| 1 |
package IsisDB; |
package Biblio::Isis; |
| 2 |
use strict; |
use strict; |
| 3 |
|
|
| 4 |
use Carp; |
use Carp; |
| 5 |
|
use File::Glob qw(:globally :nocase); |
| 6 |
|
|
| 7 |
use Data::Dumper; |
use Data::Dumper; |
| 8 |
|
|
| 9 |
BEGIN { |
BEGIN { |
| 10 |
use Exporter (); |
use Exporter (); |
| 11 |
use vars qw ($VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS); |
use vars qw ($VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS); |
| 12 |
$VERSION = 0.01; |
$VERSION = 0.12; |
| 13 |
@ISA = qw (Exporter); |
@ISA = qw (Exporter); |
| 14 |
#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 |
| 15 |
@EXPORT = qw (); |
@EXPORT = qw (); |
| 20 |
|
|
| 21 |
=head1 NAME |
=head1 NAME |
| 22 |
|
|
| 23 |
IsisDB - Read CDS/ISIS database |
Biblio::Isis - Read CDS/ISIS, WinISIS and IsisMarc database |
| 24 |
|
|
| 25 |
=head1 SYNOPSIS |
=head1 SYNOPSIS |
| 26 |
|
|
| 27 |
use IsisDB |
use Biblio::Isis; |
| 28 |
my $isis = new IsisDB( |
|
| 29 |
|
my $isis = new Biblio::Isis( |
| 30 |
isisdb => './cds/cds', |
isisdb => './cds/cds', |
| 31 |
); |
); |
| 32 |
|
|
| 33 |
|
for(my $mfn = 1; $mfn <= $isis->count; $mfn++) { |
| 34 |
|
print $isis->to_ascii($mfn),"\n"; |
| 35 |
|
} |
| 36 |
|
|
| 37 |
=head1 DESCRIPTION |
=head1 DESCRIPTION |
| 38 |
|
|
| 39 |
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 |
| 40 |
It can be used as perl-only alternative to OpenIsis module. |
IsisMarc. It can be used as perl-only alternative to OpenIsis module which |
| 41 |
|
seems to depriciate it's old C<XS> bindings for perl. |
| 42 |
|
|
| 43 |
|
It can create hash values from data in ISIS database (using C<to_hash>), |
| 44 |
|
ASCII dump (using C<to_ascii>) or just hash with field names and packed |
| 45 |
|
values (like C<^asomething^belse>). |
| 46 |
|
|
| 47 |
|
Unique feature of this module is ability to C<include_deleted> records. |
| 48 |
|
It will also skip zero sized fields (OpenIsis has a bug in XS bindings, so |
| 49 |
|
fields which are zero sized will be filled with random junk from memory). |
| 50 |
|
|
| 51 |
|
It also has support for identifiers (only if ISIS database is created by |
| 52 |
|
IsisMarc), see C<to_hash>. |
| 53 |
|
|
| 54 |
|
This module will always be slower than OpenIsis module which use C |
| 55 |
|
library. However, since it's written in perl, it's platform independent (so |
| 56 |
|
you don't need C compiler), and can be easily modified. I hope that it |
| 57 |
|
creates data structures which are easier to use than ones created by |
| 58 |
|
OpenIsis, so reduced time in other parts of the code should compensate for |
| 59 |
|
slower performance of this module (speed of reading ISIS database is |
| 60 |
|
rarely an issue). |
| 61 |
|
|
| 62 |
=head1 METHODS |
=head1 METHODS |
| 63 |
|
|
| 77 |
# some binary reads |
# some binary reads |
| 78 |
# |
# |
| 79 |
|
|
|
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); |
|
|
} |
|
|
|
|
| 80 |
=head2 new |
=head2 new |
| 81 |
|
|
| 82 |
Open CDS/ISIS database |
Open ISIS database |
| 83 |
|
|
| 84 |
my $isis = new IsisDB( |
my $isis = new Biblio::Isis( |
| 85 |
isisdb => './cds/cds', |
isisdb => './cds/cds', |
| 86 |
