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 = 0.01; |
$VERSION = 0.14; |
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 = shift; |
88 |
|
$v =~ s#foo#bar#g; |
89 |
|
}, |
90 |
debug => 1, |
debug => 1, |
91 |
); |
); |
92 |
|
|
96 |
|
|
97 |
=item isisdb |
=item isisdb |
98 |
|
|
99 |
Prefix path to CDS/ISIS. It should contain full or relative path to database |
This is full or relative path to ISIS database files which include |
100 |
and common prefix of C<.FDT>, C<.MST>, C<.CNT>, C<.XRF> and C<.MST> files. |
common prefix of C<.MST>, and C<.XRF> and optionally C<.FDT> (if using |
101 |
|
C<read_fdt> option) files. |
102 |
|
|
103 |
|
In this example it uses C<./cds/cds.MST> and related files. |
104 |
|
|
105 |
=item read_fdt |
=item read_fdt |
106 |
|
|
107 |
Boolean flag to specify if field definition table should be read. It's off |
Boolean flag to specify if field definition table should be read. It's off |
108 |
by default. |
by default. |
109 |
|
|
110 |
|
=item include_deleted |
111 |
|
|
112 |
|
Don't skip logically deleted records in ISIS. |
113 |
|
|
114 |
|
=item hash_filter |
115 |
|
|
116 |
|
Filter code ref which will be used before data is converted to hash. |
117 |
|
|
118 |
=item debug |
=item debug |
119 |
|
|
120 |
Dump a C<lot> of debugging output. |
Dump a B<lot> of debugging output. |
121 |
|
|
122 |
=back |
=back |
123 |
|
|
|
It will also set C<$isis-E<gt>{'maxmfn'}> which is maximum MFN stored in database. |
|
|
|
|
124 |
=cut |
=cut |
125 |
|
|
126 |
sub new { |
sub new { |
128 |
my $self = {}; |
my $self = {}; |
129 |
bless($self, $class); |
bless($self, $class); |
130 |
|
|
131 |
$self->{isisdb} = {@_}->{isisdb} || croak "new needs database name as argument!"; |
croak "new needs database name (isisdb) as argument!" unless ({@_}->{isisdb}); |
132 |
|
|
133 |
|
foreach my $v (qw{isisdb debug include_deleted hash_filter}) { |
134 |
|
$self->{$v} = {@_}->{$v}; |
135 |
|
} |
136 |
|
|
137 |
$self->{debug} = {@_}->{debug}; |
my @isis_files = grep(/\.(FDT|MST|XRF|CNT)$/i,glob($self->{isisdb}."*")); |
138 |
|
|
139 |
|
foreach my $f (@isis_files) { |
140 |
|
my $ext = $1 if ($f =~ m/\.(\w\w\w)$/); |
141 |
|
$self->{lc($ext)."_file"} = $f; |
142 |
|
} |
143 |
|
|
144 |
|
my @must_exist = qw(mst xrf); |
145 |
|
push @must_exist, "fdt" if ($self->{read_fdt}); |
146 |
|
|
147 |
|
foreach my $ext (@must_exist) { |
148 |
|
unless ($self->{$ext."_file"}) { |
149 |
|
carp "missing ",uc($ext)," file in ",$self->{isisdb}; |
150 |
|
return; |
151 |
|
} |
152 |
|
} |
153 |
|
|
154 |
|
if ($self->{debug}) { |
155 |
|
print STDERR "## using files: ",join(" ",@isis_files),"\n"; |
156 |
|
eval "use Data::Dump"; |
157 |
|
|
158 |
|
if (! $@) { |
159 |
|
*Dumper = *Data::Dump::dump; |
160 |
|
} else { |
161 |
|
use Data::Dumper; |
162 |
|
} |
163 |
|
} |
164 |
|
|
165 |
# 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! |
166 |
if ({@_}->{read_fdt} && -e $self->{isisdb}.".FDT") { |
if ($self->{read_fdt} && -e $self->{fdt_file}) { |
167 |
|
|
168 |
# read the $db.FDT file for tags |
# read the $db.FDT file for tags |
169 |
my $fieldzone=0; |
my $fieldzone=0; |
170 |
|
|
171 |
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}': $!"; |
172 |
|
binmode($fileFDT); |
173 |
|
|
174 |
while (<fileFDT>) { |
while (<$fileFDT>) { |
175 |
chomp; |
chomp; |
176 |
if ($fieldzone) { |
if ($fieldzone) { |
177 |
my $name=substr($_,0,30); |
my $name=substr($_,0,30); |
188 |
} |
} |
189 |
} |
} |
190 |
|
|
191 |
close(fileFDT); |
close($fileFDT); |
192 |
} |
} |
193 |
|
|
194 |
