/[webpac2]/trunk/lib/WebPAC/Input.pm
This is repository of my old source code which isn't updated any more. Go to git.rot13.org for current projects!
ViewVC logotype

Diff of /trunk/lib/WebPAC/Input.pm

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 285 by dpavlin, Sun Dec 18 21:06:39 2005 UTC revision 292 by dpavlin, Sun Dec 18 23:34:30 2005 UTC
# Line 11  use Text::Iconv; Line 11  use Text::Iconv;
11    
12  =head1 NAME  =head1 NAME
13    
14  WebPAC::Input - core module for input file format  WebPAC::Input - read different file formats into WebPAC
15    
16  =head1 VERSION  =head1 VERSION
17    
18  Version 0.02  Version 0.03
19    
20  =cut  =cut
21    
22  our $VERSION = '0.02';  our $VERSION = '0.03';
23    
24  =head1 SYNOPSIS  =head1 SYNOPSIS
25    
26  This module is used as base class for all database specific modules  This module implements input as database which have fixed and known
27  (basically, files which have one handle, fixed size while indexing and some  I<size> while indexing and single unique numeric identifier for database
28  kind of numeric idefinirier which goes from 1 to filesize).  position ranging from 1 to I<size>.
29    
30    Simply, something that is indexed by unmber from 1 .. I<size>.
31    
32    Examples of such databases are CDS/ISIS files, MARC files, lines in
33    text file, and so on.
34    
35    Specific file formats are implemented using low-level interface modules,
36    located in C<WebPAC::Input::*> namespace which export C<open_db>,
37    C<fetch_rec> and optional C<init> functions.
38    
39  Perhaps a little code snippet.  Perhaps a little code snippet.
40    
41      use WebPAC::Input;      use WebPAC::Input;
42    
43      my $db = WebPAC::Input->new(      my $db = WebPAC::Input->new(
44          format => 'NULL',          module => 'WebPAC::Input::ISIS',
45          config => $config,                  config => $config,
46          lookup => $lookup_obj,                  lookup => $lookup_obj,
47          low_mem => 1,                  low_mem => 1,
48      );      );
49    
50      $db->open('/path/to/database');      $db->open('/path/to/database');
51      print "database size: ",$db->size,"\n";      print "database size: ",$db->size,"\n";
52      while (my $row = $db->fetch) {      while (my $rec = $db->fetch) {
         ...  
53      }      }
54    
55    
56    
57  =head1 FUNCTIONS  =head1 FUNCTIONS
58    
59  =head2 new  =head2 new
# Line 51  Perhaps a little code snippet. Line 61  Perhaps a little code snippet.
61  Create new input database object.  Create new input database object.
62    
63    my $db = new WebPAC::Input(    my $db = new WebPAC::Input(
64          format => 'NULL'          module => 'WebPAC::Input::MARC',
65          code_page => 'ISO-8859-2',          code_page => 'ISO-8859-2',
66          low_mem => 1,          low_mem => 1,
67    );    );
68    
69    C<module> is low-level file format module. See L<WebPAC::Input::Isis> and
70    L<WebPAC::Input::MARC>.
71    
72  Optional parametar C<code_page> specify application code page (which will be  Optional parametar C<code_page> specify application code page (which will be
73  used internally). This should probably be your terminal encoding, and by  used internally). This should probably be your terminal encoding, and by
74  default, it C<ISO-8859-2>.  default, it C<ISO-8859-2>.
# Line 74  sub new { Line 87  sub new {
87    
88          my $log = $self->_get_logger;          my $log = $self->_get_logger;
89    
90            $log->logconfess("specify low-level file format module") unless ($self->{module});
91            my $module = $self->{module};
92            $module =~ s#::#/#g;
93            $module .= '.pm';
94            $log->debug("require low-level module $self->{module} from $module");
95    
96            require $module;
97            #eval $self->{module} .'->import';
98    
99          # check if required subclasses are implemented          # check if required subclasses are implemented
100          foreach my $subclass (qw/open_db fetch_rec/) {          foreach my $subclass (qw/open_db fetch_rec init/) {
101                  $log->logdie("missing implementation of $subclass") unless ($self->SUPER::can($subclass));                  my $n = $self->{module} . '::' . $subclass;
102                    if (! defined &{ $n }) {
103                            my $missing = "missing $subclass in $self->{module}";
104                            $log->logwarn($missing);
105                            $self->{$subclass} = sub { warn "$missing\n" };
106                    } else {
107                            $self->{$subclass} = \&{ $n };
108                    }
109          }          }
110    
111          if ($self->can('init')) {          if ($self->{init}) {
112                  $log->debug("calling init");                  $log->debug("calling init");
113                  $self->init(@_);                  $self->{init}->($self, @_);
114          }          }
115    
116          $self->{'code_page'} ||= 'ISO-8859-2';          $self->{'code_page'} ||= 'ISO-8859-2';
# Line 119  sub new { Line 148  sub new {
148    
149  This function will read whole database in memory and produce lookups.  This function will read whole database in memory and produce lookups.