read_fdt => 1, |
read_fdt => 1, |
| 87 |
|
include_deleted => 1, |
| 88 |
|
hash_filter => sub { |
| 89 |
|
my $v = shift; |
| 90 |
|
$v =~ s#foo#bar#g; |
| 91 |
|
}, |
| 92 |
debug => 1, |
debug => 1, |
| 93 |
); |
); |
| 94 |
|
|
| 95 |
|
Options are described below: |
| 96 |
|
|
| 97 |
|
=over 5 |
| 98 |
|
|
| 99 |
=item isisdb |
=item isisdb |
| 100 |
|
|
| 101 |
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 |
| 102 |
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 |
| 103 |
|
C<read_fdt> option) files. |
| 104 |
|
|
| 105 |
|
In this example it uses C<./cds/cds.MST> and related files. |
| 106 |
|
|
| 107 |
=item read_fdt |
=item read_fdt |
| 108 |
|
|
| 109 |
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 |
| 110 |
by default. |
by default. |
| 111 |
|
|
| 112 |
|
=item include_deleted |
| 113 |
|
|
| 114 |
|
Don't skip logically deleted records in ISIS. |
| 115 |
|
|
| 116 |
|
=item hash_filter |
| 117 |
|
|
| 118 |
|
Filter code ref which will be used before data is converted to hash. |
| 119 |
|
|
| 120 |
=item debug |
=item debug |
| 121 |
|
|
| 122 |
Dump a C<lot> of debugging output. |
Dump a B<lot> of debugging output. |
| 123 |
|
|
| 124 |
|
=back |
| 125 |
|
|
| 126 |
=cut |
=cut |
| 127 |
|
|
| 130 |
my $self = {}; |
my $self = {}; |
| 131 |
bless($self, $class); |
bless($self, $class); |
| 132 |
|
|
| 133 |
$self->{isisdb} = {@_}->{isisdb} || croak "new needs database name as argument!"; |
croak "new needs database name (isisdb) as argument!" unless ({@_}->{isisdb}); |
| 134 |
|
|
| 135 |
|
foreach my $v (qw{isisdb debug include_deleted hash_filter}) { |
| 136 |
|
$self->{$v} = {@_}->{$v}; |
| 137 |
|
} |
| 138 |
|
|
| 139 |
|
my @isis_files = grep(/\.(FDT|MST|XRF|CNT)$/i,glob($self->{isisdb}."*")); |
| 140 |
|
|
| 141 |
$self->{debug} = {@_}->{debug} || 1; # XXX remove debug always! |
foreach my $f (@isis_files) { |
| 142 |
|
my $ext = $1 if ($f =~ m/\.(\w\w\w)$/); |
| 143 |
|
$self->{lc($ext)."_file"} = $f; |
| 144 |
|
} |
| 145 |
|
|
| 146 |
|
my @must_exist = qw(mst xrf); |
| 147 |
|
push @must_exist, "fdt" if ($self->{read_fdt}); |
| 148 |
|
|
| 149 |
|
foreach my $ext (@must_exist) { |
| 150 |
|
unless ($self->{$ext."_file"}) { |
| 151 |
|
carp "missing ",uc($ext)," file in ",$self->{isisdb}; |
| 152 |
|
return; |
| 153 |
|
} |
| 154 |
|
} |
| 155 |
|
|
| 156 |
|
print STDERR "## using files: ",join(" ",@isis_files),"\n" if ($self->{debug}); |
| 157 |
|
|
| 158 |
# 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! |
| 159 |
if ({@_}->{read_fdt} && -e $self->{isisdb}.".FDT") { |
if ($self->{read_fdt} && -e $self->{fdt_file}) { |
| 160 |
|
|
| 161 |
# read the $db.FDT file for tags |
# read the $db.FDT file for tags |
| 162 |
my $fieldzone=0; |
my $fieldzone=0; |
| 163 |
|
|
| 164 |
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}': $!"; |
| 165 |
|
binmode($fileFDT); |
| 166 |
|
|
| 167 |
while (<fileFDT>) { |
while (<$fileFDT>) { |
| 168 |
chomp; |
chomp; |
| 169 |
if ($fieldzone) { |
if ($fieldzone) { |
| 170 |
my $name=substr($_,0,30); |
my $name=substr($_,0,30); |
| 181 |
} |
} |
| 182 |
} |
} |
| 183 |
|
|
| 184 |
close(fileFDT); |
close($fileFDT); |
