7 PSPP-Perl - Perl extension to PSPP
15 PSPP-Perl provides an interface to the libraries used by pspp to read and
24 $PSPP::VERSION='@VERSION_FOR_PERL@';
26 XSLoader::load('PSPP', $PSPP::VERSION);
29 PSPP::onBoot($PSPP::VERSION);
33 =head1 PROGRAMMER'S INTERFACE
35 The subroutines in this package return zero or unref on error.
36 When errors occur, a string describing the error is written
44 use constant { SYSMIS => -(POSIX::DBL_MAX),
45 PERL_EPOCH => 12219379200 # Number of seconds between
57 =head2 PSPP::Dict::new
59 Creates a new dictionary. This returned dictionary will be empty.
60 Returns undef on failure.
62 =head3 set_documents ($string)
64 Sets the documents (comments) to C<string>.
66 =head3 add_document ($string)
68 Appends C<string> to the documents.
70 =head3 clear_documents ()
72 Removes all documents.
74 =head3 set_weight ($var)
76 Sets the weighting variable to C<var>.
83 my $self = pxs_dict_new ();
84 bless ($self, $class);
92 Returns the number of variables in the dictionary.
96 Returns the C<idx>th variable from the dictionary.
97 Returns undef if C<idx> is greater than or equal to the number
98 of variables in the dictionary.
106 my $var = pxs_get_variable ($dict, $idx);
110 bless ($var, "PSPP::Var");
117 =head3 get_var_by_name ($name)
119 Returns the variable from the dictionary whose name is C<name>.
120 If there is no such variable, a null reference will be returned.
128 my $var = pxs_get_var_by_name ($dict, $name);
132 bless ($var, "PSPP::Var");
144 Contains constants used to denote variable format types.
145 The identifiers are the same as those used in pspp to denote formats.
146 For example C<PSPP::Fmt::F> defines floating point format, and
147 C<PSPP::Fmt::A> denotes string format.
151 # These must correspond to the values in src/data/format.h
197 =head3 new ($dict, $name, %input_fmt)
199 Creates and returns a new variable in the dictionary C<dict>. The
200 new variable will have the name C<name>. C<name> must be a valid UTF8 string.
201 The input format is set by the C<input_fmt> parameter
203 By default, the write and print formats are the same as the input format.
204 The write and print formats may be changed (See L</set_write_format>),
205 L</set_print_format>). The input format may not be changed after
206 the variable has been created.
207 If the variable cannot be created, undef is returned.
217 my $self = pxs_dict_create_var ($dict, $name, \%format);
220 bless ($self, $class);
227 =head3 set_label ($label)
229 Sets the variable label to C<label>, which must be a valid UTF8 string.
236 =head3 set_write_format (%fmt)
238 Sets the write format to C<fmt>. <fmt> is a hash containing the keys:
244 A constant denoting the format type. See L</PSPP::Fmt>.
248 An integer denoting the number of decimal places for the format.
252 An integer denoting the width of the format.
256 On error the subroutine returns zero.
264 pxs_set_write_format ($var, \%format);
269 =head3 set_print_format (%fmt)
271 Sets the print format to C<fmt>.
272 On error the subroutine returns zero.
280 pxs_set_print_format ($var, \%format);
286 =head3 get_write_format ()
288 Returns a reference to a hash containing the write format for the variable.
291 =head3 get_print_format ()
293 Returns a reference to a hash containing the print format for the variable.
295 =head3 set_output_format (%fmt)
297 Sets the write and print formats to C<fmt>. This is the same as
298 calling set_write_format followed by set_print_format.
299 On error the subroutine returns zero.
304 sub set_output_format
308 pxs_set_output_format ($var, \%format);
313 =head3 clear_value_labels ()
315 Removes all value labels from the variable.
322 =head3 add_value_label ($key, $label)
324 Adds the value label C<label> to the variable for the value C<key>.
325 C<label> must be a valid UTF8 string.
326 On error the subroutine returns zero.
328 =head3 add_value_labels (@array)
339 while ( @li = each %values )
341 if ( $var->add_value_label ($li[0], "$li[1]") )
352 =head3 set_value_labels ($key, $label)
354 C<Set_value_labels> is identical to calling L</clear_value_labels>
355 followed by L</add_value_labels>.
