1 2007-07-22 Ben Pfaff <blp@gnu.org>
3 Fix problems with uniqueness of short names in system files with
4 very long string variables. Now a variable may have multiple
7 * automake.mk (src_data_libdata_a_SOURCES): Add new files
8 short-names.c, short-names.h.
10 * dictionary.c (dict_clone): Clone all the short names.
11 (compare_strings): Move into short-names.c.
12 (hash_strings): Ditto.
13 (set_var_short_name_suffix): Ditto.
14 (dict_assign_short_names): Ditto, rename short_names_assign,
15 change to assign all short names.
17 * por-file-writer.c (write_variables): Use short_names_assign
18 instead of dict_assign_short_names.
20 * short-names.c: New file.
22 * short-names.h: New file.
24 * sys-file-private.c (sfm_width_to_segments): New function.
26 * sys-file-reader.c (read_long_var_name_map): Save and restore all
27 the short names, not just the first one.
29 * sys-file-writer.c (cont_var_name): Removed.
30 (sfm_open_writer): Use short_names_assign instead of
31 dict_assign_short_names. Use unique short names assigned by
32 short_names_assign instead of those generated by cont_var_name.
34 * variable.c (struct variable): Remove `short_name' member,
35 replace by `short_names' and `short_name_cnt'.
36 (var_create) Initialize new members.
37 (var_get_short_name_cnt): New function.
38 (var_get_short_name): Now takes an index argument. Changed most
40 (var_set_short_name): Ditto.
41 (var_clear_short_name): Renamed var_clear_short_names, changed to
42 clear all short names.
44 2007-07-22 Ben Pfaff <blp@gnu.org>
46 * variable.c (var_set_width): Use new var_set_width function.
48 * missing-values.c (mv_n_values): Drop assertion, which was not
51 * format.c (fmt_default_for_width): New function.
52 (fmt_resize): New function.
54 2007-07-18 John Darrington <john@darrington.wattle.id.au>
56 * datasheet.c (datasheet_delete_columns): Added assertion to check
57 we're not deleting outside the range of the sheet.
60 * dictionary.c dictionary.h variable.c: Added the ability for string
61 variables to be resized.
63 * vardict.h: Added some prototypes (moved from dictionary.h) as
64 these should only be called by variable.c
67 2007-07-14 John Darrington <john@darrington.wattle.id.au>
69 * sfm-reader.c: Respect case_cnt field in file header.
71 2007-07-01 John Darrington <john@darrington.wattle.id.au>
73 * transformation.c transformation.h (trns_chain_execute): Changed the
74 signature (Patch #6057)
76 2007-06-10 Ben Pfaff <blp@gnu.org>
78 * casereader-filter.c (casereader_filter_destroy): Make sure to
79 write all the remaining excluded cases to the casewriter, if any.
81 * caseinit.c (init_list_destroy): Rewrite.
82 (init_list_clear): Ditto.
84 * casegrouper.c (casegrouper_get_next_group): Always set *reader
85 to null when returning false.
87 2007-06-06 Ben Pfaff <blp@gnu.org>
89 Actually implement the new procedure code and adapt all of its
90 clients to match. Also adapt all of the other case sources and
91 sinks in the tree and their clients to use the
92 casereader/casewriter infrastructure.
94 * automake.mk: Add and remove files.
96 * any-reader.c: Change into a casereader.
97 * por-file-reader.c: Ditto.
98 * scratch-reader.c: Ditto.
99 * sys-file-reader.c: Ditto.
101 * any-writer.c: Change into a casewriter.
102 * por-file-writer.c: Ditto.
103 * scratch-writer.c: Ditto.
104 * sys-file-writer.c: Ditto.
106 * procedure.c: Change to use casereader, casewriter, caseinit, and
107 other new infrastructure.
109 * scratch-handle.c: Adapt to new infrastructure.
111 * case-sink.c: Removed, now dead code.
112 * case-sink.h: Ditto.
113 * case-source.c: Ditto.
114 * case-source.h: Ditto.
115 * casefile-factory.c: Ditto.
116 * casefile-private.h: Ditto.
119 * casefilter.c: Ditto.
120 * casefilter.h: Ditto.
123 * fastfile-factory.c: Ditto.
124 * fastfile-factory.h: Ditto.
125 * storage-stream.c: Ditto.
126 * storage-stream.h: Ditto.
128 2007-06-06 Ben Pfaff <blp@gnu.org>
132 * automake.mk: Add new files.
134 * datasheet.c: New file.
136 * datasheet.h: New file.
138 2007-06-06 Ben Pfaff <blp@gnu.org>
140 Until now, the procedure code has provided a case to the
141 case_source, which has filled in the data values that come from
142 the active file. "Left" data values that don't come from the
143 active file naturally stay the same from case to case, because the
144 procedure code keeps using that same case.
146 One of the compromises that comes with the new procedure code is
147 that the active file allocates and provides its own case, which
148 the procedure code then has to resize to provide room for any
149 other variables that should go in the case and then fill in the
150 values of "left" variables. Then, when we're done with that case,
151 we have to save the values of "left" variables to copy into the
152 next case read from the active file.
154 The caseinit code helps with this.
156 * automake.mk: Add new files.
158 * caseinit.c: New file.
160 * caseinit.h: New file.
162 2007-06-06 Ben Pfaff <blp@gnu.org>
164 * value.h (value_cnt_from_width): New function.
166 * variable.c (var_get_value_cnt): Use new function.
168 2007-06-06 Ben Pfaff <blp@gnu.org>
170 Add casegrouper, to allow cases read from a given casereader to be
171 broken into groups, each of which has its own casereader.
172 Generally cases are grouped based on having equal values for some
175 * automake.mk: Add new files.
177 * casegrouper.c: New file.
179 * casegrouper.h: New file.
181 2007-06-06 Ben Pfaff <blp@gnu.org>
183 Add interface to lexicographical ordering of cases.
185 * automake.mk: Add new files.
187 * case-ordering.c: New file.
189 * case-ordering.h: New file.
191 2007-06-06 Ben Pfaff <blp@gnu.org>
193 Add casereaders and casewriters, the basis of the new data processing
194 implementation. A casereader is a uniform interface to reading cases
195 from a data source; a casewriter is a uniform interface to writing
196 cases to a data sink.
198 * automake.mk: Add new files.
200 * casereader-filter.c: New file.
202 * casereader-provider.h: New file.
204 * casereader-translator.c: New file.
206 * casereader.c: New file.
208 * casereader.h: New file.
210 * casewriter-provider.h: New file.
212 * casewriter-translator.c: New file.
214 * casewriter.c: New file.
216 * casewriter.h: New file.
218 2007-06-06 Ben Pfaff <blp@gnu.org>
220 "casewindow" data structure that extends the deque (from libpspp)
221 of cases with the ability to dump cases to disk if we get too many
224 * automake.mk: Add new files.
226 * casewindow.c: New file.
228 * casewindow.h: New file.
230 2007-06-06 Ben Pfaff <blp@gnu.org>
232 sparse_cases data structure that augments a sparse_array of cases
233 with the ability to dump cases to disk if we get too many cases in
236 * automake.mk: Add new files.
238 * sparse-cases.c: New file.
240 * sparse-cases.h: New file.
242 2007-06-06 Ben Pfaff <blp@gnu.org>
244 Adds a low-level on-disk case array data structure.
246 * automake.mk: Add new files.
248 * case-tmpfile.c: New file.
250 * case-tmpfile.h: New file.
252 2007-06-06 Ben Pfaff <blp@gnu.org>
254 In a couple of places we calculate the maximum number of cases to
255 keep in memory based on the user-defined workspace. Enable
256 centralizing the calculation through a new function.
258 * settings.c (get_workspace_cases): New function.