| 185 |
} |
} |
| 186 |
|
|
| 187 |
# Get the Maximum MFN from $db.MST |
# Get the Maximum MFN from $db.MST |
| 188 |
|
|
| 189 |
open(fileMST,$self->{isisdb}.".MST") || croak "can't read '$self->{isisdb}.MST': $!"; |
open($self->{'fileMST'}, $self->{mst_file}) || croak "can't open '$self->{mst_file}': $!"; |
| 190 |
|
binmode($self->{'fileMST'}); |
| 191 |
|
|
| 192 |
# MST format: (* = 32 bit signed) |
# MST format: (* = 32 bit signed) |
| 193 |
# CTLMFN* always 0 |
# CTLMFN* always 0 |
| 195 |
# NXTMFB* last block allocated to master file |
# NXTMFB* last block allocated to master file |
| 196 |
# NXTMFP offset to next available position in last block |
# NXTMFP offset to next available position in last block |
| 197 |
# MFTYPE always 0 for user db file (1 for system) |
# MFTYPE always 0 for user db file (1 for system) |
| 198 |
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"; |
|
| 199 |
|
|
| 200 |
close(fileMST); |
my $buff; |
| 201 |
|
|
| 202 |
# Get the index information from $db.CNT |
read($self->{'fileMST'}, $buff, 4) || croak "can't read NXTMFN from MST: $!"; |
| 203 |
|
$self->{'NXTMFN'}=unpack("V",$buff) || croak "NXTNFN is zero"; |
|
open(fileCNT, $self->{isisdb}.".CNT") || croak "can't read '$self->{isisdb}.CNT': $!"; |
|
| 204 |
|
|
| 205 |
# There is two 26 Bytes fixed lenght records |
print STDERR Dumper($self),"\n" if ($self->{debug}); |
| 206 |
|
|
| 207 |
# 0: IDTYPE BTree type 16 |
# open files for later |
| 208 |
# 2: ORDN Nodes Order 16 |
open($self->{'fileXRF'}, $self->{xrf_file}) || croak "can't open '$self->{xrf_file}': $!"; |
| 209 |
# 4: ORDF Leafs Order 16 |
binmode($self->{'fileXRF'}); |
|
# 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); |
|
|
} |
|
|
} |
|
| 210 |
|
|
| 211 |
my $buff; |
$self ? return $self : return undef; |
| 212 |
read(fileCNT, $buff, 26); |
} |
|
$self->unpack_cnt($buff); |
|
| 213 |
|
|
| 214 |
read(fileCNT, $buff, 26); |
=head2 count |
|
$self->unpack_cnt($buff); |
|
| 215 |
|
|
| 216 |
|
Return number of records in database |
| 217 |
|
|
| 218 |
close(fileCNT); |
print $isis->count; |
| 219 |
|
|
| 220 |
print Dumper($self) if ($self->{debug}); |
=cut |
| 221 |
|
|
| 222 |
$self ? return $self : return undef; |
sub count { |
| 223 |
|
my $self = shift; |
| 224 |
|
return $self->{'NXTMFN'} - 1; |
| 225 |
} |
} |
| 226 |
|
|
| 227 |
|
=head2 fetch |
| 228 |
|
|
| 229 |
|
Read record with selected MFN |
| 230 |
|
|
| 231 |
# Get a record from the MFN |
my $rec = $isis->fetch(55); |
| 232 |
# Return the number of fields in the record. |
|
| 233 |
# Return -1 if the record is marked for deletion |
Returns hash with keys which are field names and values are unpacked values |
| 234 |
# The record is then extracted with call to GETs |
for that field like this: |
| 235 |
|
|
| 236 |
|
$rec = { |
| 237 |
|
'210' => [ '^aNew York^cNew York University press^dcop. 1988' ], |
| 238 |
|
'990' => [ '2140', '88', 'HAY' ], |
| 239 |
|
}; |
| 240 |
|
|
| 241 |
|
=cut |
| 242 |
|
|
| 243 |
sub GetMFN { |
sub fetch { |
| 244 |
my $self = shift; |
my $self = shift; |
| 245 |
|
|
| 246 |
my $mfn = shift || croak "GetMFN needs MFN as argument!"; |