150    
151   $isis->open(   $input->open(
152          path => '/path/to/database/file',          path => '/path/to/database/file',
153          code_page => '852',          code_page => '852',
154          limit_mfn => 500,          limit => 500,
155          start_mfn => 6000,          offset => 6000,
156          lookup => $lookup_obj,          lookup => $lookup_obj,
157   );   );
158    
159  By default, C<code_page> is assumed to be C<852>.  By default, C<code_page> is assumed to be C<852>.
160    
161  If optional parametar C<start_mfn> is set, this will be first MFN to read  C<offset> is optional parametar to position at some offset before reading from database.
 from database (so you can skip beginning of your database if you need to).  
162    
163  If optional parametar C<limit_mfn> is set, it will read just 500 records  C<limit> is optional parametar to read just C<limit> records from database
 from database in example above.  
164    
165  Returns size of database, regardless of C<start_mfn> and C<limit_mfn>  Returns size of database, regardless of C<offset> and C<limit>
166  parametars, see also C<$isis->size>.  parametars, see also C<size>.
167    
168  =cut  =cut
169    
# Line 150  sub open { Line 177  sub open {
177          my $code_page = $arg->{'code_page'} || '852';          my $code_page = $arg->{'code_page'} || '852';
178    
179          # store data in object          # store data in object
180          $self->{'code_page'} = $code_page;          $self->{'input_code_page'} = $code_page;
181          foreach my $v (qw/path start_mfn limit_mfn/) {          foreach my $v (qw/path offset limit/) {
182                  $self->{$v} = $arg->{$v} if ($arg->{$v});                  $self->{$v} = $arg->{$v} if ($arg->{$v});
183          }          }
184    
185          # create Text::Iconv object          # create Text::Iconv object
186          $self->{iconv} = Text::Iconv->new($code_page,$self->{'code_page'});          $self->{iconv} = Text::Iconv->new($code_page,$self->{'code_page'});
187    
188          my ($db, $size) = $self->open_db(          my ($db, $size) = $self->{open_db}->( $self,
189                  path => $arg->{path},                  path => $arg->{path},
190          );          );
191    
# Line 172  sub open { Line 199  sub open {
199                  return;                  return;
200          }          }
201    
202          my $startmfn = 1;          my $offset = 1;
203          my $maxmfn = $size;          my $limit = $size;
204    
205          if (my $s = $self->{start_mfn}) {          if (my $s = $self->{offset}) {
206                  $log->info("skipping to MFN $s");                  $log->info("skipping to MFN $s");
207                  $startmfn = $s;                  $offset = $s;
208          } else {          } else {
209                  $self->{start_mfn} = $startmfn;                  $self->{offset} = $offset;
210          }          }
211    
212          if ($self->{limit_mfn}) {          if ($self->{limit}) {
213                  $log->info("limiting to ",$self->{limit_mfn}," records");                  $log->info("limiting to ",$self->{limit}," records");
214                  $maxmfn = $startmfn + $self->{limit_mfn} - 1;                  $limit = $offset + $self->{limit} - 1;
215                  $maxmfn = $size if ($maxmfn > $size);                  $limit = $size if ($limit > $size);
216          }          }
217    
218          # store size for later          # store size for later
219          $self->{size} = ($maxmfn - $startmfn) ? ($maxmfn - $startmfn + 1) : 0;          $self->{size} = ($limit - $offset) ? ($limit - $offset + 1) : 0;
220    
221          $log->info("processing $self->{size} records in $code_page, convert to $self->{code_page}");          $log->info("processing $self->{size} records in $code_page, convert to $self->{code_page}");
222    
223          # read database          # read database
224          for (my $mfn = $startmfn; $mfn <= $maxmfn; $mfn++) {          for (my $pos = $offset; $pos <= $limit; $pos++) {
225    
226                  $log->debug("mfn: $mfn\n");                  $log->debug("position: $pos\n");
227    
228                  my $rec = $self->fetch_rec( $db, $mfn );                  my $rec = $self->{fetch_rec}->($self, $db, $pos );
229    
230                  if (! $rec) {                  if (! $rec) {
231                          $log->warn("record $mfn empty? skipping...");                          $log->warn("record $pos empty? skipping...");
232                          next;                          next;
233                  }                  }
234    
235                  # store                  # store
236                  if ($self->{'low_mem'}) {                  if ($self->{low_mem}) {
237                          $self->{'db'}->put($mfn, $rec);                          $self->{db}->put($pos, $rec);
238                  } else {                  } else {
239                          $self->{'data'}->{$mfn} = $rec;                          $self->{data}->{$pos} = $rec;
240                  }                  }
241    
242                  # create lookup                  # create lookup
243                  $self->{'lookup'}->add( $rec ) if ($rec && $self->{'lookup'});                  $self->{'lookup'}->add( $rec ) if ($rec && $self->{'lookup'});
244    
245                  $self->progress_bar($mfn,$maxmfn);                  $self->progress_bar($pos,$limit);
246    
247          }          }
248    
249          $self->{'current_mfn'} = -1;          $self->{pos} = -1;
250          $self->{'last_pcnt'} = 0;          $self->{last_pcnt} = 0;
   
         $log->debug("max mfn: $maxmfn");  
251    
252          # store max mfn and return it.          # store max mfn and return it.