# Get the Maximum MFN from $db.MST |
# Get the Maximum MFN from $db.MST |
195 |
|
|
196 |
open(fileMST,$self->{isisdb}.".MST") || croak "can't read '$self->{isisdb}.MST': $!"; |
open($self->{'fileMST'}, $self->{mst_file}) || croak "can't open '$self->{mst_file}': $!"; |
197 |
|
binmode($self->{'fileMST'}); |
198 |
|
|
199 |
# MST format: (* = 32 bit signed) |
# MST format: (* = 32 bit signed) |
200 |
# CTLMFN* always 0 |
# CTLMFN* always 0 |
202 |
# NXTMFB* last block allocated to master file |
# NXTMFB* last block allocated to master file |
203 |
# NXTMFP offset to next available position in last block |
# NXTMFP offset to next available position in last block |
204 |
# MFTYPE always 0 for user db file (1 for system) |
# MFTYPE always 0 for user db file (1 for system) |
205 |
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"; |
|
206 |
|
|
207 |
# save maximum MFN |
my $buff; |
|
$self->{'maxmfn'} = $self->{'NXTMFN'} - 1; |
|
208 |
|
|
209 |
close(fileMST); |
read($self->{'fileMST'}, $buff, 4) || croak "can't read NXTMFN from MST: $!"; |
210 |
|
$self->{'NXTMFN'}=unpack("V",$buff) || croak "NXTNFN is zero"; |
211 |
|
|
212 |
# Get the index information from $db.CNT |
print STDERR "## self ",Dumper($self),"\n" if ($self->{debug}); |
|
|
|
|
open(fileCNT, $self->{isisdb}.".CNT") || croak "can't read '$self->{isisdb}.CNT': $!"; |
|
213 |
|
|
214 |
# There is two 26 Bytes fixed lenght records |
# open files for later |
215 |
|
open($self->{'fileXRF'}, $self->{xrf_file}) || croak "can't open '$self->{xrf_file}': $!"; |
216 |
|
binmode($self->{'fileXRF'}); |
217 |
|
|
218 |
# 0: IDTYPE BTree type 16 |
$self ? return $self : return undef; |
219 |
# 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); |
|
|
} |
|
|
} |
|
|
|
|
|
my $buff; |
|
|
read(fileCNT, $buff, 26); |
|
|
$self->unpack_cnt($buff); |
|
220 |
|
|
221 |
read(fileCNT, $buff, 26); |
=head2 count |
|
$self->unpack_cnt($buff); |
|
222 |
|
|
223 |
|
Return number of records in database |
224 |
|
|
225 |
close(fileCNT); |
print $isis->count; |
226 |
|
|
227 |
print Dumper($self) if ($self->{debug}); |
=cut |
228 |
|
|
229 |
$self ? return $self : return undef; |
sub count { |
230 |
|
my $self = shift; |
231 |
|
return $self->{'NXTMFN'} - 1; |
232 |
} |
} |
233 |
|
|
234 |
=head2 GetMFN |
=head2 fetch |
235 |
|
|
236 |
Read record with selected MFN |
Read record with selected MFN |
237 |
|
|
238 |
my $rec = $isis->GetMFN(55); |
my $rec = $isis->fetch(55); |
239 |
|
|
240 |
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 |
241 |
for that field. |
for that field like this: |
242 |
|
|
243 |
|
$rec = { |
244 |
|
'210' => [ '^aNew York^cNew York University press^dcop. 1988' ], |
245 |
|
'990' => [ '2140', '88', 'HAY' ], |
246 |
|
}; |
247 |
|
|
248 |
=cut |
=cut |
249 |
|
|
250 |
sub GetMFN { |
sub fetch { |
251 |
my $self = shift; |
my $self = shift; |
252 |
|
|
253 |
my $mfn = shift || croak "GetMFN needs MFN as argument!"; |
my $mfn = shift || croak "fetch needs MFN as argument!"; |
254 |
|
|
255 |
print "GetMFN: $mfn\n" if ($self->{debug}); |
# is mfn allready in memory? |
256 |
|
my $old_mfn = $self->{'current_mfn'} || -1; |
257 |
|
return $self->{record} if ($mfn == $old_mfn); |
258 |
|
|
259 |
open(fileXRF, $self->{isisdb}.".XRF") || croak "can't open '$self->{isisdb}.XRF': $!"; |
print STDERR "## fetch: $mfn\n" if ($self->{debug}); |
260 |
|
|
261 |
# XXX check this? |
# XXX check this? |
262 |
my $mfnpos=($mfn+int(($mfn-1)/127))*4; |
my $mfnpos=($mfn+int(($mfn-1)/127))*4; |
263 |
|
|
264 |
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 ($self->{debug}); |
265 |