my $mfn = shift || croak "fetch needs MFN as argument!"; |
| 247 |
|
|
| 248 |
print "GetMFN: $mfn\n" if ($self->{debug}); |
# is mfn allready in memory? |
| 249 |
|
my $old_mfn = $self->{'current_mfn'} || -1; |
| 250 |
|
return $self->{record} if ($mfn == $old_mfn); |
| 251 |
|
|
| 252 |
open(fileXRF, $self->{isisdb}.".XRF") || croak "can't open '$self->{isisdb}.XRF': $!"; |
print STDERR "## fetch: $mfn\n" if ($self->{debug}); |
| 253 |
|
|
| 254 |
# XXX check this? |
# XXX check this? |
| 255 |
my $mfnpos=($mfn+int(($mfn-1)/127))*4; |
my $mfnpos=($mfn+int(($mfn-1)/127))*4; |
| 256 |
|
|
| 257 |
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 ($self->{debug}); |
| 258 |
seek(fileXRF,$mfnpos,0); |
seek($self->{'fileXRF'},$mfnpos,0); |
| 259 |
|
|
| 260 |
# read XRFMFB abd XRFMFP |
my $buff; |
|
my $pointer=$self->Read32(\*fileXRF); |
|
| 261 |
|
|
| 262 |
my $XRFMFB = int($pointer/2048); |
# delete old record |
| 263 |
my $XRFMFP = $pointer - ($XRFMFB*2048); |
delete $self->{record}; |
| 264 |
|
|
| 265 |
print "XRFMFB: $XRFMFB XRFMFP: $XRFMFP\n" if ($self->{debug}); |
# read XRFMFB abd XRFMFP |
| 266 |
|
read($self->{'fileXRF'}, $buff, 4); |
| 267 |
|
my $pointer=unpack("V",$buff) || croak "pointer is null"; |
| 268 |
|
|
| 269 |
|
# check for logically deleted record |
| 270 |
|
if ($pointer & 0x80000000) { |
| 271 |
|
print STDERR "## record $mfn is logically deleted\n" if ($self->{debug}); |
| 272 |
|
$self->{deleted} = $mfn; |
| 273 |
|
|
| 274 |
# XXX fix this to be more readable!! |
return unless $self->{include_deleted}; |
|
# e.g. (XRFMFB - 1) * 512 + XRFMFP |
|
| 275 |
|
|
| 276 |
my $offset = $pointer; |
# abs |
| 277 |
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; |
|
| 278 |
} |
} |
|
my $offset4=($offset2*512)+$offset3; |
|
| 279 |
|
|
| 280 |
print "$offset - $offset2 - $offset3 - $offset4\n" if ($self->{debug}); |
my $XRFMFB = int($pointer/2048); |
| 281 |
|
my $XRFMFP = $pointer - ($XRFMFB*2048); |
| 282 |
|
|
| 283 |
close(fileXRF); |
# (XRFMFB - 1) * 512 + XRFMFP |
| 284 |
|
# why do i have to do XRFMFP % 1024 ? |
| 285 |
|
|
| 286 |
|
my $blk_off = (($XRFMFB - 1) * 512) + ($XRFMFP % 512); |
| 287 |
|
|
| 288 |
|
print STDERR "## pointer: $pointer XRFMFB: $XRFMFB XRFMFP: $XRFMFP offset: $blk_off\n" if ($self->{'debug'}); |
| 289 |
|
|
| 290 |
# Get Record Information |
# Get Record Information |
| 291 |
|
|
| 292 |
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: $!"; |
| 293 |
|
|
| 294 |
seek(fileMST,$offset4,0); |
read($self->{'fileMST'}, $buff, 4) || croak "can't read 4 bytes at offset $blk_off from MST file: $!"; |
| 295 |
|
my $value=unpack("V",$buff); |
| 296 |
|
|
| 297 |
my $value=$self->Read32(\*fileMST); |
print STDERR "## offset for rowid $value is $blk_off (blk $XRFMFB off $XRFMFP)\n" if ($self->{debug}); |
| 298 |
|
|
| 299 |
if ($value!=$mfn) { |
if ($value!=$mfn) { |
| 300 |
print ("Error: The MFN:".$mfn." is not found in MST(".$value.")"); |
if ($value == 0) { |
| 301 |
return -1; # XXX deleted record? |
print STDERR "## record $mfn is physically deleted\n" if ($self->{debug}); |
| 302 |
|
$self->{deleted} = $mfn; |
| 303 |
|
return; |
| 304 |
|
} |
| 305 |
|
|
| 306 |
