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 = 0.03; |
$VERSION = 0.20; |
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, |
|
debug => 1, |
|
85 |
include_deleted => 1, |
include_deleted => 1, |
86 |
|
hash_filter => sub { |
87 |
|
my $v = shift; |
88 |
|
$v =~ s#foo#bar#g; |
89 |
|
}, |
90 |
|
debug => 1, |
91 |
); |
); |
92 |
|
|
93 |
Options are described below: |
Options are described below: |
96 |
|
|
97 |
=item isisdb |
=item isisdb |
98 |
|
|
99 |
Prefix path to CDS/ISIS. It should contain full or relative path to database |
This is full or relative path to ISIS database files which include |
100 |
and common prefix of C<.FDT>, C<.MST>, C<.CNT>, C<.XRF> and C<.MST> files. |
common prefix of C<.MST>, and C<.XRF> and optionally C<.FDT> (if using |
101 |
|
C<read_fdt> option) files. |
102 |
|
|
103 |
|
In this example it uses C<./cds/cds.MST> and related files. |
104 |
|
|
105 |
=item read_fdt |
=item read_fdt |
106 |
|
|
107 |
Boolean flag to specify if field definition table should be read. It's off |
Boolean flag to specify if field definition table should be read. It's off |
108 |
by default. |
by default. |
109 |
|
|
110 |
=item debug |
=item include_deleted |
111 |
|
|
112 |
Dump a C<lot> of debugging output. |
Don't skip logically deleted records in ISIS. |
113 |
|
|
114 |
=item include_deleted |
=item hash_filter |
115 |
|
|
116 |
Don't skip logically deleted records. |
Filter code ref which will be used before data is converted to hash. |
117 |
|
|
118 |
=back |
=item debug |
119 |
|
|
120 |
|
Dump a B<lot> of debugging output even at level 1. For even more increase level. |
121 |
|
|
122 |
It will also set C<$isis-E<gt>{'maxmfn'}> which is maximum MFN stored in database. |
=back |
123 |
|
|
124 |
=cut |
=cut |
125 |
|
|
130 |
|
|
131 |
croak "new needs database name (isisdb) as argument!" unless ({@_}->{isisdb}); |
croak "new needs database name (isisdb) as argument!" unless ({@_}->{isisdb}); |
132 |
|
|
133 |
foreach my $v (qw{isisdb debug include_deleted}) { |
foreach my $v (qw{isisdb debug include_deleted hash_filter}) { |
134 |
$self->{$v} = {@_}->{$v}; |
$self->{$v} = {@_}->{$v}; |
135 |
} |
} |
136 |
|
|
137 |
|
my @isis_files = grep(/\.(FDT|MST|XRF|CNT)$/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"; |
|
|
|
|
|
# save maximum MFN |
|
|
$self->{'maxmfn'} = $self->{'NXTMFN'} - 1; |
|
|
|
|
|
close(fileMST); |
|
|
|
|
|
# Get the index information from $db.CNT |
|
|
|
|
|
open(fileCNT, $self->{isisdb}.".CNT") || croak "can't read '$self->{isisdb}.CNT': $!"; |
|
206 |
|
|
207 |
# There is two 26 Bytes fixed lenght records |
my $buff; |
208 |
|
|
209 |
# 0: IDTYPE BTree type 16 |
read($self->{'fileMST'}, $buff, 4) || croak "can't read NXTMFN from MST: $!"; |
210 |
# 2: ORDN Nodes Order 16 |
$self->{'NXTMFN'}=unpack("V",$buff) || croak "NXTNFN is zero"; |
|
# 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", $buff); |
|
|
|
|
|
print "unpack_cnt: ",join(" ",@arr),"\n" if ($self->{'debug'}); |
|
|
|
|
|
my $IDTYPE = shift @arr; |
|
|
foreach (@flds) { |
|
|