seek(fileXRF,$mfnpos,0); |
seek($self->{'fileXRF'},$mfnpos,0); |
266 |
|
|
267 |
# read XRFMFB abd XRFMFP |
my $buff; |
|
my $pointer=$self->Read32(\*fileXRF); |
|
268 |
|
|
269 |
my $XRFMFB = int($pointer/2048); |
# delete old record |
270 |
my $XRFMFP = $pointer - ($XRFMFB*2048); |
delete $self->{record}; |
271 |
|
|
272 |
|
# read XRFMFB abd XRFMFP |
273 |
|
read($self->{'fileXRF'}, $buff, 4); |
274 |
|
my $pointer=unpack("V",$buff); |
275 |
|
if (! $pointer) { |
276 |
|
if ($self->{include_deleted}) { |
277 |
|
return; |
278 |
|
} else { |
279 |
|
warn "pointer for MFN $mfn is null\n"; |
280 |
|
return; |
281 |
|
} |
282 |
|
} |
283 |
|
|
284 |
print "XRFMFB: $XRFMFB XRFMFP: $XRFMFP\n" if ($self->{debug}); |
# check for logically deleted record |
285 |
|
if ($pointer & 0x80000000) { |
286 |
|
print STDERR "## record $mfn is logically deleted\n" if ($self->{debug}); |
287 |
|
$self->{deleted} = $mfn; |
288 |
|
|
289 |
# XXX fix this to be more readable!! |
return unless $self->{include_deleted}; |
|
# e.g. (XRFMFB - 1) * 512 + XRFMFP |
|
290 |
|
|
291 |
my $offset = $pointer; |
# abs |
292 |
my $offset2=int($offset/2048)-1; |
$pointer = ($pointer ^ 0xffffffff) + 1; |
|
my $offset22=int($offset/4096); |
|
|
my $offset3=$offset-($offset22*4096); |
|
|
if ($offset3>512) { |
|
|
$offset3=$offset3-2048; |
|
293 |
} |
} |
|
my $offset4=($offset2*512)+$offset3; |
|
294 |
|
|
295 |
print "$offset - $offset2 - $offset3 - $offset4\n" if ($self->{debug}); |
my $XRFMFB = int($pointer/2048); |
296 |
|
my $XRFMFP = $pointer - ($XRFMFB*2048); |
297 |
|
|
298 |
|
# (XRFMFB - 1) * 512 + XRFMFP |
299 |
|
# why do i have to do XRFMFP % 1024 ? |
300 |
|
|
301 |
|
my $blk_off = (($XRFMFB - 1) * 512) + ($XRFMFP % 512); |
302 |
|
|
303 |
close(fileXRF); |
print STDERR "## pointer: $pointer XRFMFB: $XRFMFB XRFMFP: $XRFMFP offset: $blk_off\n" if ($self->{'debug'}); |
304 |
|
|
305 |
# Get Record Information |
# Get Record Information |
306 |
|
|
307 |
open(fileMST, $self->{isisdb}.".MST") || croak "can't open '$self->{isisdb}.MST': $!"; |
seek($self->{'fileMST'},$blk_off,0) || croak "can't seek to $blk_off: $!"; |
308 |
|
|
309 |
seek(fileMST,$offset4,0); |
read($self->{'fileMST'}, $buff, 4) || croak "can't read 4 bytes at offset $blk_off from MST file: $!"; |
310 |
|
my $value=unpack("V",$buff); |
311 |
|
|
312 |
my $value=$self->Read32(\*fileMST); |
print STDERR "## offset for rowid $value is $blk_off (blk $XRFMFB off $XRFMFP)\n" if ($self->{debug}); |
313 |
|
|
314 |
if ($value!=$mfn) { |
if ($value!=$mfn) { |
315 |
print ("Error: The MFN:".$mfn." is not found in MST(".$value.")"); |
if ($value == 0) { |
316 |
return -1; # XXX deleted record? |
print STDERR "## record $mfn is physically deleted\n" if ($self->{debug}); |
317 |
|
$self->{deleted} = $mfn; |
318 |
|
return; |
319 |
|
} |
320 |
|
|
321 |
|
carp "Error: MFN ".$mfn." not found in MST file, found $value"; |
322 |
|
return; |
323 |
} |
} |
324 |
|
|
325 |
# $MFRL=$self->Read16($fileMST); |
read($self->{'fileMST'}, $buff, 14); |
|
# $MFBWB=$self->Read32($fileMST); |
|
|
# $MFBWP=$self->Read16($fileMST); |
|
|
# $BASE=$self->Read16($fileMST); |
|
|
# $NVF=$self->Read16($fileMST); |
|
|
# $STATUS=$self->Read16($fileMST); |
|
326 |
|
|
327 |
my $buff; |
my ($MFRL,$MFBWB,$MFBWP,$BASE,$NVF,$STATUS) = unpack("vVvvvv", $buff); |
328 |
read(fileMST, $buff, 14); |
|
329 |
|
print STDERR "## MFRL: $MFRL MFBWB: $MFBWB MFBWP: $MFBWP BASE: $BASE NVF: $NVF STATUS: $STATUS\n" if ($self->{debug}); |
330 |
|
|
331 |
my ($MFRL,$MFBWB,$MFBWP,$BASE,$NVF,$STATUS) = unpack("slssss", $buff); |