|
carp "Error: MFN ".$mfn." not found in MST file, found $value"; |
| 307 |
|
return; |
| 308 |
} |
} |
| 309 |
|
|
| 310 |
# $MFRL=$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); |
|
| 311 |
|
|
| 312 |
my $buff; |
my ($MFRL,$MFBWB,$MFBWP,$BASE,$NVF,$STATUS) = unpack("vVvvvv", $buff); |
| 313 |
read(fileMST, $buff, 14); |
|
| 314 |
|
print STDERR "## MFRL: $MFRL MFBWB: $MFBWB MFBWP: $MFBWP BASE: $BASE NVF: $NVF STATUS: $STATUS\n" if ($self->{debug}); |
| 315 |
|
|
| 316 |
my ($MFRL,$MFBWB,$MFBWP,$BASE,$NVF,$STATUS) = unpack("slssss", $buff); |
warn "MFRL $MFRL is not even number" unless ($MFRL % 2 == 0); |
| 317 |
|
|
| 318 |
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); |
| 319 |
|
|
| 320 |
# Get Directory Format |
# Get Directory Format |
| 321 |
|
|
| 323 |
my @FieldLEN; |
my @FieldLEN; |
| 324 |
my @FieldTAG; |
my @FieldTAG; |
| 325 |
|
|
| 326 |
for (my $i = 0 ; $i < $NVF ; $i++) { |
read($self->{'fileMST'}, $buff, 6 * $NVF); |
| 327 |
|
|
| 328 |
|
my $rec_len = 0; |
| 329 |
|
|
| 330 |
# $TAG=$self->Read16($fileMST); |
for (my $i = 0 ; $i < $NVF ; $i++) { |
|
# $POS=$self->Read16($fileMST); |
|
|
# $LEN=$self->Read16($fileMST); |
|
| 331 |
|
|
| 332 |
read(fileMST, $buff, 6); |
my ($TAG,$POS,$LEN) = unpack("vvv", substr($buff,$i * 6, 6)); |
|
my ($TAG,$POS,$LEN) = unpack("sss", $buff); |
|
| 333 |
|
|
| 334 |
print "TAG: $TAG POS: $POS LEN: $LEN\n" if ($self->{debug}); |
print STDERR "## TAG: $TAG POS: $POS LEN: $LEN\n" if ($self->{debug}); |
| 335 |
|
|
| 336 |
# 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. |
| 337 |
# |
# |
| 346 |
push @FieldTAG,$TAG; |
push @FieldTAG,$TAG; |
| 347 |
push @FieldPOS,$POS; |
push @FieldPOS,$POS; |
| 348 |
push @FieldLEN,$LEN; |
push @FieldLEN,$LEN; |
| 349 |
|
|
| 350 |
|
$rec_len += $LEN; |
| 351 |
} |
} |
| 352 |
|
|
| 353 |
# Get Variable Fields |
# Get Variable Fields |
| 354 |
|
|
| 355 |
|
read($self->{'fileMST'},$buff,$rec_len); |
| 356 |
|
|
| 357 |
|
print STDERR "## rec_len: $rec_len poc: ",tell($self->{'fileMST'})."\n" if ($self->{debug}); |
| 358 |
|
|
| 359 |
for (my $i = 0 ; $i < $NVF ; $i++) { |
for (my $i = 0 ; $i < $NVF ; $i++) { |
| 360 |
my $rec; |
# skip zero-sized fields |
| 361 |
read(fileMST,$rec,$FieldLEN[$i]); |
next if ($FieldLEN[$i] == 0); |
|
$self->{record}->{$FieldTAG[$i]} = $rec; |
|
|
} |
|
|
close(fileMST); |
|
| 362 |
|
|
| 363 |
# The record is marked for deletion |
push @{$self->{record}->{$FieldTAG[$i]}}, substr($buff,$FieldPOS[$i],$FieldLEN[$i]); |
|
if ($STATUS==1) { |
|
|
return -1; |
|
| 364 |
} |
} |
| 365 |
|
|
| 366 |
print Dumper($self) if ($self->{debug}); |
$self->{'current_mfn'} = $mfn; |
| 367 |
|
|
| 368 |
|
print STDERR Dumper($self),"\n" if ($self->{debug}); |
| 369 |
|
|
| 370 |
return $NVF; |
return $self->{'record'}; |
| 371 |
} |
} |
| 372 |
|
|
| 373 |
=begin php |
=head2 to_ascii |
| 374 |
|
|
| 375 |
# Load the dictionary from the $db.L0x files. |
Returns ASCII output of record with specified 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,$offset); |
|
| 376 |
|
|
| 377 |
} While (!feof($fileL01)); |
print $isis->to_ascii(42); |
| 378 |
|
|
| 379 |
fclose($fileL01); |