$self->{$IDTYPE}->{$_} = abs(shift @arr); |
|
|
} |
|
|
} |
|
211 |
|
|
212 |
my $buff; |
print STDERR "## self ",Dumper($self),"\n" if ($self->{debug}); |
|
read(fileCNT, $buff, 26); |
|
|
$self->unpack_cnt($buff); |
|
213 |
|
|
214 |
read(fileCNT, $buff, 26); |
# open files for later |
215 |
$self->unpack_cnt($buff); |
open($self->{'fileXRF'}, $self->{xrf_file}) || croak "can't open '$self->{xrf_file}': $!"; |
216 |
|
binmode($self->{'fileXRF'}); |
217 |
|
|
218 |
|
$self ? return $self : return undef; |
219 |
|
} |
220 |
|
|
221 |
close(fileCNT); |
=head2 count |
222 |
|
|
223 |
print Dumper($self) if ($self->{debug}); |
Return number of records in database |
224 |
|
|
225 |
# open files for later |
print $isis->count; |
|
open($self->{'fileXRF'}, $self->{isisdb}.".XRF") || croak "can't open '$self->{isisdb}.XRF': $!"; |
|
226 |
|
|
227 |
open($self->{'fileMST'}, $self->{isisdb}.".MST") || croak "can't open '$self->{isisdb}.MST': $!"; |
=cut |
228 |
|
|
229 |
$self ? return $self : return undef; |
sub count { |
230 |
|
my $self = shift; |
231 |
|
return $self->{'NXTMFN'} - 1; |
232 |
} |
} |
233 |
|
|
234 |
=head2 fetch |
=head2 fetch |
238 |
my $rec = $isis->fetch(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 |
|
|
252 |
|
|
253 |
my $mfn = shift || croak "fetch needs MFN as argument!"; |
my $mfn = shift || croak "fetch needs MFN as argument!"; |
254 |
|
|
255 |
print "fetch: $mfn\n" if ($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 |
|
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 ($self->{debug}); |
265 |
seek($self->{'fileXRF'},$mfnpos,0); |
seek($self->{'fileXRF'},$mfnpos,0); |
266 |
|
|
267 |
|
my $buff; |
268 |
|
|
269 |
|
# delete old record |
270 |
|
delete $self->{record}; |
271 |
|
|
272 |
# read XRFMFB abd XRFMFP |
# read XRFMFB abd XRFMFP |
273 |
my $pointer=$self->Read32(\*{$self->{'fileXRF'}}); |
read($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 $mfn is null\n"; |
280 |
|
return; |
281 |
|
} |
282 |
|
} |
283 |
|
|
284 |
|
# check for logically deleted record |
285 |
|
if ($pointer & 0x80000000) { |
286 |
|
print STDERR "## record $mfn is logically deleted\n" if ($self->{debug}); |
287 |
|
$self->{deleted} = $mfn; |
288 |
|
|
289 |
|
return unless $self->{include_deleted}; |
290 |
|
|
291 |
|
# abs |
292 |
|
$pointer = ($pointer ^ 0xffffffff) + 1; |
293 |
|
} |
294 |
|
|
295 |
my $XRFMFB = int($pointer/2048); |
my $XRFMFB = int($pointer/2048); |
296 |
my $XRFMFP = $pointer - ($XRFMFB*2048); |
my $XRFMFP = $pointer - ($XRFMFB*2048); |
297 |
|
|
298 |
print "XRFMFB: $XRFMFB XRFMFP: $XRFMFP\n" if ($self->{debug}); |
# (XRFMFB - 1) * 512 + XRFMFP |
299 |
|
# why do i have to do XRFMFP % 1024 ? |
300 |
|
|
301 |
# XXX fix this to be more readable!! |
my $blk_off = (($XRFMFB - 1) * 512) + ($XRFMFP % 512); |
|
# e.g. (XRFMFB - 1) * 512 + XRFMFP |
|
302 |
|
|
303 |
my $offset = $pointer; |
print STDERR "## pointer: $pointer XRFMFB: $XRFMFB XRFMFP: $XRFMFP offset: $blk_off\n" if ($self->{'debug'}); |
|
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)+$offset3; |
|
|
|
|
|