260 2007-06-06 Ben Pfaff <blp@gnu.org>
262 The casenumber type is defined in transformations.h, but case.h is
263 a more sensible place. Move it.
265 * case.h (CASENUMBER_MAX): New macro.
266 (typedef casenumber): Move here, from transformations.h.
268 2007-06-03 Ben Pfaff <blp@gnu.org>
270 Slightly generalize case_to_values and case_from_values functions.
272 * case.c (case_to_values): Rename case_copy_out, change interface.
273 (case_from_values): Rename case_copy_in, change interface.
275 * fastfile.c (fastfilereader_get_next_case): Update caller.
276 (write_case_to_disk): Ditto.
278 2007-06-02 Ben Pfaff <blp@gnu.org>
280 Clean up after a forgotten part of patch #5829.
282 * casedeque.h: Remove unused file.
284 * automake.mk: Remove casedeque.h from sources.
286 2007-05-10 Jason Stover <jhs@math.gcsu.edu>
288 * category.c: Removed redundant #include
290 2007-05-06 Ben Pfaff <blp@gnu.org>
292 Abstract the documents within a dictionary a little better.
293 Thanks to John Darrington for suggestion, initial version, and
296 * dictionary.c (struct dictionary): Change `documents' member from
297 char * to struct string.
298 (dict_clear): Destroy struct string.
299 (dict_get_documents): Convert struct string to char *.
300 (dict_set_documents): Set struct string. Pad to 80-character
302 (dict_clear_documents): New function.
303 (dict_add_document_line): New function.
304 (dict_get_document_line_cnt): New function.
305 (dict_get_document_line): New function.
307 * dictionary.h (macro DOC_LINE_LENGTH): New macro.
309 * sys-file-reader.c (read_documents): Use new document functions.
311 2007-04-19 John Darrington <john@darrington.wattle.id.au>
313 * sys-file-reader.c: When reading a system file which has no
314 long name table, automatically create one where the long names
315 are the lower case versions of the short names.
317 2007-04-22 Ben Pfaff <blp@gnu.org>
319 * dictionary.c (dict_set_split_vars): dict_destroy expects that
320 dict_clear will free most data related to the dictionary.
321 dict_clear does a decent job, except that dict_set_split_vars on
322 some systems won't actually free the dict's "split" member.
323 Instead, it'll allocate a 1-byte region. Fix this.
325 * value.c (value_copy): New function.
326 (value_set_missing): Ditto.
328 2007-04-22 John Darrington <john@darrington.wattle.id.au>
330 * Deleted existing category.h and moved cat-routines.h into
331 category.h Encapsulated struct cat_vals better.
333 2007-04-19 John Darrington <john@darrington.wattle.id.au>
335 * sys-file-reader.c: When reading a system file which has no
336 long name table, automatically create one where the long names
337 are the lower case versions of the short names.
339 2007-04-16 John Darrington <john@darrington.wattle.id.au>
341 * sys-file-reader.c: Some versions of Other Software seem to
342 produce system files with string variables' measure set to
343 zero. We'll assume these are supposed to be nominal variables.
345 2007-03-30 Ben Pfaff <blp@gnu.org>
347 * procedure.c: Adapt to new deque data structure.
349 Mon Feb 19 10:53:21 2007 John McCabe-Dansted <gmatht@gmail.com>
350 Ben Pfaff <blp@gnu.org>
352 * file-name.c: Mingw compatibility fixes.
353 (fn_search_path): Use ISSLASH instead of comparing against '/'
355 (fn_dir_name): Use dir_name from gnulib.
356 (fn_is_absolute): Use IS_ABSOLUTE_FILE_NAME from gnulib.
357 (fn_get_identity): Use GetFullPathName instead of canonicalize
358 from gnulib, because the latter does not fully support
359 Windows-style path names. Use this implementation based on the
360 detected presence of Windows instead of the absence of Unix, since
361 the new implementation is Windows-specific.
362 (fn_compare_file_identities): In Windows implementation, compare
363 names case-insensitively.
365 Sun Feb 18 13:28:02 2007 Ben Pfaff <blp@gnu.org>
367 * make-file.c: Don't include mkstemp.h, because gnulib now causes
368 <stdlib.h> to have the same effect.
370 Sun Feb 18 11:20:24 2007 Ben Pfaff <blp@gnu.org>
372 * por-file-reader.c: Add missing _() around messages.
374 Sun Feb 11 20:44:13 2007 Ben Pfaff <blp@gnu.org>
376 * make-file.c: Include "mkstemp.h", without which linking on
379 Thu Feb 8 14:59:05 2007 Ben Pfaff <blp@gnu.org>
381 Reduce platform dependence.
383 * file-name.c (fn_tilde_expand): Removed, and removed calls to it.
384 Everywhere we using this, we really should have just depended on
385 the shell to expand tildes.
386 (fn_search_path): Simplify, given that we don't do tilde expansion
388 (fn_normalize): Removed. Caller changed to use the canonicalize
390 (fn_get_cwd): Removed. Only user was fn_normalize.
391 (fn_is_absolute): Really only test for absolute names.
392 (fn_is_special): Use pipe files if HAVE_POPEN, not if we're in
394 (fn_readlink): Removed, as it was only used fn_normalize.
395 (fn_exists): Assume the stat function is available; gnulib does.
396 (fn_open): Use pipe files if HAVE_POPEN, not if we're in unix.
398 Sat Feb 3 21:52:17 2007 Ben Pfaff <blp@gnu.org>
400 * dictionary.c (dict_create_vector_assert): New function.
402 Wed Feb 7 21:25:15 2007 Ben Pfaff <blp@gnu.org>
404 * file-name.c (fn_normalize): Correct name of function
405 fn_is_special. Thanks to John McCabe-Dansted <gmatht@gmail.com>
406 for pointing this out.
408 Thu Feb 1 16:53:37 2007 Ben Pfaff <blp@gnu.org>
410 We are using a single member in struct file_handle, the "name"
411 field, for more than one purpose. When it begins with '"', it's a
412 file name; otherwise, it's a token that can be used to identify
413 it. When that assertion fires, it's because we searched for the
414 name case-sensitively as a file name (so that there was no match),
415 and then we try to insert it case-insensitively as a token, which
416 fails because duplicates aren't allowed.
418 Solution: break the two purposes into two separate fields. This
419 fixes the problem and likely makes the code easier to read too.
421 Fixes bug #18922. Thanks to John Darrington for bug report and
424 * file-handle-def.c (struct file_handle): New `id' member.
425 (fh_from_name): Rename fh_from_id. Update all callers.
426 (create_handle): New `id' parameter. Update all callers.
427 (fh_create_file): Ditto.
428 (fh_get_id): New function.
430 Mon Jan 15 16:18:10 2007 Ben Pfaff <blp@gnu.org>
432 * case.c (case_is_null): Change return type to bool.
434 Mon Jan 15 10:57:28 2007 Ben Pfaff <blp@gnu.org>
438 * case.c (case_clone) [DEBUGGING]: When debugging, don't use
439 reference counting to share data. This makes it easy for
440 valgrind, etc. to find accesses to cases that have been destroyed
441 but have been kept around by another user's ref-count. This often
442 happens when the data set is small enough to find in memory; if a
443 bigger data set that would overflow to disk were used, then data
444 corruption would occur.
446 Mon Jan 15 10:55:18 2007 Ben Pfaff <blp@gnu.org>
450 * case.c (case_unshare): Make it check internally whether the
451 ref_cnt is greater than 1, so that the callers don't have to.
452 Update callers not to check.
454 Mon Jan 15 10:53:01 2007 Ben Pfaff <blp@gnu.org>
456 Before, I was thinking that I might want to get rid of reference
457 counting at some point. Now, I'm pretty sure that it's here to
458 stay. Thus, because we have to store the value_cnt anyway for
459 reference-counted cases, we might as well expose it to users.
