definition of record: |
698 |
|
|
699 |
|
off key description size |
700 |
|
0: IDTYPE BTree type s |
701 |
|
2: ORDN Nodes Order s |
702 |
|
4: ORDF Leafs Order s |
703 |
|
6: N Number of Memory buffers for nodes s |
704 |
|
8: K Number of buffers for first level index s |
705 |
|
10: LIV Current number of Index Levels s |
706 |
|
12: POSRX Pointer to Root Record in N0x l |
707 |
|
16: NMAXPOS Next Available position in N0x l |
708 |
|
20: FMAXPOS Next available position in L0x l |
709 |
|
24: ABNORMAL Formal BTree normality indicator s |
710 |
|
length: 26 bytes |
711 |
|
|
712 |
|
This will fill C<$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; |
my $self = shift; |
718 |
return $self->{NXTMFN}; |
|
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 |
|
=item 0.22 |
766 |
|
|
767 |
|
Added field number when calling C<hash_filter> |
768 |
|
|
769 |
|
=item 0.21 |
770 |
|
|
771 |
|
Added C<join_subfields_with> to L</new> and L</to_hash>. |
772 |
|
|
773 |
|
Added C<include_subfields> to L</to_hash>. |
774 |
|
|
775 |
|
=item 0.20 |
776 |
|
|
777 |
|
Added C<< $isis->mfn >>, support for repeatable subfields and |
778 |
|
C<< $isis->to_hash({ mfn => 42, ... }) >> calling convention |
779 |
|
|
780 |
|
=back |
781 |
|
|
782 |
=head1 AUTHOR |
=head1 AUTHOR |
783 |
|
|
786 |
dpavlin@rot13.org |
dpavlin@rot13.org |
787 |
http://www.rot13.org/~dpavlin/ |
http://www.rot13.org/~dpavlin/ |
788 |
|
|
789 |
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 |
790 |
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. |
791 |
|
|
792 |
=head1 COPYRIGHT |
=head1 COPYRIGHT |
793 |
|
|
800 |
|
|
801 |
=head1 SEE ALSO |
=head1 SEE ALSO |
802 |
|
|
803 |
L<http://www.openisis.org|OpenIsis>, perl(1). |
L<Biblio::Isis::Manual> for CDS/ISIS manual appendix F, G and H which describe file format |
804 |
|
|
805 |
|
OpenIsis web site L<http://www.openisis.org> |
806 |
|
|
807 |
|
perl4lib site L<http://perl4lib.perl.org> |
808 |
|
|