print "$offset - $offset2 - $offset3 - $offset4\n" if ($self->{debug}); |
|
304 |
|
|
305 |
# Get Record Information |
# Get Record Information |
306 |
|
|
307 |
seek($self->{'fileMST'},$offset4,0); |
seek($self->{'fileMST'},$blk_off,0) || croak "can't seek to $blk_off: $!"; |
308 |
|
|
309 |
|
read($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(\*{$self->{'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 |
# $MFRL=$self->Read16($fileMST); |
carp "Error: MFN ".$mfn." not found in MST file, found $value"; |
322 |
# $MFBWB=$self->Read32($fileMST); |
return; |
323 |
# $MFBWP=$self->Read16($fileMST); |
} |
|
# $BASE=$self->Read16($fileMST); |
|
|
# $NVF=$self->Read16($fileMST); |
|
|
# $STATUS=$self->Read16($fileMST); |
|
324 |
|
|
|
my $buff; |
|
325 |
read($self->{'fileMST'}, $buff, 14); |
read($self->{'fileMST'}, $buff, 14); |
326 |
|
|
327 |
my ($MFRL,$MFBWB,$MFBWP,$BASE,$NVF,$STATUS) = unpack("slssss", $buff); |
my ($MFRL,$MFBWB,$MFBWP,$BASE,$NVF,$STATUS) = unpack("vVvvvv", $buff); |
328 |
|
|
329 |
print "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}); |
330 |
|
|
331 |
# delete old record |
warn "MFRL $MFRL is not even number" unless ($MFRL % 2 == 0); |
|
delete $self->{record}; |
|
332 |
|
|
333 |
if (! $self->{'include_deleted'} && $MFRL < 0) { |
warn "BASE is not 18+6*NVF" unless ($BASE == 18 + 6 * $NVF); |
|
print "## logically deleted record $mfn, skipping...\n" if ($self->{debug}); |
|
|
return; |
|
|
} |
|
334 |
|
|
335 |
# Get Directory Format |
# Get Directory Format |
336 |
|
|
340 |
|
|
341 |
read($self->{'fileMST'}, $buff, 6 * $NVF); |
read($self->{'fileMST'}, $buff, 6 * $NVF); |
342 |
|
|
343 |
my $fld_len = 0; |
my $rec_len = 0; |
344 |
|
|
345 |
for (my $i = 0 ; $i < $NVF ; $i++) { |
for (my $i = 0 ; $i < $NVF ; $i++) { |
346 |
|
|
347 |
# $TAG=$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)); |
|
348 |
|
|
349 |
print "TAG: $TAG POS: $POS LEN: $LEN\n" if ($self->{debug}); |
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 |
# |
# |
362 |
push @FieldPOS,$POS; |
push @FieldPOS,$POS; |
363 |
push @FieldLEN,$LEN; |
push @FieldLEN,$LEN; |
364 |
|
|
365 |
$fld_len += $LEN; |
$rec_len += $LEN; |
366 |
} |
} |
367 |
|
|
368 |
# Get Variable Fields |
# Get Variable Fields |
369 |
|
|
370 |
read($self->{'fileMST'},$buff,$fld_len); |
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 ; $i++) { |
375 |
# skip zero-sized fields |
# skip zero-sized fields |
377 |
|
|
378 |
push @{$self->{record}->{$FieldTAG[$i]}}, substr($buff,$FieldPOS[$i],$FieldLEN[$i]); |
push @{$self->{record}->{$FieldTAG[$i]}}, substr($buff,$FieldPOS[$i],$FieldLEN[$i]); |
379 |
} |
} |
|
close(fileMST); |
|
380 |
|
|
381 |
print Dumper($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 mfn |
389 |
|
|
390 |
|
Returns current MFN position |
391 |
|
|
392 |
|
my $mfn = $isis->mfn; |
393 |
|
|
394 |
|
=cut |
395 |
|
|
396 |
|
# This function should be simple return $self->{current_mfn}, |
397 |
|
# but if new is called with _hack_mfn it becomes setter. |
398 |
|
# It's useful in tests when setting $isis->{record} directly |
399 |