461 * case.c (case_get_value_cnt): New function.
462 (case_resize): Drop OLD_CNT argument. Update all callers. Only
463 resize case if its size actually changed.
465 * casefile.c (casefile_append_xfer): Use case_get_value_cnt
466 instead of peeking inside struct case directly.
467 (casefile_append): Ditto.
469 Mon Jan 15 10:50:22 2007 Ben Pfaff <blp@gnu.org>
471 Get rid of the inlines for the case functions, which made the
472 header file hard to read. (Also, in testing with "-O2 -DNDEBUG",
473 the inlines didn't speed up "make check" at all, which is not a
474 perfect benchmark but seems indicative.)
476 * case.c: Remove #ifdef DEBUGGING...#endif around many function
477 definitions. Remove some assertions on nonnull pointers that were
478 redundant with a pointer dereference soon after in the function.
480 (struct case_data): Move definition here from case.h.
484 (case_data_wr): Ditto.
486 Sun Jan 14 21:41:12 2007 Ben Pfaff <blp@gnu.org>
488 * automake.mk: Add casedeque.h to sources.
490 * casedeque.h: New file.
492 * procedure.c: (struct dataset) Change lag_count, lag_head,
493 lag_queue member into single struct casedeque member. Update all
494 users to use the casedeque instead.
497 Sun Jan 14 21:43:12 2007 Ben Pfaff <blp@gnu.org>
499 * procedure.c: Simplify lagged cases interface. Updated all
500 clients--well, the only client--to use the simplified interface.
501 (dataset_n_lag) Removed.
502 (dataset_set_n_lag) Removed.
503 (dataset_need_lag) New function.
505 Tue Jan 9 07:20:05 WST 2007 John Darrington <john@darrington.wattle.id.au>
507 * dictionary.c procedure.c: More changes to ensure that callbacks occur
508 whenever appropriate, but only when the dataset/dictionary is in a
511 Sun Jan 7 08:33:04 WST 2007 John Darrington <john@darrington.wattle.id.au>
513 * dictionary.c dictionary.h : Added callbacks for change of filter and
514 split variables. Refactored some code to ensure that callbacks get
515 invoked when appropriate.
517 * procedure.c (proc_cancel_temporary_transformations): Make sure that
518 replace_dict callback occurs when permanent_dict replaces the current
521 Wed Jan 3 11:02:11 WST 2007 John Darrington <john@darrington.wattle.id.au>
523 * dictionary.c dictionary.h : Added callback for when the weight
524 variable of a dictionary changes.
526 Mon Jan 1 10:36:26 WST 2007 John Darrington <john@darrington.wattle.id.au>
528 * dictionary.c dictionary.h : Added replace_source and replace_dict
529 callbacks, and functions to deal with them.
531 Fri Dec 22 13:56:08 2006 Ben Pfaff <blp@gnu.org>
533 Simplify missing value handling.
535 * missing-values.h (enum mv_class): New type.
536 (enum mv_type): Moved definition into missing-values.c and renamed
537 each MV_* to MVT_*, to distinguish them from the exposed mv_class
538 enums. Updated all uses.
539 (struct missing_values): Changed type of `type' from `enum
540 mv_type' to `int' because the definition is no longer exposed.
542 * missing-values.c (mv_is_value_missing): Add new enum mv_class
543 parameter. Update all callers.
544 (mv_is_num_missing): Ditto.
545 (mv_is_str_missing): Ditto.
546 (mv_is_value_user_missing): Removed. Changed callers to use
548 (mv_is_num_user_missing): Removed. Changed callers to use
550 (mv_is_str_user_missing): Removed. Changed callers to use
552 (mv_is_value_system_missing): Removed. Changed callers to use
554 (mv_set_type): Removed. Changed callers to use mv_clear.
555 (mv_clear): New function.
557 * variable.c (var_is_value_missing): Add new enum mv_class
558 parameter. Update all callers.
559 (var_is_num_missing): Ditto.
560 (var_is_str_missing): Ditto.
561 (var_is_value_user_missing): Removed. Changed callers to use
562 var_is_value_missing.
563 (var_is_num_user_missing): Removed. Changed callers to use
565 (var_is_str_user_missing): Removed. Changed callers to use
567 (var_is_value_system_missing): Removed. Changed callers to use
568 var_is_value_missing.
570 * casefilter.c (struct casefilter): Use enum mv_class in place of
572 (casefilter_variable_missing): Adapt to new member.
573 (casefilter_create): Change signature to take enum mv_class,
576 Fri Dec 22 20:08:38 WST 2006 John Darrington <john@darrington.wattle.id.au>
578 * casefile-factory.h fastfile-factory.c fastfile-factory.h: New files.
580 * case-sink.c case-sink.h procedure.c procedure.h
581 storage-stream.c: Now uses the factory.
583 Sat Dec 16 22:05:18 2006 Ben Pfaff <blp@gnu.org>
585 Make it possible to pull cases from the active file with a
586 function call, instead of requiring indirection through a callback
589 * case-source.h (struct case_source_class): Change ->read function
590 to return a single case, instead of calling a callback function
591 for each case. Change ->destroy function to return an error
594 * case-source.c (free_case_source): Pass along the value returned
595 by the case_source ->destroy function.
597 * procedure.c (struct write_case_data): Removed.
598 (struct dataset): Added some members to track procedure state.
599 (procedure): Optimize the trivial case at this level.
600 (internal_procedure): Re-implement in terms of proc_open,
601 proc_read, proc_close.
602 (proc_open) New function.
603 (proc_read) New function.
604 (proc_close) New function.
605 (write_case) Moved into proc_read.
606 (close_active_file) Moved closing of data source into proc_close.
608 * storage-source.c: Rewrote to conform with modified
609 case_source_class interface.
611 * transformations.c (trns_chain_execute): Added argument to allow
612 starting execution from an arbitrary transformation. Updated
615 * transformations.h (enum TRNS_NEXT_CASE) Renamed TRNS_END_CASE.
617 Sat Dec 16 14:09:25 2006 Ben Pfaff <blp@gnu.org>
619 * sys-file-reader.c (read_display_parameters): Don't assume that
620 MEASURE_* and ALIGN_* have the same values found in system files.
622 * sys-file-writer.c (write_variable_display_parameters): Ditto.
624 * variable.h: Change MEASURE_NOMINAL, MEASURE_ORDINAL,
625 MEASURE_SCALE to be 0-based instead of 1-based. This also fixes
626 the value of n_MEASURES, which was off by 1 (at least from my
629 Sat Dec 16 12:17:34 WST 2006 John Darrington <john@darrington.wattle.id.au>
631 * dictionary.c dictionary.h vardict.h variable.c: Added optional
632 callbacks which are invoked when the dictionary or its
633 variables are changed.
635 * missing-values.c missing-values.h value-labels.c: Tidied up
636 consistency checks, and made some of them return false
637 instead of assert-failing.
639 Wed Dec 13 19:30:11 2006 Ben Pfaff <blp@gnu.org>
641 * calendar.c (calendar_days_in_month): New function.
643 Mon Dec 11 07:53:39 2006 Ben Pfaff <blp@gnu.org>
645 * value-labels.c (hash_int_val_lab): Only hash as many bytes as
646 the value label's width.
648 Sun Dec 10 14:21:29 2006 Ben Pfaff <blp@gnu.org>
650 * sfm-private.h: Move contents into sys-file-writer.c, which is
651 the only remaining user. Removed Borland C++-specific directives.
653 * sys-file-reader.c: Clean up and rewrite entire file. The
654 rewritten version is simpler and better abstracted, and should be
655 easier to maintain and extend. It avoids using structures to read
656 file data, which is prone to padding variations among compilers.