warn "MFRL $MFRL is not even number" unless ($MFRL % 2 == 0); |
332 |
|
|
333 |
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); |
334 |
|
|
335 |
# Get Directory Format |
# Get Directory Format |
336 |
|
|
338 |
my @FieldLEN; |
my @FieldLEN; |
339 |
my @FieldTAG; |
my @FieldTAG; |
340 |
|
|
341 |
for (my $i = 0 ; $i < $NVF ; $i++) { |
read($self->{'fileMST'}, $buff, 6 * $NVF); |
342 |
|
|
343 |
# $TAG=$self->Read16($fileMST); |
my $rec_len = 0; |
|
# $POS=$self->Read16($fileMST); |
|
|
# $LEN=$self->Read16($fileMST); |
|
344 |
|
|
345 |
read(fileMST, $buff, 6); |
for (my $i = 0 ; $i < $NVF ; $i++) { |
|
my ($TAG,$POS,$LEN) = unpack("sss", $buff); |
|
346 |
|
|
347 |
print "TAG: $TAG POS: $POS LEN: $LEN\n" if ($self->{debug}); |
my ($TAG,$POS,$LEN) = unpack("vvv", substr($buff,$i * 6, 6)); |
348 |
|
|
349 |
|
print STDERR "## TAG: $TAG POS: $POS LEN: $LEN\n" if ($self->{debug}); |
350 |
|
|
351 |
# 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. |
352 |
# |
# |
361 |
push @FieldTAG,$TAG; |
push @FieldTAG,$TAG; |
362 |
push @FieldPOS,$POS; |
push @FieldPOS,$POS; |
363 |
push @FieldLEN,$LEN; |
push @FieldLEN,$LEN; |
364 |
|
|
365 |
|
$rec_len += $LEN; |
366 |
} |
} |
367 |
|
|
368 |
# Get Variable Fields |
# Get Variable Fields |
369 |
|
|
370 |
delete $self->{record}; |
read($self->{'fileMST'},$buff,$rec_len); |
371 |
|
|
372 |
|
print STDERR "## rec_len: $rec_len poc: ",tell($self->{'fileMST'})."\n" if ($self->{debug}); |
373 |
|
|
374 |
for (my $i = 0 ; $i < $NVF ; $i++) { |
for (my $i = 0 ; $i < $NVF ; $i++) { |
375 |
my $rec; |
# skip zero-sized fields |
376 |
read(fileMST,$rec,$FieldLEN[$i]); |
next if ($FieldLEN[$i] == 0); |
|
push @{$self->{record}->{$FieldTAG[$i]}}, $rec; |
|
|
} |
|
|
close(fileMST); |
|
377 |
|
|
378 |
# The record is marked for deletion |
push @{$self->{record}->{$FieldTAG[$i]}}, substr($buff,$FieldPOS[$i],$FieldLEN[$i]); |
|
if ($STATUS==1) { |
|
|
return -1; |
|
379 |
} |
} |
380 |
|
|
381 |
print Dumper($self) if ($self->{debug}); |
$self->{'current_mfn'} = $mfn; |
382 |
|
|
383 |
|
print STDERR Dumper($self),"\n" if ($self->{debug}); |
384 |
|
|
385 |
return $self->{'record'}; |
return $self->{'record'}; |
386 |
} |
} |
387 |
|
|
388 |
=head2 to_ascii |
=head2 to_ascii |
389 |
|
|
390 |
Dump ascii output of selected MFN |
Returns ASCII output of record with specified MFN |
391 |
|
|
392 |
|
print $isis->to_ascii(42); |
393 |
|
|
394 |
print $isis->to_ascii(55); |
This outputs something like this: |
395 |
|
|
396 |
|
210 ^aNew York^cNew York University press^dcop. 1988 |
397 |
|
990 2140 |
398 |
|
990 88 |
399 |
|
990 HAY |
400 |
|
|
401 |
|
If C<read_fdt> is specified when calling C<new> it will display field names |
402 |
|
from C<.FDT> file instead of numeric tags. |
403 |
|
|
404 |
=cut |
=cut |
405 |
|
|
408 |
|
|
409 |
my $mfn = shift || croak "need MFN"; |
my $mfn = shift || croak "need MFN"; |
410 |
|
|
411 |
my $rec = $self->GetMFN($mfn); |
my $rec = $self->fetch($mfn) || return; |
412 |
|
|
413 |
my $out = "0\t$mfn"; |
my $out = "0\t$mfn"; |
414 |
|
|
415 |
foreach my $f (sort keys %{$rec}) { |
foreach my $f (sort keys %{$rec}) { |
416 |
$out .= "\n$f\t".join("\n$f\t",@{$self->{record}->{$f}}); |
my $fn = $self->tag_name($f); |
417 |
|
$out .= "\n$fn\t".join("\n$fn\t",@{$self->{record}->{$f}}); |
418 |
} |
} |
419 |
|
|
420 |
$out .= "\n"; |
$out .= "\n"; |
422 |
return $out; |
return $out; |
423 |
} |
} |
424 |
|
|
425 |
################# old cruft which is not ported from php to perl |
=head2 to_hash |
426 |
|
|
427 |
=begin php |
Read record with specified MFN and convert it to hash |