|
|
400 |
|
sub mfn { |
401 |
|
my $self = shift; |
402 |
|
return $self->{current_mfn}; |
403 |
|
}; |
404 |
|
|
405 |
|
|
406 |
=head2 to_ascii |
=head2 to_ascii |
407 |
|
|
408 |
Dump ascii output of selected MFN |
Returns ASCII output of record with specified MFN |
409 |
|
|
410 |
|
print $isis->to_ascii(42); |
411 |
|
|
412 |
|
This outputs something like this: |
413 |
|
|
414 |
print $isis->to_ascii(55); |
210 ^aNew York^cNew York University press^dcop. 1988 |
415 |
|
990 2140 |
416 |
|
990 88 |
417 |
|
990 HAY |
418 |
|
|
419 |
|
If C<read_fdt> is specified when calling C<new> it will display field names |
420 |
|
from C<.FDT> file instead of numeric tags. |
421 |
|
|
422 |
=cut |
=cut |
423 |
|
|
426 |
|
|
427 |
my $mfn = shift || croak "need MFN"; |
my $mfn = shift || croak "need MFN"; |
428 |
|
|
429 |
my $rec = $self->fetch($mfn); |
my $rec = $self->fetch($mfn) || return; |
430 |
|
|
431 |
my $out = "0\t$mfn"; |
my $out = "0\t$mfn"; |
432 |
|
|
433 |
foreach my $f (sort keys %{$rec}) { |
foreach my $f (sort keys %{$rec}) { |
434 |
$out .= "\n$f\t".join("\n$f\t",@{$self->{record}->{$f}}); |
my $fn = $self->tag_name($f); |
435 |
|
$out .= "\n$fn\t".join("\n$fn\t",@{$self->{record}->{$f}}); |
436 |
} |
} |
437 |
|
|
438 |
$out .= "\n"; |
$out .= "\n"; |
440 |
return $out; |
return $out; |
441 |
} |
} |
442 |
|
|
443 |
# |
=head2 to_hash |
444 |
# XXX porting from php left-over: |
|
445 |
# |
Read record with specified MFN and convert it to hash |
446 |
# do I *REALLY* need those methods, or should I use |
|
447 |
# $self->{something} directly? |
my $hash = $isis->to_hash($mfn); |
448 |
# |
|
449 |
# Probably direct usage is better! |
It has ability to convert characters (using C<hash_filter>) from ISIS |
450 |
# |
database before creating structures enabling character re-mapping or quick |
451 |
|
fix-up of data. |
452 |
|
|
453 |
|
This function returns hash which is like this: |
454 |
|
|
455 |
|
$hash = { |
456 |
|
'210' => [ |
457 |
|
{ |
458 |
|
'c' => 'New York University press', |
459 |
|
'a' => 'New York', |
460 |
|
'd' => 'cop. 1988' |
461 |
|
} |
462 |
|
], |
463 |
|
'990' => [ |
464 |
|
'2140', |
465 |
|
'88', |
466 |
|
'HAY' |
467 |
|
], |
468 |
|
}; |
469 |
|
|
470 |
|
You can later use that hash to produce any output from ISIS data. |
471 |
|
|
472 |
|
If database is created using IsisMarc, it will also have to special fields |
473 |
|
which will be used for identifiers, C<i1> and C<i2> like this: |
474 |
|
|
475 |
|
'200' => [ |
476 |
|
{ |
477 |
|
'i1' => '1', |
478 |
|
'i2' => ' ' |
479 |
|
'a' => 'Goa', |
480 |
|
'f' => 'Valdo D\'Arienzo', |
481 |
|
'e' => 'tipografie e tipografi nel XVI secolo', |
482 |
|
} |
483 |
|
], |
484 |
|
|
485 |
|
In case there are repeatable subfields in record, this will create |
486 |
|
following structure: |
487 |
|
|
488 |
|
'900' => [ { |
489 |
|
'a' => [ 'foo', 'bar', 'baz' ], |
490 |
|
}] |
491 |
|
|
492 |
|
This method will also create additional field C<000> with MFN. |
493 |
|
|
494 |
|
=cut |
495 |
|
|
496 |
|
sub to_hash { |