657 It should also handle non-IEEE 754 system files, although I
658 haven't been able to find any. It has been tested against many
659 .sav files obtained from the Web and found to produce the same
660 results as the earlier version of the code, or in some cases
661 improved results. It is more tolerant of format variations found
664 * sys-file-reader.h (struct sfm_read_info): Removed `big_endian'
665 member, putting an enum integer_format in its place. New member
666 `float_format'. Changed `compressed' member to type bool.
668 Sun Dec 10 13:48:53 2006 Ben Pfaff <blp@gnu.org>
670 * dictionary.c (dict_delete_consecutive_vars): New function.
672 Sat Dec 9 20:08:25 2006 Ben Pfaff <blp@gnu.org>
674 * file-name.c (fn_search_path): Remove prefix arg that was unused
675 by any caller. Updated all callers.
677 Sat Dec 9 20:04:22 2006 Ben Pfaff <blp@gnu.org>
679 * format.c (fmt_dollar_template): Use user's decimal point
680 character. Add assertion.
682 Sat Dec 9 20:02:25 2006 Ben Pfaff <blp@gnu.org>
684 * format.c (fmt_dollar_template): New function, based on
685 dollar_format_template from var-type-dialog.c.
687 Sat Dec 9 18:05:59 2006 Ben Pfaff <blp@gnu.org>
689 * data-out.c (output_scientific): Fix bad assumption that "buf" is
692 Sat Dec 9 17:23:23 2006 Ben Pfaff <blp@gnu.org>
694 Finish converting struct variable to an opaque type. In this
695 phase, we add remaining setter and getter functions, convert the
696 remaining PSPP code to use them, and do a bunch of cleanup. The
697 resulting changes are pervasive but mostly trivial, and only the
698 notable changes are logged.
700 * automake.mk (src_data_libdata_a_SOURCES): Add the new source
703 * case.c (case_data): Renamed case_data_idx.
704 (case_num): Renamed case_num_idx.
705 (case_str): Renamed case_str_idx.
706 (case_data_rw): Renamed case_data_rw_idx.
708 * case.h (case_data): New function with old name and an interface
709 that takes a variable instead of an index, which is easier to
710 use. Updated all callers to use the new interface, or to use the
711 new *_idx function (see above).
714 (case_data_rw): Ditto.
716 * category.c (cat_stored_values_destroy): Changed interface to
717 take a struct cat_vals * instead of a struct variable *.
719 * dictionary.c (dict_clone): Use new vector_clone function.
720 (dict_clear) Use new var_destroy function.
721 (add_var) New function.
722 (dict_create_var) Rewrite in terms of dict_create_var_assert.
723 (dict_create_var_assert) Rewrite in terms of add_var.
724 (dict_clone_var) Rewrite in terms of dict_clone_var_assert.
725 (dict_clone_var_assert) Rewrite in terms of var_clone, add_var.
726 (dict_lookup_var) Use new var_create, var_destroy functions.
727 (dict_contains_var) Rewrite in terms of new vardict functionality.
728 (set_var_dict_index) New function.
729 (set_var_case_index) New function.
730 (reindex_vars) New function.
731 (dict_delete_var) Rewrite in terms of new vardict functionality.
732 (dict_reorder_var) Ditto.
733 (dict_reorder_vars) Ditto.
734 (rename_var) New function.
735 (dict_rename_var) Use rename_var.
736 (dict_rename_vars) Use pool to simplify code. Use rename_var.
737 (dict_get_compacted_idx_to_fv) Rename
738 dict_get_compacted_dict_index_to_case_index, update callers.
739 (dict_create_vector) Use new vector_create function.
740 (dict_clear_vectors) Use new vector_destroy function.
741 (set_var_short_name_suffix) Move here from variable.c, renamed
742 from var_set_short_name_suffix, make static, update caller.
744 * sys-file-private.c: New file.
745 (sfm_width_to_bytes) Moved here from variable.c, renamed from
746 width_to_bytes, update callers.
748 * sys-file-private.h: New file. Later it will supplant
749 sfm-private.h; for now it supplements it.
750 (macro MIN_VERY_LONG_STRING) New macro.
751 (macro EFFECTIVE_LONG_STRING_LENGTH) New macro, from value.h.
753 * sys-file-reader.c: Use MIN_VERY_LONG_STRING - 1 where
754 MAX_LONG_STRING was used before.
756 * sys-file-writer.c: Ditto.
758 * value-labels.c: Change the paradigm here to be that a null
759 pointer is OK for a struct val_labs * in most cases; it just
760 represents an empty set of value labels.
761 (val_labs_copy) A copy of a null set is a null set.
762 (val_labs_count) A null set has 0 labels.
763 (val_labs_replace) Change return type to void. Rewrite for
765 (val_labs_find) A null set does not contain the value.
766 (value_to_string) Moved to variable.c, renamed var_get_value_name,
767 transposed argument order, updated all callers.
770 (value_dup) Moved here from variable.c.
771 (compare_values) Ditto.
774 * value.h: (macro MAX_SHORT_STRING) Rewrote for simplicity.
775 (macro MAX_LONG_STRING) Removed, because it was only interesting
776 for system files, not for general code.
777 (macro MAX_VERY_LONG_STRING) Ditto.
778 (macro EFFECTIVE_LONG_STRING_LENGTH) Moved to sys-file-private.h.
779 (macro MAX_ELEMS_PER_VALUE) Removed, as it was unused.
781 * vardict.h: New file, for an interface between variables and
784 * variable.c: A lot of functions were moved around, for better
786 (struct variable) Move definition here, from variable.h.
787 (var_type_adj) Removed--makes i18n hard.
788 (var_type_noun) Ditto.
789 (var_create) New function.
790 (var_clone) New function.
791 (var_destroy) New function.
792 (var_set_name) Assert that variable is not in a dictionary.
793 (compare_var_names) Rename compare_vars_by_name and fix a couple
794 of callers who thought the args were strings.
795 (hash_var_name) Rename hash_var_by_name.
796 (compare_var_ptr_names) Rename compare_var_ptrs_by_name.
797 (hash_var_ptr_name) Rename hash_var_ptr_by_name.
798 (var_is_very_long_string) Removed, because it was only interesting
800 (var_set_missing_values) Allow the argument to be the wrong width,
801 as long as we can resize it. Simplify callers who were doing the
803 (var_get_value_labels) New function.
804 (var_has_value_labels) New function.
805 (var_set_value_labels) New function.
806 (alloc_value_labels) New function.
807 (var_add_value_label) New function.
808 (var_replace_value_label) New function.
809 (var_clear_value_labels) New function.
810 (var_lookup_value_label) New function.
811 (var_get_value_name) Moved here from variable.c, renamed from
812 var_get_value_name, transposed argument order, updated all
814 (var_to_string) Moved here, from variable-label.c.
815 (var_set_leave) New function.
816 (var_get_leave) New function.
817 (var_must_leave) New function.
818 (var_set_short_name_suffix) Moved to dictionary.c, renamed
819 set_var_short_name_suffix.
820 (var_get_dict_index) New function.
821 (var_get_case_index) New function.
822 (var_get_obs_vals) New function.
823 (var_set_obs_vals) New function.
824 (var_has_obs_vals) New function.
825 (var_get_vardict) New function.
826 (var_set_vardict) New function.
827 (var_has_vardict) New function.
828 (var_clear_vardict) New function.
829 (value_dup) Moved to value.c.
830 (compare_values) Ditto.
833 * variable.h: (enum NUMERIC) Rename VAR_NUMERIC, update all users.
834 (enum ALPHA) Rename VAR_STRING, update all users.
836 * vector.c: New file.
837 (struct vector) Moved here, from variable.h.
838 (check_widths) New function.
839 (vector_create) New function.
840 (vector_clone) New function.
841 (vector_destroy) New function.
842 (vector_get_name) New function.