This outputs something like this: |
|
} |
|
| 380 |
|
|
| 381 |
# self function search through the tree and returns an array of pointers to IFP |
210 ^aNew York^cNew York University press^dcop. 1988 |
| 382 |
# The function must be recursive |
990 2140 |
| 383 |
|
990 88 |
| 384 |
|
990 HAY |
| 385 |
|
|
| 386 |
sub SearchTree($search,$fileNB,$PUNT) |
If C<read_fdt> is specified when calling C<new> it will display field names |
| 387 |
{ |
from C<.FDT> file instead of numeric tags. |
|
$offset=(($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 $OLDPUNT; |
|
|
} |
|
| 388 |
|
|
| 389 |
# Search ISIS for record containing search |
=cut |
|
# Return a sorted array of MFN |
|
| 390 |
|
|
| 391 |
sub Search($search) |
sub to_ascii { |
| 392 |
{ |
my $self = shift; |
| 393 |
|
|
| 394 |
$search=strtoupper($search); |
my $mfn = shift || croak "need MFN"; |
|
#print "Searching....".$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; |
|
|
} |
|
| 395 |
|
|
| 396 |
=cut |
my $rec = $self->fetch($mfn); |
| 397 |
|
|
| 398 |
# |
my $out = "0\t$mfn"; |
|
# XXX porting from php left-over: |
|
|
# |
|
|
# do I *REALLY* need those methods, or should I use |
|
|
# $self->{something} directly? |
|
|
# |
|
|
# Probably direct usage is better! |
|
|
# |
|
| 399 |
|
|
| 400 |
sub GetFieldName { |
foreach my $f (sort keys %{$rec}) { |
| 401 |
my $self = shift; |
my $fn = $self->tag_name($f); |
| 402 |
return $self->{FieldName}; |
$out .= "\n$fn\t".join("\n$fn\t",@{$self->{record}->{$f}}); |
| 403 |
} |
} |
| 404 |
|
|
| 405 |
sub GetTagName { |
$out .= "\n"; |
|
my $self = shift; |
|
|
return $self->{TagName}; |
|
|
} |
|
| 406 |
|
|
| 407 |
sub GetFieldTag { |
return $out; |
|
my $self = shift; |
|
|
return $self->{FieldTAG}; |
|
| 408 |
} |
} |
| 409 |
|
|
| 410 |
sub GetNextMFN { |
=head2 to_hash |
| 411 |
|
|
| 412 |
|
Read record with specified MFN and convert it to hash |
| 413 |
|
|
| 414 |
|
my $hash = $isis->to_hash($mfn); |
| 415 |
|
|
| 416 |
|
It has ability to convert characters (using C<hash_filter>) from ISIS |
| 417 |
|
database before creating structures enabling character re-mapping or quick |
| 418 |
|
fix-up of data. |
| 419 |
|
|
| 420 |
|
This function returns hash which is like this: |
| 421 |
|
|
| 422 |
|
$hash = { |
| 423 |
|
'210' => [ |
| 424 |
|
{ |
| 425 |
|
'c' => 'New York University press', |
| 426 |
|
'a' => 'New York', |
| 427 |
|
'd' => 'cop. 1988' |
| 428 |
|
} |
| 429 |
|
], |
| 430 |
|
'990' => [ |
| 431 |
|
'2140', |
| 432 |
|
'88', |
| 433 |
|
'HAY' |
| 434 |
|
], |
| 435 |
|
}; |
| 436 |
|
|
| 437 |
|
You can later use that hash to produce any output from ISIS data. |
| 438 |
|
|
| 439 |
|
If database is created using IsisMarc, it will also have to special fields |
| 440 |
|
which will be used for identifiers, C<i1> and C<i2> like this: |
| 441 |
|
|
| 442 |
|
'200' => [ |
| 443 |
|
{ |
| 444 |
|
'i1' => '1', |
| 445 |
|
'i2' => ' ' |
| 446 |
|
'a' => 'Goa', |
| 447 |
|
'f' => 'Valdo D\'Arienzo', |
| 448 |
|
'e' => 'tipografie e tipografi nel XVI secolo', |
| 449 |
|
} |
| 450 |
|
], |
| 451 |
|
|
| 452 |
|
This method will also create additional field C<000> with MFN. |
| 453 |
|