497 |
|
my $self = shift; |
498 |
|
|
499 |
|
my $mfn = shift || confess "need mfn!"; |
500 |
|
|
501 |
|
# init record to include MFN as field 000 |
502 |
|
my $rec = { '000' => [ $mfn ] }; |
503 |
|
|
504 |
|
my $row = $self->fetch($mfn) || return; |
505 |
|
|
506 |
|
foreach my $k (keys %{$row}) { |
507 |
|
foreach my $l (@{$row->{$k}}) { |
508 |
|
|
509 |
|
# filter output |
510 |
|
if ($self->{'hash_filter'}) { |
511 |
|
$l = $self->{'hash_filter'}->($l); |
512 |
|
next unless defined($l); |
513 |
|
} |
514 |
|
|
515 |
|
my $val; |
516 |
|
|
517 |
|
# has identifiers? |
518 |
|
($val->{'i1'},$val->{'i2'}) = ($1,$2) if ($l =~ s/^([01 #])([01 #])\^/\^/); |
519 |
|
|
520 |
|
# has subfields? |
521 |
|
if ($l =~ m/\^/) { |
522 |
|
foreach my $t (split(/\^/,$l)) { |
523 |
|
next if (! $t); |
524 |
|
my ($sf,$v) = (substr($t,0,1), substr($t,1)); |
525 |
|
# FIXME make this option ! |
526 |
|
next unless ($v); |
527 |
|
# warn "### $k^$sf:$v",$/ if ($self->{debug} > 1); |
528 |
|
|
529 |
|
# FIXME array return optional, by default unroll to ' ; ' |
530 |
|
if (ref( $val->{$sf} ) eq 'ARRAY') { |
531 |
|
|
532 |
|
push @{ $val->{$sf} }, $v; |
533 |
|
} elsif (defined( $val->{$sf} )) { |
534 |
|
# convert scalar field to array |
535 |
|
$val->{$sf} = [ $val->{$sf}, $v ]; |
536 |
|
} else { |
537 |
|
$val->{$sf} = $v; |
538 |
|
} |
539 |
|
} |
540 |
|
} else { |
541 |
|
$val = $l; |
542 |
|
} |
543 |
|
|
544 |
|
push @{$rec->{$k}}, $val; |
545 |
|
} |
546 |
|
} |
547 |
|
|
548 |
|
return $rec; |
549 |
|
} |
550 |
|
|
551 |
|
=head2 tag_name |
552 |
|
|
553 |
|
Return name of selected tag |
554 |
|
|
555 |
|
print $isis->tag_name('200'); |
556 |
|
|
557 |
|
=cut |
558 |
|
|
559 |
|
sub tag_name { |
560 |
|
my $self = shift; |
561 |
|
my $tag = shift || return; |
562 |
|
return $self->{'TagName'}->{$tag} || $tag; |
563 |
|
} |
564 |
|
|
565 |
|
|
566 |
sub TagName { |
=head2 read_cnt |
567 |
|
|
568 |
|
Read content of C<.CNT> file and return hash containing it. |
569 |
|
|
570 |
|
print Dumper($isis->read_cnt); |
571 |
|
|
572 |
|
This function is not used by module (C<.CNT> files are not required for this |
573 |
|
module to work), but it can be useful to examine your index (while debugging |
574 |
|
for example). |
575 |
|
|
576 |
|
=cut |
577 |
|
|
578 |
|
sub read_cnt { |
579 |
my $self = shift; |
my $self = shift; |
580 |
return $self->{TagName}; |
|
581 |
|
croak "missing CNT file in ",$self->{isisdb} unless ($self->{cnt_file}); |
582 |
|
|
583 |
|
# Get the index information from $db.CNT |
584 |
|
|
585 |
|
open(my $fileCNT, $self->{cnt_file}) || croak "can't read '$self->{cnt_file}': $!"; |
586 |
|
binmode($fileCNT); |
587 |
|
|
588 |
|
my $buff; |
589 |
|
|
590 |
|
read($fileCNT, $buff, 26) || croak "can't read first table from CNT: $!"; |
591 |
|
$self->unpack_cnt($buff); |
592 |
|
|
593 |
|
read($fileCNT, $buff, 26) || croak "can't read second table from CNT: $!"; |
594 |
|
$self->unpack_cnt($buff); |
595 |
|
|
596 |
|
close($fileCNT); |
597 |
|
|
598 |
|
return $self->{cnt}; |
599 |
} |
} |
600 |
|
|
601 |