843 (vector_get_var) New function.
844 (vector_get_var_cnt) New function.
845 (compare_vector_ptrs_by_name) New function.
847 * vector.h: New file.
849 Sun Dec 10 11:32:56 WST 2006 John Darrington <john@darrington.wattle.id.au>
851 * casefilter.c (casefilter_variable_missing): Avoided comparision of
852 string variables to SYSMIS. Thanks to Ben Pfaff for reporting this
855 Sat Dec 9 07:18:03 WST 2006 John Darrington <john@darrington.wattle.id.au>
857 * value-labels.c (destroy_atoms): New function.
858 * value-labels.c (atom_create): Call destroy_atoms in atexit handler.
860 Thu Dec 7 17:38:26 2006 Ben Pfaff <blp@gnu.org>
862 Thanks to Jason Stover for pointing out this problem.
864 * data-out.c (output_number): Use gsl_finite from GSL, which is
865 portable, instead of isfinite, which is not.
868 Thu Dec 7 15:22:38 WST 2006 John Darrington <john@darrington.wattle.id.au>
870 * variable.c variable.h (value_dup): New function.
872 Mon Dec 4 22:20:17 2006 Ben Pfaff <blp@gnu.org>
874 Start converting struct variable to an opaque type. In this
875 phase, we add a bunch of setter and getter functions and convert
876 most of the PSPP code to use them. The resulting changes are
877 pervasive but mostly trivial, and only the notable changes are
880 * format.c (fmt_equal): New function.
882 * variable.c (var_type_is_valid): New function.
883 (measure_is_valid) Moved here, from format.c.
884 (alignment_is_valid) Moved here, from format.c.
885 (var_get_name) New function.
886 (var_set_name) New function.
887 (width_to_type) New function.
888 (var_get_type) New function.
889 (var_get_width) New function.
890 (var_set_width) New function.
891 (var_is_numeric) New function.
892 (var_is_alpha) New function.
893 (var_is_short_string) New function.
894 (var_is_long_string) New function.
895 (var_is_very_long_string) New function.
896 (var_get_missing_values) New function.
897 (var_set_missing_values) New function.
898 (var_clear_missing_values) New function.
899 (var_has_missing_values) New function.
900 (var_is_value_missing) New function.
901 (var_is_num_missing) New function.
902 (var_is_str_missing) New function.
903 (var_is_value_user_missing) New function.
904 (var_is_num_user_missing) New function.
905 (var_is_str_user_missing) New function.
906 (var_is_value_system_missing) New function.
907 (var_get_print_format) New function.
908 (var_set_print_format) New function.
909 (var_get_write_format) New function.
910 (var_set_write_format) New function.
911 (var_set_both_formats) New function.
912 (var_get_label) New function.
913 (var_set_label) New function.
914 (var_clear_label) New function.
915 (var_has_label) New function.
916 (var_get_measure) New function.
917 (var_set_measure) New function.
918 (var_get_display_width) New function.
919 (var_set_display_width) New function.
920 (var_get_alignment) New function.
921 (var_set_alignment) New function.
922 (var_get_value_cnt) New function.
923 (var_get_leave) New function.
924 (var_get_short_name) New function.
926 * variable.h: (struct variable) Removed "type" and "nv" members;
927 they are now computed from "width" where needed.
929 Mon Dec 4 21:38:40 2006 Ben Pfaff <blp@gnu.org>
931 * missing-values.c (mv_resize): Don't write beyond end of the
932 allocated buffer when resizing a long string.
934 Sat Dec 2 16:28:32 2006 Ben Pfaff <blp@gnu.org>
936 Clean up identifier code: don't require identifier enumerations to
937 be in a particular order; make better use of string library;
938 expose less of the internals.
940 * identifier.c: (lex_skip_identifier) Rename lex_id_get_length,
941 change interface. Updated all callers.
942 (lex_id_match) Change interface to use struct substring, update
944 (lex_id_match_len) Removed. Update callers to use lex_id_match.
945 (global array keywords[]) Make static, change form. Update all
946 users to use lex_id_name instead.
947 (lex_is_keyword) New function.
948 (lex_id_to_token) Change interface to use struct substring, update
950 (lex_id_name) New function.
952 * identifier.h: (T_FIRST_KEYWORD) Removed. Changed users to call
953 lex_is_keyword instead.
954 (T_LAST_KEYWORD) Removed.
955 (T_N_KEYWORDS) Removed.
957 Sat Nov 18 20:46:35 2006 Ben Pfaff <blp@gnu.org>
959 * format.c: (fmt_date_template) Distinguish characters for which a
960 space is output and any date delimiter is allowed on input, from
961 those for which a space is output and only a space is allowed on
962 input. The former is represented by X, the latter by a space.
963 Also, drop distinction between h and H, changing the former to the
966 * data-in.c: Completely rewrite internals to conform to SPSS input
967 formats as closely as possible.
968 (data_in) Changed external interface by replacing the structure
969 that was used as a single argument by a set of arguments. Updated
971 (data_in_finite_line) Removed. Converted all callers to use plain
973 (data_in_get_integer_format) New function.
974 (data_in_set_integer_format) New function.
975 (data_in_get_float_format) New function.
976 (data_in_set_float_format) New function.
978 * data-in.h: (enums DI_IGNORE_ERROR, DI_IMPLIED_DECIMALS) Removed.
979 (struct data_in) Removed.
981 * data-out.c: (output_date) Drop each component from the input as
982 it is output, to allow us to drop the distinction between h (a
983 count of hours) and H (the hour of day) template characters.
984 Also, handle new X template character.
985 (output_scientific) Follow more rational rule on when to drop
986 fraction introduced between SPSS 13 and 15. Updated test case to
989 Sat Nov 11 11:41:26 2006 Ben Pfaff <blp@gnu.org>
991 Fix buffer overflow reported by John Darrington.
993 * data-out.c (output_bcd_integer): In case of SYSMIS, etc.,
994 realize that DIGITS is a count of nibbles, not of bytes.
996 Sat Nov 4 15:59:56 2006 Ben Pfaff <blp@gnu.org>
998 * calendar.c (calendar_offset_to_gregorian) Also return the
999 year-of-day. Change callers to new interface.
1001 * data-out.c: Completely rewrite internals to conform to SPSS
1002 output formats as completely as possible.
1003 (data_out) Change interface to put input parameters before output
1004 parameters, for consistency with the style I now prefer. Update
1006 (data_out_get_integer_format) New public function.
1007 (data_out_set_integer_format) New public function.
1008 (data_out_get_float_format) New public function.
1009 (data_out_set_float_format) New public function.
1011 * data-out.h: New file. Move prototype for data_out here, from
1014 * format.c: (fmt_step_width) Use equality comparison instead of
1015 bitwise and, for clarity.
1016 (fmt_is_string) Ditto.
1017 (fmt_input_to_output) Fix categories that are translated to F
1020 Sun Nov 5 08:29:34 WST 2006 John Darrington <john@darrington.wattle.id.au>
1022 * casefilter.c casefilter.h (new files), casefile.c casefile.h
1023 casefile-private.h: Added casefilter to assist commands with missing
1026 Sat Nov 4 11:47:09 2006 Ben Pfaff <blp@gnu.org>
1028 Implement SET ERRORS, SHOW ERRORS. Fixes bug #17609.
1030 * settings.c: (route_errors_to_terminal) New variable.
1031 (route_errors_to_listing) New variable.
1032 (get_error_routing_to_terminal) New function.
1033 (set_error_routing_to_terminal) New function.
1034 (get_error_routing_to_listing) New function.
1035 (set_error_routing_to_listing) New function.
1037 * settings.h: (SET_ROUTE_* enums) Removed, because unused.
1039 Tue Oct 31 19:58:27 2006 Ben Pfaff <blp@gnu.org>
1041 * format.c: Completely rewrite, to achieve better abstraction.