428 |
|
|
429 |
# Load the dictionary from the $db.L0x files. |
my $hash = $isis->to_hash($mfn); |
|
# Not usefull Yet |
|
|
|
|
|
sub LoadDictionary() |
|
|
{ |
|
|
$fileL01=fopen($self->{isisdb}.".L01","r"); |
|
|
rewind($fileL01); |
|
|
|
|
|
do |
|
|
{ |
|
|
|
|
|
$POS=$self->Read32($fileL01); |
|
|
$OCK=$self->Read16($fileL01); |
|
|
$IT=$self->Read16($fileL01); |
|
|
$PS=$self->Read32($fileL01); |
|
|
print "<br>PS:".$PS." ".$self->{ORDF}->{1}." "; |
|
|
for ($i=0;$i<$OCK;$i++) |
|
|
{ |
|
|
$KEY=fread($fileL01,10); |
|
|
|
|
|
print $KEY." ### "; |
|
|
|
|
|
$INFO1=$self->Read32($fileL01); |
|
|
$INFO2=$self->Read32($fileL01); |
|
|
|
|
|
#L01Key->{$key}=array($INFO1,$INFO2); |
|
|
} |
|
|
|
|
|
rewind($fileL01); |
|
|
$offset=($PS-1)*(12+$self->{ORDF}->{1}*18*2); |
|
|
fseek($fileL01,$offset); |
|
430 |
|
|
431 |
} While (!feof($fileL01)); |
It has ability to convert characters (using C<hash_filter>) from ISIS |
432 |
|
database before creating structures enabling character re-mapping or quick |
433 |
|
fix-up of data. |
434 |
|
|
435 |
|
This function returns hash which is like this: |
436 |
|
|
437 |
|
$hash = { |
438 |
|
'210' => [ |
439 |
|
{ |
440 |
|
'c' => 'New York University press', |
441 |
|
'a' => 'New York', |
442 |
|
'd' => 'cop. 1988' |
443 |
|
} |
444 |
|
], |
445 |
|
'990' => [ |
446 |
|
'2140', |
447 |
|
'88', |
448 |
|
'HAY' |
449 |
|
], |
450 |
|
}; |
451 |
|
|
452 |
|
You can later use that hash to produce any output from ISIS data. |
453 |
|
|
454 |
|
If database is created using IsisMarc, it will also have to special fields |
455 |
|
which will be used for identifiers, C<i1> and C<i2> like this: |
456 |
|
|
457 |
|
'200' => [ |
458 |
|
{ |
459 |
|
'i1' => '1', |
460 |
|
'i2' => ' ' |
461 |
|
'a' => 'Goa', |
462 |
|
'f' => 'Valdo D\'Arienzo', |
463 |
|
'e' => 'tipografie e tipografi nel XVI secolo', |
464 |
|
} |
465 |
|
], |
466 |
|
|
467 |
fclose($fileL01); |
This method will also create additional field C<000> with MFN. |
|
} |
|
468 |
|
|
469 |
# self function search through the tree and returns an array of pointers to IFP |
=cut |
|
# The function must be recursive |
|
470 |
|
|
471 |
sub SearchTree($search,$fileNB,$PUNT) |
sub to_hash { |
472 |
{ |
my $self = shift; |
|
$offset=(($PUNT-1)*(8+2*$self->{ORDN}->{1}*14)); |
|
|
|
|
|
rewind($fileNB1); |
|
|
|
|
|
fseek($fileNB,$offset); |
|
|
|
|
|
$POS=$self->Read32($fileNB); |
|
|
$OCK=$self->Read16($fileNB); |
|
|
$IT=$self->Read16($fileNB); |
|
|
|
|
|
#print "<br>".$POS." - ".$OCK." - ".$IT; |
|
|
|
|
|
$OLDPUNT=$POS; |
|
|
$j=0; |
|
|
for ($i=0;$i<$OCK;$i++) |
|
|
{ |
|
|
$KEY=fread($fileNB,10); |
|
|
|
|
|
$PUNT=$self->Read32($fileNB); |
|
|
|
|
|
#print " ## ".chop($KEY)."(".$PUNT."-".$OLDPUNT.") ## "; |
|
|
|
|
|
If (strcmp($search,chop($KEY))<0) |
|
|
{ |
|
|
break; |
|
|
} |
|
|
$OLDPUNT=$PUNT; |
|
|
} |
|
|
#print $OLDPUNT; |
|
|
Return $OLDPUNT; |
|
|
} |
|
473 |
|
|
474 |
# Search ISIS for record containing search |
my $mfn = shift || confess "need mfn!"; |
|
# Return a sorted array of MFN |
|
475 |
|
|
476 |
sub Search($search) |
# init record to include MFN as field 000 |
477 |
{ |
my $rec = { '000' => [ $mfn ] }; |
478 |
|
|
479 |
$search=strtoupper($search); |
my $row = $self->fetch($mfn) || return; |
|
#print "Searching....".$search." - ".$self->{POSRX}->{1}."<br>"; |
|
|
# first search .x01 |
|
|
|
|
|
|
|
|
# Search in .N01 |
|
|
|
|
|
|
|
|
$fileN01=fopen($self->{isisdb}.".N01","r"); |
|
|
$offset=(($self->{POSRX}->{1}-1)*(8+2*$self->{ORDN}->{1}*14)); |
|
|
|
|
|
do |
|
|
{ |
|
|
rewind($fileN01); |
|
|
|