sub NextMFN { |
=head2 unpack_cnt |
602 |
|
|
603 |
|
Unpack one of two 26 bytes fixed length record in C<.CNT> file. |
604 |
|
|
605 |
|
Here is definition of record: |
606 |
|
|
607 |
|
off key description size |
608 |
|
0: IDTYPE BTree type s |
609 |
|
2: ORDN Nodes Order s |
610 |
|
4: ORDF Leafs Order s |
611 |
|
6: N Number of Memory buffers for nodes s |
612 |
|
8: K Number of buffers for first level index s |
613 |
|
10: LIV Current number of Index Levels s |
614 |
|
12: POSRX Pointer to Root Record in N0x l |
615 |
|
16: NMAXPOS Next Available position in N0x l |
616 |
|
20: FMAXPOS Next available position in L0x l |
617 |
|
24: ABNORMAL Formal BTree normality indicator s |
618 |
|
length: 26 bytes |
619 |
|
|
620 |
|
This will fill C<$self> object under C<cnt> with hash. It's used by C<read_cnt>. |
621 |
|
|
622 |
|
=cut |
623 |
|
|
624 |
|
sub unpack_cnt { |
625 |
my $self = shift; |
my $self = shift; |
626 |
return $self->{NXTMFN}; |
|
627 |
|
my @flds = qw(ORDN ORDF N K LIV POSRX NMAXPOS FMAXPOS ABNORMAL); |
628 |
|
|
629 |
|
my $buff = shift || return; |
630 |
|
my @arr = unpack("vvvvvvVVVv", $buff); |
631 |
|
|
632 |
|
print STDERR "unpack_cnt: ",join(" ",@arr),"\n" if ($self->{'debug'}); |
633 |
|
|
634 |
|
my $IDTYPE = shift @arr; |
635 |
|
foreach (@flds) { |
636 |
|
$self->{cnt}->{$IDTYPE}->{$_} = abs(shift @arr); |
637 |
|
} |
638 |
} |
} |
639 |
|
|
640 |
1; |
1; |
641 |
|
|
642 |
=head1 BUGS |
=head1 BUGS |
643 |
|
|
644 |
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 |
645 |
|
some variations in input databases which has been tested with this module. |
646 |
|
When I was in doubt, I assumed that OpenIsis's implementation was right |
647 |
|
(except for obvious bugs). |
648 |
|
|
649 |
|
However, every effort has been made to test this module with as much |
650 |
|
databases (and programs that create them) as possible. |
651 |
|
|
652 |
|
I would be very greatful for success or failure reports about usage of this |
653 |
|
module with databases from programs other than WinIsis and IsisMarc. I had |
654 |
|
tested this against ouput of one C<isis.dll>-based application, but I don't |
655 |
|
know any details about it's version. |
656 |
|
|
657 |
|
=head1 VERSIONS |
658 |
|
|
659 |
|
You can find version dependencies documented here |
660 |
|
|
661 |
|
=over 8 |
662 |
|
|
663 |
|
=item 0.20 |
664 |
|
|
665 |
|
Added C<< $isis->mfn >> and support for repeatable subfields |
666 |
|
|
667 |
|
=back |
668 |
|
|
669 |
=head1 AUTHOR |
=head1 AUTHOR |
670 |
|
|
673 |
dpavlin@rot13.org |
dpavlin@rot13.org |
674 |
http://www.rot13.org/~dpavlin/ |
http://www.rot13.org/~dpavlin/ |
675 |
|
|
676 |
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 |
677 |
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. |
678 |
|
|
679 |
=head1 COPYRIGHT |
=head1 COPYRIGHT |
680 |
|
|
687 |
|
|
688 |
=head1 SEE ALSO |
=head1 SEE ALSO |
689 |
|
|
690 |
L<http://www.openisis.org|OpenIsis>, perl(1). |
OpenIsis web site L<http://www.openisis.org> |
691 |
|
|
692 |
|
perl4lib site L<http://perl4lib.perl.org> |
693 |
|
|