1042 Rewrite all references to formats in other files.
1044 * format.def: Rewrite and reorganize.
1046 * settings.c: Move everything related to custom currency formats
1047 into format.[ch], changing them in form, so as to group related
1048 code and definitions better. Changed all references to use the
1050 (static var decimal) Removed.
1051 (static var grouping) Removed.
1052 (static var cc) Removed.
1053 (get_decimal) Removed.
1054 (set_decimal) Removed.
1055 (get_grouping) Removed.
1056 (set_grouping) Removed.
1060 * settings.h: (macro CC_CNT) Removed.
1061 (macro CC_WIDTH) Removed.
1062 (struct custom_currency) Removed.
1064 Tue Oct 31 19:56:19 2006 Ben Pfaff <blp@gnu.org>
1066 * data-in.c (data_in): Use switch statement instead of table, to
1067 avoid dependence on the order of the FMT_* enums.
1069 Tue Oct 31 19:35:36 2006 Ben Pfaff <blp@gnu.org>
1071 * data-out.c: (num_to_string) Removed, because it was dead code.
1073 Tue Oct 31 18:09:24 2006 Ben Pfaff <blp@gnu.org>
1075 * data-in.c (parse_trailer): Fix error message.
1077 Sat Oct 28 11:56:50 2006 Ben Pfaff <blp@gnu.org>
1079 * format.c (fmt_is_binary): New function.
1081 Thu Oct 19 22:59:56 WST 2006 John Darrington <john@darrington.wattle.id.au>
1083 * procedure.c procedure.h: Encapsulated the static data into a single
1086 Sat Oct 14 16:56:44 2006 Ben Pfaff <blp@gnu.org>
1088 * casefile.c (casereader_read_xfer): Always initialize the case,
1089 even on an error condition.
1091 Wed Sep 27 09:37:49 WST 2006 John Darrington <john@darrington.wattle.id.au>
1093 * procedure.c (case_limit_trns_proc): Fixed buglet which rendered the
1094 entire function useless.
1096 Mon Sep 25 17:11:46 WST 2006 John Darrington <john@darrington.wattle.id.au>
1098 * casefile-private.h casefile.c casefile.h fastfile.c: Created new
1099 casereader method casereader_clone.
1101 * procedure.c pransformations.h: Introduced new type casenum_t
1103 Thu Sep 21 07:00:30 2006 Ben Pfaff <blp@gnu.org>
1105 * variable.c: (width_to_bytes) Rephrase code for clarify.
1107 Sun Jul 16 19:52:03 2006 Ben Pfaff <blp@gnu.org>
1109 * format.c: (fmt_type_from_string) New function.
1110 (fmt_to_string) Include decimals in output if the format has
1111 decimals, even if the format type does not. This way, we can
1112 accurately reproduce incorrect formats in user output.
1113 (check_common_specifier) Make the check for a bad format type an
1114 assertion, so we get bug reports if they show up. Fix message.
1115 Check for decimal places with a format type that doesn't allow
1117 (check_input_specifier) Remove check for FMT_X, which has been
1119 (check_output_specifier) Ditto.
1121 * format.def: Remove FMT_T, FMT_X, FMT_DESCEND, FMT_NEWREC.
1123 * format.h: (macro FMT_TYPE_LEN_MAX) New macro.
1124 (struct fmt_desc) Use FMT_TYPE_LEN_MAX in definition.
1125 (enum fmt_parse_flags) Removed.
1127 Mon Jul 17 18:26:21 WST 2006 John Darrington <john@darrington.wattle.id.au>
1129 * casefile.c casefile.h: Converted to an abstract base class.
1130 * casefile-private.h fastfile.c fastfile.h: New files.
1131 * automake.mk procedure.c scratch-writer.c storage-stream.c
1133 Wed Jul 12 21:02:26 2006 Ben Pfaff <blp@gnu.org>
1135 * procedure.c (internal_procedure): Create sink_case with only as
1136 many values as the compacted dictionary.
1138 Wed Jul 12 21:01:00 2006 Ben Pfaff <blp@gnu.org>
1140 Remove "debugging" code that caused plenty of false positives and
1143 * case.h (struct ccase): [DEBUGGING] Remove `this' member.
1145 * case.c: Remove all references to `this' member.
1147 Thu Jul 6 19:09:53 2006 Ben Pfaff <blp@gnu.org>
1149 Fix link error noted by Jason Stover.
1151 * storage-stream.c: Include <assert.h>.
1153 Tue Jul 4 08:47:35 2006 Ben Pfaff <blp@gnu.org>
1155 Fix bug #15766 (/KEEP subcommand on SAVE doesn't fully support
1156 ALL) and additional underlying system file issues.
1158 Thanks to John Darrington for review.
1160 First problem: var_hash points to variables not owned by the
1161 sys-file-reader, which the caller may free or modify. Use an
1162 array of sfm_vars instead, as done earlier (e.g. CVS version
1165 * sys-file-reader.c (struct sfm_reader): Remove var_hash, svars
1166 members and remove all code that references it. Add vars, var_cnt
1167 members. Remove fix_specials member, which was unused.
1168 (struct sfm_var) Remove name member, which was unused.
1169 (sfm_close_reader) Free vars member instead of var_hash.
1170 (compare_var_shortnames) Removed.
1171 (hash_var_shortname) Removed.
1172 (sfm_open_reader) Fill out vars array.
1173 (compare_var_index) Removed.
1174 (sfm_read_case) Use vars instead of var_hash.
1176 Second problem: we're confused about when we actually have very
1177 long strings, causing us to choose incorrectly between slow path
1178 and fast path in sfm_read_case.
1180 * sys-file-reader.c: (sfm_open_reader) Only mark has_vls if we
1181 have very long strings, not when we have long variable names,
1182 which is an unrelated feature.
1184 Tue Jun 27 12:06:49 2006 Ben Pfaff <blp@gnu.org>
1186 * variable.h: Move var_set and variable parsing declarations to
1187 new header, src/language/lexer/variable-parser.h. Modified lots
1188 of files to include the new header.
1190 Sun Jun 25 22:39:32 2006 Ben Pfaff <blp@gnu.org>
1192 * value-labels.c (value_to_string): When there's no value label,
1193 format the variable according to its print format, instead of
1194 always effectively using A or F format.
1196 Mon Jun 19 18:05:42 WST 2006 John Darrington <john@darrington.wattle.id.au>
1198 * casefile.c (casefile_get_random_reader): Nasty hack to get around
1201 * format.c: Removed tortological assertion.
1203 Fri Jun 9 12:20:09 2006 Ben Pfaff <blp@gnu.org>
1205 Reform string library.
1207 * file-name.c (fn_interp_vars): Change interface to take a
1208 substring as input. Updated all users.
1210 Fri Jun 9 12:11:24 2006 Ben Pfaff <blp@gnu.org>
1212 * format.c (measure_is_valid): Really return false when m >=
1215 Tue Jun 6 18:46:26 2006 Ben Pfaff <blp@gnu.org>
1217 Implement random access to casefiles, for use in GUI.
1219 * casefile.c: (struct casereader) Add `random', `file_ofs',
1220 `buffer_ofs' members.
1221 (casefile_get_random_reader) New function.
1222 (read_open_file) Break part into new function
1223 seek_and_fill_buffer().
1224 (fill_buffer) Update buffer_ofs, file_ofs.
1225 (casereader_seek) New function.
1227 Tue May 30 19:52:33 WST 2006 John Darrington <john@darrington.wattle.id.au>
1229 * settings.c: Added call to i18n{done, init}.
1231 Tue May 9 21:09:17 2006 Ben Pfaff <blp@gnu.org>
1233 * procedure.h: Add WARN_UNUSED_RESULT to procedure function
1236 Tue May 9 21:08:05 2006 Ben Pfaff <blp@gnu.org>
1238 * casefile.c: Convert many uses of `int' to `bool'.