|
|
fseek($fileN01,$offset); |
|
|
|
|
|
$POS=$self->Read32($fileN01); |
|
|
$OCK=$self->Read16($fileN01); |
|
|
$IT=$self->Read16($fileN01); |
|
|
|
|
|
#print "<br>".$POS." - ".$OCK." - ".$IT; |
|
|
|
|
|
$OLDPUNT=$POS; |
|
|
for ($i=0;$i<$OCK;$i++) |
|
|
{ |
|
|
$KEY=fread($fileN01,10); |
|
|
|
|
|
$PUNT=$self->Read32($fileN01); |
|
|
|
|
|
#print " ## ".chop($KEY)."(".$PUNT."-".$OLDPUNT.") ## "; |
|
|
|
|
|
If (strcmp($search,chop($KEY))<0) |
|
|
{ |
|
|
break; |
|
|
} |
|
|
$OLDPUNT=$PUNT; |
|
|
} |
|
|
$offset=(($OLDPUNT-1)*(8+2*$self->{ORDN}->{1}*14)); |
|
|
} while ($OLDPUNT>0); |
|
|
#print $OLDPUNT; |
|
|
|
|
|
|
|
|
fclose($fileN01); |
|
|
|
|
|
# Now look for records in .L01 file |
|
|
$fileL01=fopen($self->{isisdb}.".L01","r"); |
|
|
rewind($fileL01); |
|
|
|
|
|
$offset=(-$OLDPUNT-1)*(12+$self->{ORDF}->{1}*18*2); |
|
|
fseek($fileL01,$offset); |
|
|
|
|
|
$POS=$self->Read32($fileL01); |
|
|
$OCK=$self->Read16($fileL01); |
|
|
$IT=$self->Read16($fileL01); |
|
|
$PS=$self->Read32($fileL01); |
|
|
#print "<br>POS:".$POS." ".$self->{ORDF}->{1}." "; |
|
|
for ($i=0;$i<$OCK;$i++) |
|
|
{ |
|
|
$KEY=fread($fileL01,10); |
|
|
|
|
|
#print $KEY." ### "; |
|
|
|
|
|
$INFO1=$self->Read32($fileL01); |
|
|
$INFO2=$self->Read32($fileL01); |
|
|
|
|
|
If (strcmp($search,chop($KEY))==0) |
|
|
{ |
|
|
break; |
|
|
} |
|
|
} |
|
|
|
|
|
fclose($fileL01); |
|
|
|
|
|
#print $INFO1."--".$INFO2; |
|
|
|
|
|
# Now look in .IFP for the MFN |
|
|
$fileIFP=fopen($self->{isisdb}.".IFP","r"); |
|
|
rewind($fileIFP); |
|
|
$offset=($INFO1-1)*512+($INFO2*4); |
|
|
fseek($fileIFP,$offset); |
|
|
|
|
|
$IFPBLK=$self->Read32($fileIFP); |
|
|
|
|
|
$IFPNXTB=$self->Read32($fileIFP); |
|
|
$IFPNXTP=$self->Read32($fileIFP); |
|
|
$IFPTOTP=$self->Read32($fileIFP); |
|
|
$IFPSEGP=$self->Read32($fileIFP); |
|
|
$IFPSEGC=$self->Read32($fileIFP); |
|
|
|
|
|
|
|
|
#print "<br>IFP:".$IFPBLK." # ".$IFPNXTB." - ".$IFPNXTP." - ".$IFPTOTP." - ".$IFPSEGP." - ".$IFPSEGC; |
|
|
|
|
|
rewind($fileIFP); |
|
|
$offset=($INFO1-1)*512+24+($INFO2*4); |
|
|
fseek($fileIFP,$offset); |
|
|
|
|
|
$j=24+($INFO2*4); |
|
|
$k=0; |
|
|
$l=1; |
|
|
$OLDPMFN=""; |
|
|
for ($i=0;$i<$IFPSEGP;$i++) |
|
|
{ |
|
|
$B1=$self->Read8($fileIFP); |
|
|
$B2=$self->Read8($fileIFP); |
|
|
$B3=$self->Read8($fileIFP); |
|
|
$B4=$self->Read8($fileIFP); |
|
|
$B5=$self->Read8($fileIFP); |
|
|
$B6=$self->Read8($fileIFP); |
|
|
$B7=$self->Read8($fileIFP); |
|
|
$B8=$self->Read8($fileIFP); |
|
|
|
|
|
$PMFN=$B1*65536+$B2*256+$B3; |
|
|
$PTAG=$B4*256+$B5; |
|
|
$POCC=$B6; |
|
|
$PCNT=$B7*256+$B8; |
|
|
|
|
|
if ($OLDPMFN!=$PMFN) |
|
|
{ |
|
|
if ($PMFN!=0) |
|
|
{ |
|
|
$self->{MFNArray}->{$l}=$PMFN; |
|
|
$OLDPMFN=$PMFN; |
|
|
$l+=1; |
|
|
} |
|
|
} |
|
|
|
|
|
$j=$j+8; |
|
|
#print "<br>".$PMFN."-".$PTAG." - ".$POCC." - ".$PCNT; |
|
|
#print "@@".$j."@@@@"; |
|
|
if ($j>=504) |
|
|
{ |
|
|
if ($IFPNXTB==0 && $IFPNXTP==0) |
|
|
{ |
|
|
$k=$k+1; |
|
|
rewind($fileIFP); |
|
|
$offset=($INFO1-1+$k)*512; |
|
|
fseek($fileIFP,$offset); |
|
|
$B=$self->Read32($fileIFP); |
|
|
#print "<br>-".$B."-<br>"; |
|
|
$j=0; |
|
|
} else |
|
|
{ |
|
|
rewind($fileIFP); |
|
|
$offset=($IFPNXTB-1)*512; |
|
|
fseek($fileIFP,$offset); |
|
|
|
|
|
$OLDIFPNXTB=$IFPNXTB; |
|
|
$OLDIFPNXTP=$IFPNXTP; |
|
|
|
|
|
$IFPBLK=$self->Read32($fileIFP); |
|
|
|
|
|
$IFPNXTB=$self->Read32($fileIFP); |
|
|
$IFPNXTP=$self->Read32($fileIFP); |
|
|
$IFPTOTP=$self->Read32($fileIFP); |
|
|
$IFPSEGP=$self->Read32($fileIFP); |
|
|
$IFPSEGC=$self->Read32($fileIFP); |
|
|
|
|
|
rewind($fileIFP); |
|
|
$offset=($OLDIFPNXTB-1)*512+24+($OLDIFPNXTP*4); |