|
| 454 |
|
=cut |
| 455 |
|
|
| 456 |
|
sub to_hash { |
| 457 |
my $self = shift; |
my $self = shift; |
| 458 |
return $self->{NXTMFN}; |
|
| 459 |
|
my $mfn = shift || confess "need mfn!"; |
| 460 |
|
|
| 461 |
|
# init record to include MFN as field 000 |
| 462 |
|
my $rec = { '000' => [ $mfn ] }; |
| 463 |
|
|
| 464 |
|
my $row = $self->fetch($mfn); |
| 465 |
|
|
| 466 |
|
foreach my $k (keys %{$row}) { |
| 467 |
|
foreach my $l (@{$row->{$k}}) { |
| 468 |
|
|
| 469 |
|
# filter output |
| 470 |
|
$l = $self->{'hash_filter'}->($l) if ($self->{'hash_filter'}); |
| 471 |
|
|
| 472 |
|
my $val; |
| 473 |
|
|
| 474 |
|
# has identifiers? |
| 475 |
|
($val->{'i1'},$val->{'i2'}) = ($1,$2) if ($l =~ s/^([01 #])([01 #])\^/\^/); |
| 476 |
|
|
| 477 |
|
# has subfields? |
| 478 |
|
if ($l =~ m/\^/) { |
| 479 |
|
foreach my $t (split(/\^/,$l)) { |
| 480 |
|
next if (! $t); |
| 481 |
|
$val->{substr($t,0,1)} = substr($t,1); |
| 482 |
|
} |
| 483 |
|
} else { |
| 484 |
|
$val = $l; |
| 485 |
|
} |
| 486 |
|
|
| 487 |
|
push @{$rec->{$k}}, $val; |
| 488 |
|
} |
| 489 |
|
} |
| 490 |
|
|
| 491 |
|
return $rec; |
| 492 |
} |
} |
| 493 |
|
|
| 494 |
sub GetMFNArray { |
=head2 tag_name |
| 495 |
|
|
| 496 |
|
Return name of selected tag |
| 497 |
|
|
| 498 |
|
print $isis->tag_name('200'); |
| 499 |
|
|
| 500 |
|
=cut |
| 501 |
|
|
| 502 |
|
sub tag_name { |
| 503 |
my $self = shift; |
my $self = shift; |
| 504 |
return $self->{MFNArray}; |
my $tag = shift || return; |
| 505 |
|
return $self->{'TagName'}->{$tag} || $tag; |
| 506 |
} |
} |
|
=begin php |
|
| 507 |
|
|
|
sub Read32($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.")"; |
|
| 508 |
|
|
| 509 |
return $value; |
=head2 read_cnt |
|
} |
|
| 510 |
|
|
| 511 |
sub Read24($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)); |
|
| 512 |
|
|
| 513 |
$value=$B1+$B2*256+$B3*65536; |
print Dumper($isis->read_cnt); |
| 514 |
|
|
| 515 |
# print "(".$B1.",".$B2.",".$B3.":".$value.")"; |
This function is not used by module (C<.CNT> files are not required for this |
| 516 |
|
module to work), but it can be useful to examine your index (while debugging |
| 517 |
|
for example). |
| 518 |
|
|
| 519 |
return $value; |
=cut |
|
} |
|
| 520 |
|
|
| 521 |
sub Read16($fileNB) |
sub read_cnt { |
| 522 |
{ |
my $self = shift; |
|
$B1=ord(fread($fileNB,1)); |
|
|
$B2=ord(fread($fileNB,1)); |
|
| 523 |
|
|
| 524 |
$value=$B1+$B2*256; |
croak "missing CNT file in ",$self->{isisdb} unless ($self->{cnt_file}); |
|
# print "(".$B1.",".$B2.":".$value.")"; |
|
| 525 |
|
|
| 526 |
return $value; |
# Get the index information from $db.CNT |
| 527 |
} |
|
| 528 |
|
open(my $fileCNT, $self->{cnt_file}) || croak "can't read '$self->{cnt_file}': $!"; |
| 529 |
|
binmode($fileCNT); |
| 530 |
|
|
| 531 |
sub Read8($fileNB) |
my $buff; |
|
{ |
|
|
$B1=ord(fread($fileNB,1)); |
|
| 532 |
|
|
| 533 |
$value=$B1; |
read($fileCNT, $buff, 26) || croak "can't read first table from CNT: $!"; |
| 534 |
# print "(".$value.")"; |
$self->unpack_cnt($buff); |
| 535 |
|
|
| 536 |
return $value; |
read($fileCNT, $buff, 26) || croak "can't read second table from CNT: $!"; |
| 537 |
} |
$self->unpack_cnt($buff); |