1240 Sat May 6 22:49:43 2006 Ben Pfaff <blp@gnu.org>
1242 * transformations.c (trns_chain_destroy): Destroy chain's trns
1243 member, to fix memory leak.
1245 Sat May 6 22:48:30 2006 Ben Pfaff <blp@gnu.org>
1247 * storage-stream.c (storage_source_decapsulate): Destroy case
1248 source to fix memory leak.
1250 Sat May 6 22:46:47 2006 Ben Pfaff <blp@gnu.org>
1252 * scratch-reader.c (scratch_reader_read_case): Copy into existing
1253 case passed as argument instead of initializing the argument as a
1254 case. Fixes memory leak that showed up in
1255 tests/command/aggregate.sh with scratch files.
1257 Sat May 6 22:45:55 2006 Ben Pfaff <blp@gnu.org>
1259 * procedure.c (proc_done): Destroy default_dict, to fix memory
1262 Sat May 6 22:44:44 2006 Ben Pfaff <blp@gnu.org>
1264 Simplify procedure_with_splits().
1266 * procedure.c (struct split_aux_data): Removed case_count member.
1267 (procedure_with_splits) Don't initialize case_count.
1268 (split_procedure_case_func) Check whether prev_case is null
1269 instead of case_count.
1270 (split_procedure_end_func) Ditto.
1272 Sat May 6 22:42:23 2006 Ben Pfaff <blp@gnu.org>
1274 * case.c (case_move): Do nothing if dst and src are the same
1276 (case_try_create) Merge two similar cases.
1277 (case_copy) Unshare only if data must be actually copied.
1279 Sun May 7 10:04:06 WST 2006 John Darrington <john@darrington.wattle.id.au>
1281 * data-in.c data-out.c dictionary.c sys-file-reader.c
1282 sys-file-writer.c variable.c variable.h: Reworked very long string
1283 support for better encapsulation.
1285 Sat May 6 19:02:00 2006 Ben Pfaff <blp@gnu.org>
1287 * value-labels.c (val_labs_can_set_width): New function.
1288 (val_labs_set_width) Clear labels if increasing width to long
1290 (val_labs_destroy) Remove unneeded test for null.
1292 Sat May 6 16:14:08 2006 Ben Pfaff <blp@gnu.org>
1294 * value-labels.h: Remove unneeded dependency on variable.h.
1296 Sat May 6 15:58:36 2006 Ben Pfaff <blp@gnu.org>
1298 Get rid of `char *c' member in union value, for cleanliness.
1300 * value.h: (union value) Remove `c' member.
1302 Sat May 6 15:36:59 2006 Ben Pfaff <blp@gnu.org>
1304 Make dictionary compacting functions a little more general.
1306 * sys-file-writer.c (sfm_open_writer): Use
1307 dict_compacting_would_change().
1308 (does_dict_need_translation) Removed.
1310 Sat May 6 15:35:42 2006 Ben Pfaff <blp@gnu.org>
1312 Make dictionary compacting functions a little more general.
1314 * dictionary.c (dict_needs_compaction): Rename
1315 dict_compacting_would_shrink(). Update all callers.
1316 (dict_compacting_would_change) New function.
1318 Sat May 6 14:25:49 2006 Ben Pfaff <blp@gnu.org>
1320 * sys-file-writer.c: (does_dict_need_translation) Fix bug:
1321 inverted return value (!).
1323 Sat May 6 13:37:52 2006 Ben Pfaff <blp@gnu.org>
1325 Continue reforming procedure execution.
1327 * procedure.c: Search and replace "vfm" by "proc". Notably:
1328 (static var vfm_source) Rename proc_source. Update all
1330 (static var vfm_sink) Rename proc_sink. Update all references.
1332 Sat May 6 12:38:55 2006 Ben Pfaff <blp@gnu.org>
1334 Continue reforming procedure execution. In this phase, remove
1335 PROCESS IF, which was deprecated anyway and can be easily
1336 simulated with TEMPORARY followed by SELECT IF.
1338 * procedure.c: (open_active_file) Don't call
1339 add_process_if_trns().
1340 (discard_variables) Get rid of redundant call to
1341 proc_cancel_all_transformations().
1342 (add_process_if_trns) Removed.
1343 (process_if_trns_proc) Removed.
1344 (process_if_trns_free) Removed.
1346 Sat May 6 10:58:05 2006 Ben Pfaff <blp@gnu.org>
1348 Continue reforming procedure execution. In this phase, add
1349 `const' to the case passed to procedure()'s callback.
1351 Updated all users of procedure() as well.
1353 * procedure.c: (struct write_case_data) Add "const" to ccase
1354 parameter for case_func member.
1355 (procedure) Add "const" to ccase parameter for proc_func
1357 (multipass_case_func) Make ccase parameter const.
1358 (internal_procedure) Add "const" to ccase parameter for case_func
1360 (split_procedure_case_func) Make ccase parameter const.
1361 (multipass_split_case_func) Make ccase parameter const.
1363 Sat May 6 10:30:33 2006 Ben Pfaff <blp@gnu.org>
1365 Continue reforming procedure execution. In this phase, get rid of
1366 the output code for SPLIT FILE groups in procedure.c, which really
1367 shouldn't be doing any output. Move it into the individual
1368 procedures instead. This also adds some flexibility.
1370 Updated many users of procedure_with_splits() and
1371 multipass_procedure_with_splits() to call
1372 output_split_file_values() and to deal with increased use of
1375 * procedure.c: (struct split_aux_data) Add "const struct ccase *"
1376 parameter to begin_func member.
1377 (procedure_with_splits) Add "const struct ccase *" parameter to
1378 begin_func parameter. Make ccase parameter const in proc_func
1380 (split_procedure_case_func) Don't dump split file group. Pass
1382 (dump_splits) Moved to language/dictionary/split-file.c as
1383 output_split_file_values().
1384 (struct multipass_split_aux_data) Add "const struct ccase *"
1385 parameter to split_func member.
1386 (multipass_procedure_with_splits) Add "const struct ccase *"
1387 parameter to split_func parameter.
1388 (multipass_split_case_func) Save new SPLIT FILE case before
1390 (multipass_split_output) Pass saved SPLIT FILE case to split_func.
1392 Fri May 5 22:48:50 2006 Ben Pfaff <blp@gnu.org>
1394 Continue reforming procedure execution. Change
1395 internal_procedure() so that it calls open_active_file() and
1396 close_active_file(), which isolates most of the actual procedure
1399 * procedure.c: (struct write_case_data) Rename `proc_func' member
1400 to `case_func' and update all references.
1401 (procedure) Rewrite as one-line wrapper around
1402 internal_procedure().
1403 (struct multipass_aux_data) New.
1404 (multipass_callback) Renamed multipass_case_func(). Use struct
1405 multipass_aux_data as auxiliary data.
1406 (multipass_end_func) New function.
1407 (multipass_procedure) Rewrite as wrapper for internal_procedure()
1408 that uses multipass_case_func, multipass_end_func.
1409 (internal_procedure) Add `end_func' argument. Move optimization
1410 of trivial case in here. Move call to open_active_file() and
1411 close_active_file() in here. Now assert that vfm_source is
1413 (procedure_with_splits_callback) Rename
1414 split_procedure_case_func().
1415 (split_procedure_end_func) New function.
1416 (multipass_split_callback) Rename multipass_split_case_func.
1417 (multipass_split_end_func) New function.
1418 (discard_variables) No need to test for nonnull vfm_source.
1420 Fri May 5 21:34:02 2006 Ben Pfaff <blp@gnu.org>
1422 Continue reforming procedure execution. Get rid of unused member.
1424 * procedure.c: (struct write_case_data) Remove `cases_analyzed'
1426 (write_case) Don't increment cases_analyzed.