|
|
fseek($fileIFP,$offset); |
|
|
|
|
|
$j=24+($OLDIFPNXTP*4); |
|
|
$k=0; |
|
|
$j=0; |
|
|
} |
|
|
} |
|
|
|
|
|
} |
|
|
fclose($fileIFP); |
|
|
return $l-1; |
|
|
} |
|
480 |
|
|
481 |
=cut |
foreach my $k (keys %{$row}) { |
482 |
|
foreach my $l (@{$row->{$k}}) { |
483 |
|
|
484 |
# |
# filter output |
485 |
# XXX porting from php left-over: |
if ($self->{'hash_filter'}) { |
486 |
# |
$l = $self->{'hash_filter'}->($l); |
487 |
# do I *REALLY* need those methods, or should I use |
next unless defined($l); |
488 |
# $self->{something} directly? |
} |
|
# |
|
|
# Probably direct usage is better! |
|
|
# |
|
489 |
|
|
490 |
sub GetFieldName { |
my $val; |
|
my $self = shift; |
|
|
return $self->{FieldName}; |
|
|
} |
|
491 |
|
|
492 |
sub GetTagName { |
# has identifiers? |
493 |
my $self = shift; |
($val->{'i1'},$val->{'i2'}) = ($1,$2) if ($l =~ s/^([01 #])([01 #])\^/\^/); |
|
return $self->{TagName}; |
|
|
} |
|
494 |
|
|
495 |
sub GetFieldTag { |
# has subfields? |
496 |
my $self = shift; |
if ($l =~ m/\^/) { |
497 |
return $self->{FieldTAG}; |
foreach my $t (split(/\^/,$l)) { |
498 |
} |
next if (! $t); |
499 |
|
$val->{substr($t,0,1)} = substr($t,1); |
500 |
|
} |
501 |
|
} else { |
502 |
|
$val = $l; |
503 |
|
} |
504 |
|
|
505 |
sub GetNextMFN { |
push @{$rec->{$k}}, $val; |
506 |
my $self = shift; |
} |
507 |
return $self->{NXTMFN}; |
} |
508 |
|
|
509 |
|
return $rec; |
510 |
} |
} |
511 |
|
|
512 |
sub GetMFNArray { |
=head2 tag_name |
513 |
|
|
514 |
|
Return name of selected tag |
515 |
|
|
516 |
|
print $isis->tag_name('200'); |
517 |
|
|
518 |
|
=cut |
519 |
|
|
520 |
|
sub tag_name { |
521 |
my $self = shift; |
my $self = shift; |
522 |
return $self->{MFNArray}; |
my $tag = shift || return; |
523 |
|
return $self->{'TagName'}->{$tag} || $tag; |
524 |
} |
} |
|
=begin php |
|
525 |
|
|
|
sub Read32($fileNB) |
|
|
{ |
|
|
$B1=ord(fread($fileNB,1)); |
|
|
$B2=ord(fread($fileNB,1)); |
|
|
$B3=ord(fread($fileNB,1)); |
|
|
$B4=ord(fread($fileNB,1)); |
|
|
|
|
|
if ($B4<=128) |
|
|
{ |
|
|
$value=$B1+$B2*256+$B3*65536+$B4*16777216; |
|
|
} else |
|
|
{ |
|
|
$value=$self->Not8($B1)+$self->Not8($B2)*256+$self->Not8($B3)*65536+$self->Not8($B4)*16777216; |
|
|
$value=-($value+1); |
|
|
} |
|
|
# print "(".$B1.",".$B2.",".$B3.",".$B4.":".$value.")"; |
|
526 |
|
|
527 |
return $value; |
=head2 read_cnt |
|
} |
|
528 |
|
|
529 |
sub Read24($fileNB) |
Read content of C<.CNT> file and return hash containing it. |
|
{ |
|
|
$B1=ord(fread($fileNB,1)); |
|
|
$B2=ord(fread($fileNB,1)); |
|
|
$B3=ord(fread($fileNB,1)); |
|
530 |
|
|
531 |
$value=$B1+$B2*256+$B3*65536; |
print Dumper($isis->read_cnt); |
532 |
|
|
533 |
# print "(".$B1.",".$B2.",".$B3.":".$value.")"; |
This function is not used by module (C<.CNT> files are not required for this |
534 |
|
module to work), but it can be useful to examine your index (while debugging |
535 |
|
for example). |
536 |
|
|
537 |
return $value; |
=cut |
|
} |
|
538 |
|
|
539 |
sub Read16($fileNB) |
sub read_cnt { |
540 |
{ |
my $self = shift; |
|
$B1=ord(fread($fileNB,1)); |
|
|
$B2=ord(fread($fileNB,1)); |
|
541 |
|
|
542 |
$value=$B1+$B2*256; |
croak "missing CNT file in ",$self->{isisdb} unless ($self->{cnt_file}); |
|
# print "(".$B1.",".$B2.":".$value.")"; |
|
543 |
|
|
544 |
return $value; |
# Get the index information from $db.CNT |
545 |
} |
|
546 |
|
open(my $fileCNT, $self->{cnt_file}) || croak "can't read '$self->{cnt_file}': $!"; |
547 |
|
binmode($fileCNT); |
548 |
|
|
549 |
sub Read8($fileNB) |
my $buff; |
|
{ |
|
|
$B1=ord(fread($fileNB,1)); |
|
550 |