| 538 |
|
|
| 539 |
sub Not8($value) |
close($fileCNT); |
| 540 |
{ |
|
| 541 |
$value=decbin($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 $value; |
|
|
} |
|
| 542 |
} |
} |
| 543 |
|
|
| 544 |
|
=head2 unpack_cnt |
| 545 |
|
|
| 546 |
|
Unpack one of two 26 bytes fixed length record in C<.CNT> file. |
| 547 |
|
|
| 548 |
|
Here is definition of record: |
| 549 |
|
|
| 550 |
|
off key description size |
| 551 |
|
0: IDTYPE BTree type s |
| 552 |
|
2: ORDN Nodes Order s |
| 553 |
|
4: ORDF Leafs Order s |
| 554 |
|
6: N Number of Memory buffers for nodes s |
| 555 |
|
8: K Number of buffers for first level index s |
| 556 |
|
10: LIV Current number of Index Levels s |
| 557 |
|
12: POSRX Pointer to Root Record in N0x l |
| 558 |
|
16: NMAXPOS Next Available position in N0x l |
| 559 |
|
20: FMAXPOS Next available position in L0x l |
| 560 |
|
24: ABNORMAL Formal BTree normality indicator s |
| 561 |
|
length: 26 bytes |
| 562 |
|
|
| 563 |
|
This will fill C<$self> object under C<cnt> with hash. It's used by C<read_cnt>. |
| 564 |
|
|
| 565 |
=cut |
=cut |
| 566 |
|
|
| 567 |
|
sub unpack_cnt { |
| 568 |
|
my $self = shift; |
| 569 |
|
|
| 570 |
|
my @flds = qw(ORDN ORDF N K LIV POSRX NMAXPOS FMAXPOS ABNORMAL); |
| 571 |
|
|
| 572 |
|
my $buff = shift || return; |
| 573 |
|
my @arr = unpack("vvvvvvVVVv", $buff); |
| 574 |
|
|
| 575 |
|
print STDERR "unpack_cnt: ",join(" ",@arr),"\n" if ($self->{'debug'}); |
| 576 |
|
|
| 577 |
|
my $IDTYPE = shift @arr; |
| 578 |
|
foreach (@flds) { |
| 579 |
|
$self->{cnt}->{$IDTYPE}->{$_} = abs(shift @arr); |
| 580 |
|
} |
| 581 |
|
} |
| 582 |
|
|
| 583 |
1; |
1; |
|
__END__ |
|
| 584 |
|
|
| 585 |
=head1 BUGS |
=head1 BUGS |
| 586 |
|
|
| 587 |
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 |
| 588 |
|
some variations in input databases which has been tested with this module. |
| 589 |
|
When I was in doubt, I assumed that OpenIsis's implementation was right |
| 590 |
|
(except for obvious bugs). |
| 591 |
|
|
| 592 |
|
However, every effort has been made to test this module with as much |
| 593 |
|
databases (and programs that create them) as possible. |
| 594 |
|
|
| 595 |
|
I would be very greatful for success or failure reports about usage of this |
| 596 |
|
module with databases from programs other than WinIsis and IsisMarc. I had |
| 597 |
|
tested this against ouput of one C<isis.dll>-based application, but I don't |
| 598 |
|
know any details about it's version. |
| 599 |
|
|
| 600 |
=head1 AUTHOR |
=head1 AUTHOR |
| 601 |
|
|
| 604 |
dpavlin@rot13.org |
dpavlin@rot13.org |
| 605 |
http://www.rot13.org/~dpavlin/ |
http://www.rot13.org/~dpavlin/ |
| 606 |
|
|
| 607 |
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 |
| 608 |
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. |
| 609 |
|
|
| 610 |
=head1 COPYRIGHT |
=head1 COPYRIGHT |
| 611 |
|
|
| 618 |
|
|
| 619 |
=head1 SEE ALSO |
=head1 SEE ALSO |
| 620 |
|
|
| 621 |
L<http://www.openisis.org|OpenIsis>, perl(1). |
OpenIsis web site L<http://www.openisis.org> |
| 622 |
|
|
| 623 |
|
perl4lib site L<http://perl4lib.perl.org> |
| 624 |
|
|
Parent Directory
| 
Revision Log
| 
Patch