356 On error the subroutine returns zero.
364 $self->clear_value_labels () ;
365 $self->add_value_labels (%labels);
370 =head3 set_missing_values ($val1 [, $val2[, $val3] ])
372 Sets the missing values for the variable.
373 No more than three missing values may be specified.
375 =head3 get_attributes()
377 Returns a reference to a hash of the custom variable attributes.
378 Each value of the hash is a reference to an array containing the
383 Returns the name of the variable.
387 Returns the label of the variable or undef if there is no label.
389 =head3 get_value_labels ()
391 Returns a reference to a hash containing the value labels for the variable.
392 The hash is keyed by data values which correpond to the labels.
396 package PSPP::Sysfile;
402 =head3 new ($filename, $dict [,%opts])
404 Creates a new system file from the dictionary C<dict>. The file will
405 be written to the file called C<filename>.
406 C<opt>, if specified, is a hash containing optional parameters for the
407 system file. Currently, the only supported parameter is
408 C<compress>. If C<compress> is non zero, then the system file written
409 will be in the compressed format.
410 On error, undef is returned.
413 =head3 append_case (@case)
415 Appends a case to the system file.
416 C<Case> is an array of scalars, each of which are the values of
417 the variables in the dictionary corresponding to the system file.
418 If the case contains strings, then the strings must be UTF8 encoded.
419 The special value C<PSPP::SYSMIS> may be used to indicate that a value
421 If the array contains less elements than variables in the dictionary,
422 remaining values will be set to system missing.
429 my $filename = shift;
433 my $self = pxs_create_sysfile ($filename, $dict, $opts);
437 bless ($self, $class);
446 Closes the system file.
448 This subroutine closes the system file and flushes it to disk. No
449 further cases may be written once the file has been closed.
450 The system file will be automatically closed when it goes out of scope.
454 package PSPP::Reader;
465 my $filename = shift;
467 my $self = pxs_open_sysfile ($filename);
471 bless ($self, $class);
478 =head3 open ($filename)
480 Opens a system file for reading.
482 Open is used to read data from an existing system file.
483 It creates and returns a PSPP::Reader object which can be used to read
484 data and dictionary information from C<filename>.
486 =head3 get_case_cnt ()
488 Returns the number of cases in a open system file. Some files
489 do not store the number of cases. In these instances undef
490 will be returned. Therefore, then programmer must check that the
491 returned value is not undef before using it.
499 my $dict = pxs_get_dict ($reader);
501 bless ($dict, "PSPP::Dict");
510 Returns the dictionary associated with the reader.
512 =head3 get_next_case ()
514 Retrieves the next case from the reader.
515 This method returns an array of scalars, each of which are the values of
516 the data in the system file.
517 The first call to C<get_next_case> after C<open> has been called retrieves
518 the first case in the system file. Each subsequent call retrieves the next
519 case. If there are no more cases to be read, the function returns an empty
522 If the case contains system missing values, these values are set to the
525 =head2 Miscellaneous subroutines
527 The following subroutines provide (hopefully) useful information about the
528 values retrieved from a reader.
530 =head3 PSPP::format_value ($value, $variable)
532 Returns a scalar containing a string representing C<value> formatted according
533 to the print format of C<variable>.
534 In the most common usage, C<value> should be a value of C<variable>.
537 =head3 PSPP::value_is_missing ($value, $variable)
539 Returns non-zero if C<value> is either system missing, or if it matches the
540 user missing criteria for C<variable>.
550 John Darrington, E<lt>john@darrington.wattle.id.auE<gt>
552 =head1 COPYRIGHT AND LICENSE
554 Copyright (C) 2007, 2008, 2009 by Free Software Foundation
556 This program is free software: you can redistribute it and/or modify
557 it under the terms of the GNU General Public License as published by
558 the Free Software Foundation, either version 3 of the License, or
559 (at your option) any later version.
561 This program is distributed in the hope that it will be useful,
562 but WITHOUT ANY WARRANTY; without even the implied warranty of
563 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
564 GNU General Public License for more details.
566 You should have received a copy of the GNU General Public License
567 along with this program. If not, see <http://www.gnu.org/licenses/>.