253          $self->{'max_mfn'} = $maxmfn;          $self->{max_pos} = $limit;
254            $log->debug("max_pos: $limit");
255    
256          return $size;          return $size;
257  }  }
# Line 246  sub fetch { Line 272  sub fetch {
272    
273          my $log = $self->_get_logger();          my $log = $self->_get_logger();
274    
275          $log->logconfess("it seems that you didn't load database!") unless ($self->{'current_mfn'});          $log->logconfess("it seems that you didn't load database!") unless ($self->{pos});
276    
277          if ($self->{'current_mfn'} == -1) {          if ($self->{pos} == -1) {
278                  $self->{'current_mfn'} = $self->{'start_mfn'};                  $self->{pos} = $self->{offset};
279          } else {          } else {
280                  $self->{'current_mfn'}++;                  $self->{pos}++;
281          }          }
282    
283          my $mfn = $self->{'current_mfn'};          my $mfn = $self->{pos};
284    
285          if ($mfn > $self->{'max_mfn'}) {          if ($mfn > $self->{max_pos}) {
286                  $self->{'current_mfn'} = $self->{'max_mfn'};                  $self->{pos} = $self->{max_pos};
287                  $log->debug("at EOF");                  $log->debug("at EOF");
288                  return;                  return;
289          }          }
290    
291          $self->progress_bar($mfn,$self->{'max_mfn'});          $self->progress_bar($mfn,$self->{max_pos});
292    
293          my $rec;          my $rec;
294    
295          if ($self->{'low_mem'}) {          if ($self->{low_mem}) {
296                  $rec = $self->{'db'}->get($mfn);                  $rec = $self->{db}->get($mfn);
297          } else {          } else {
298                  $rec = $self->{'data'}->{$mfn};                  $rec = $self->{data}->{$mfn};
299          }          }
300    
301          $rec ||= 0E0;          $rec ||= 0E0;
# Line 287  First record in database has position 1. Line 313  First record in database has position 1.
313    
314  sub pos {  sub pos {
315          my $self = shift;          my $self = shift;
316          return $self->{'current_mfn'};          return $self->{pos};
317  }  }
318    
319    
# Line 301  Result from this function can be used to Line 327  Result from this function can be used to
327    
328   foreach my $mfn ( 1 ... $isis->size ) { ... }   foreach my $mfn ( 1 ... $isis->size ) { ... }
329    
330  because it takes into account C<start_mfn> and C<limit_mfn>.  because it takes into account C<offset> and C<limit>.
331    
332  =cut  =cut
333    
334  sub size {  sub size {
335          my $self = shift;          my $self = shift;
336          return $self->{'size'};          return $self->{size};
337  }  }
338    
339  =head2 seek  =head2 seek
# Line 329  sub seek { Line 355  sub seek {
355          if ($pos < 1) {          if ($pos < 1) {
356                  $log->warn("seek before first record");                  $log->warn("seek before first record");
357                  $pos = 1;                  $pos = 1;
358          } elsif ($pos > $self->{'max_mfn'}) {          } elsif ($pos > $self->{max_pos}) {
359                  $log->warn("seek beyond last record");                  $log->warn("seek beyond last record");
360                  $pos = $self->{'max_mfn'};                  $pos = $self->{max_pos};
361          }          }
362    
363          return $self->{'current_mfn'} = (($pos - 1) || -1);          return $self->{pos} = (($pos - 1) || -1);
364  }  }
365    
366    
Legend:
Removed from v.285  
changed lines
  Added in v.292
  ViewVC Help
Powered by ViewVC 1.1.26