|
|
551 |
$value=$B1; |
read($fileCNT, $buff, 26) || croak "can't read first table from CNT: $!"; |
552 |
# print "(".$value.")"; |
$self->unpack_cnt($buff); |
553 |
|
|
554 |
return $value; |
read($fileCNT, $buff, 26) || croak "can't read second table from CNT: $!"; |
555 |
} |
$self->unpack_cnt($buff); |
556 |
|
|
557 |
sub Not8($value) |
close($fileCNT); |
558 |
{ |
|
559 |
$value=decbin($value); |
return $self->{cnt}; |
|
if (strlen($value)<8) |
|
|
{ |
|
|
$buffer=""; |
|
|
for($i=0;$i<(8-strlen($value));$i++) |
|
|
{ |
|
|
$buffer.="0"; |
|
|
} |
|
|
$value=$buffer.$value; |
|
|
} |
|
|
$value=ereg_replace("0","3",$value); |
|
|
$value=ereg_replace("1","0",$value); |
|
|
$value=ereg_replace("3","1",$value); |
|
|
$value=bindec($value); |
|
|
return $value; |
|
|
} |
|
560 |
} |
} |
561 |
|
|
562 |
|
=head2 unpack_cnt |
563 |
|
|
564 |
|
Unpack one of two 26 bytes fixed length record in C<.CNT> file. |
565 |
|
|
566 |
|
Here is definition of record: |
567 |
|
|
568 |
|
off key description size |
569 |
|
0: IDTYPE BTree type s |
570 |
|
2: ORDN Nodes Order s |
571 |
|
4: ORDF Leafs Order s |
572 |
|
6: N Number of Memory buffers for nodes s |
573 |
|
8: K Number of buffers for first level index s |
574 |
|
10: LIV Current number of Index Levels s |
575 |
|
12: POSRX Pointer to Root Record in N0x l |
576 |
|
16: NMAXPOS Next Available position in N0x l |
577 |
|
20: FMAXPOS Next available position in L0x l |
578 |
|
24: ABNORMAL Formal BTree normality indicator s |
579 |
|
length: 26 bytes |
580 |
|
|
581 |
|
This will fill C<$self> object under C<cnt> with hash. It's used by C<read_cnt>. |
582 |
|
|
583 |
=cut |
=cut |
584 |
|
|
585 |
|
sub unpack_cnt { |
586 |
|
my $self = shift; |
587 |
|
|
588 |
|
my @flds = qw(ORDN ORDF N K LIV POSRX NMAXPOS FMAXPOS ABNORMAL); |
589 |
|
|
590 |
|
my $buff = shift || return; |
591 |
|
my @arr = unpack("vvvvvvVVVv", $buff); |
592 |
|
|
593 |
|
print STDERR "unpack_cnt: ",join(" ",@arr),"\n" if ($self->{'debug'}); |
594 |
|
|
595 |
|
my $IDTYPE = shift @arr; |
596 |
|
foreach (@flds) { |
597 |
|
$self->{cnt}->{$IDTYPE}->{$_} = abs(shift @arr); |
598 |
|
} |
599 |
|
} |
600 |
|
|
601 |
1; |
1; |
|
__END__ |
|
602 |
|
|
603 |
=head1 BUGS |
=head1 BUGS |
604 |
|
|
605 |
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 |
606 |
|
some variations in input databases which has been tested with this module. |
607 |
|
When I was in doubt, I assumed that OpenIsis's implementation was right |
608 |
|
(except for obvious bugs). |
609 |
|
|
610 |
|
However, every effort has been made to test this module with as much |
611 |
|
databases (and programs that create them) as possible. |
612 |
|
|
613 |
|
I would be very greatful for success or failure reports about usage of this |
614 |
|
module with databases from programs other than WinIsis and IsisMarc. I had |
615 |
|
tested this against ouput of one C<isis.dll>-based application, but I don't |
616 |
|
know any details about it's version. |
617 |
|
|
618 |
=head1 AUTHOR |
=head1 AUTHOR |
619 |
|
|
622 |
dpavlin@rot13.org |
dpavlin@rot13.org |
623 |
http://www.rot13.org/~dpavlin/ |
http://www.rot13.org/~dpavlin/ |
624 |
|
|
625 |
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 |
626 |
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. |
627 |
|
|
628 |
=head1 COPYRIGHT |
=head1 COPYRIGHT |
629 |
|
|
636 |
|
|
637 |
=head1 SEE ALSO |
=head1 SEE ALSO |
638 |
|
|
639 |
L<http://www.openisis.org|OpenIsis>, perl(1). |
OpenIsis web site L<http://www.openisis.org> |
640 |
|
|
641 |
|
perl4lib site L<http://perl4lib.perl.org> |
642 |
|
|