1428 Thu May 4 21:50:11 2006 Ben Pfaff <blp@gnu.org>
1430 Continue reforming procedure execution. In this phase, move
1431 procedure.c and procedure.h from src to src/data. Update
1432 makefiles and #includes accordingly.
1434 * procedure.c: Moved here from src/.
1436 * procedure.h: Moved here from src/.
1438 Wed May 3 22:42:12 2006 Ben Pfaff <blp@gnu.org>
1440 Continue reforming procedure execution. In this phase, get rid of
1441 many global variables, consolidating procedure execution in
1442 procedure.c. Encapsulate transformations in new "struct
1443 trns_chain". Also, change implementation of N OF CASES, FILTER,
1444 and PROCESS IF from special cases to transformations.
1446 * automake.mk: (src_data_libdata_a_SOURCES) Add transformations.c,
1449 * dictionary.c: (global variable default_dict) Move to
1452 * variable.h: (TRNS_*) Move to transformations.h.
1453 (struct transformation) Move to transformations.c.
1455 Thu May 4 13:47:06 WST 2006 John Darrington <john@darrington.wattle.id.au>
1457 * sys-file-reader.c: Fixed invalid read problems.
1459 Tue May 2 15:57:10 2006 Ben Pfaff <blp@gnu.org>
1461 * storage-stream.c: Add missing function comments.
1463 Tue May 2 15:50:21 2006 Ben Pfaff <blp@gnu.org>
1465 Continue reforming procedure execution. In this phase, add some
1466 new, needed functionality to storage-stream.
1468 * storage-stream.c: (storage_source_decapsulate) New function.
1470 Tue May 2 15:43:36 2006 Ben Pfaff <blp@gnu.org>
1472 * variable.c (width_to_bytes): Declarations must precede
1473 statements for C90 compliance.
1475 Tue May 2 10:42:05 WST 2006 John Darrington <john@darrington.wattle.id.au>
1477 * data-out.c, data-in.c, variable.c, variable.h: New functions
1478 copy_mangle and copy_demangle for reading/writing cases; emulates the
1479 way SPSS deals with strings > 255 bytes.
1481 * sys-file-reader.c sys-file-writer.c: Added support for Record 7,
1482 subtype 14 needed for strings longer than 255 bytes.
1484 * dictionary.c, format.def, value.c : Updated to use MAX_STRING
1485 instead of literal values. Also fixed some constness issues.
1487 * format.h: Constness
1489 * sfm-private.h: Renamed the case_size identifier, since I discovered
1490 that SPSS's respect for this variable is very nominal.
1492 Mon May 1 15:45:42 2006 Ben Pfaff <blp@gnu.org>
1494 Change case limit type from int to size_t.
1496 * dictionary.c: (struct dictionary) Change type of case_limit
1498 (dict_get_case_limit) Change return type.
1499 (dict_set_case_limit) Change parameter type.
1501 Wed Apr 26 20:01:19 2006 Ben Pfaff <blp@gnu.org>
1503 * variable.h: (struct variable) Rename `reinit' member as `leave'
1504 and invert sense. Fix up all references.
1506 Wed Apr 26 19:39:28 2006 Ben Pfaff <blp@gnu.org>
1508 Continue reforming procedure execution. In this phase, break
1509 procedure.c into multiple files.
1511 * automake.mk: (src_data_libdata_a_SOURCES) Add all the new files.
1513 * case-sink.c: New file.
1515 * case-sink.h: New file.
1517 * case-source.c: New file.
1519 * case-source.h: New file.
1521 * storage-stream.c: New file.
1523 * storage-stream.h: New file.
1525 Wed Apr 26 14:55:19 2006 Ben Pfaff <blp@gnu.org>
1527 * variable.h: (struct variable) Remove `init' member and all
1528 references to it from other files. It was initialized in several
1529 places, but nothing really ever used it for anything worthwhile.
1530 Thanks to Jason Stover for pointing out how confusing this
1533 Sun Apr 23 22:04:45 2006 Ben Pfaff <blp@gnu.org>
1535 Continue reforming error message support. In this phase, get rid
1536 of message "titles" and put the message text in `struct error'.
1537 Now `struct error' encapsulates a message more properly.
1539 * casefile.c: (io_error) Use err_msg() instead of err_vmsg().
1540 Format message ourselves.
1542 * data-in.c: (vdls_error) Ditto.
1544 * por-file-reader.c: (error) Ditto.
1546 * sys-file-reader.c: (corrupt_msg) Ditto.
1548 Sun Apr 16 18:49:51 2006 Ben Pfaff <blp@gnu.org>
1550 GNU standards require "file name" instead of "filename" in
1551 documentation. It's nice for our code to follow the convention
1554 * casefile.c: (struct casefile) Rename `filename' member to
1555 `file_name'. Updated all references.
1557 * file-name.c: [!unix] (struct file_identity) Rename
1558 normalized_filename member to normalized_file_name. Updated all
1561 Sun Apr 16 18:35:33 2006 Ben Pfaff <blp@gnu.org>
1563 We don't really support anything but Unix-like environments well,
1564 so we might as well de-obfuscate by writing directory and path
1565 separators explicitly.
1567 * file-name.h: (macro DIR_SEPARATOR) Removed. Changed all usages
1569 (macro PATH_SEPARATOR) Removed. Changed all usages to just ':'.
1570 (macro DIR_SEPARATOR_STRING) Removed. Changed all usages to just
1572 (macro PATH_SEPARATOR_STRING) Removed. Changed all usages to just
1575 Sun Apr 16 18:28:35 2006 Ben Pfaff <blp@gnu.org>
1577 GNU standards require "file name" instead of "filename" in
1578 documentation. It's nice for our code to follow the convention
1581 * filename.c: Rename to file-name.c.
1583 * filename.h: Rename to file-name.h. Update all inclusions.
1584 Update header guards.
1586 * automake.mk: Update file names.
1588 Sun Apr 16 16:42:47 2006 Ben Pfaff <blp@gnu.org>
1590 * filename.c: (fn_dirname) Renamed fn_dir_name(), all references
1592 (fn_basename) Removed (dead code).
1593 (fn_absolute_p) Renamed fn_is_absolute(), all references updated.
1594 (fn_special_p) Renamed fn_is_special(), all references updated.
1595 (fn_exists_p) Renamed fn_exists(), all references updated.
1597 Sun Apr 16 16:33:58 2006 Ben Pfaff <blp@gnu.org>
1599 * filename.c: (fn_tilde_expand) Rewrite for cleaner code.
1600 Also, now it only tilde-expands file names, not paths.
1601 (fn_search_path) Tilde-expand one directory at a time.
1603 Sun Apr 16 16:28:06 2006 Ben Pfaff <blp@gnu.org>
1605 * filename.c: (fn_search_path) rewrite for cleaner code. Also,
1606 get rid of non-Unixlike version of the code, which has probably
1608 (fn_prepend_dir) Removed (dead code).
1610 * filename.h: (macro DIR_SEPARATOR_STRING) New.
1611 (macro PATH_SEPARATOR_STRING) New.
1612 Sun Apr 16 16:05:28 2006 Ben Pfaff <blp@gnu.org>
1614 Continue reforming error message support. In this phase, we get
1615 rid of VM() and the other msg() support for "verbosity", replacing
1616 it by a new function verbose_msg().
1618 * filename.c: (fn_search_path) Use verbose_msg() instead of
1621 Sat Apr 15 19:53:19 2006 Ben Pfaff <blp@gnu.org>
1623 * sfm-private.h: Get rid of #defines after #error, which makes no
1626 Sat Apr 15 19:48:57 2006 Ben Pfaff <blp@gnu.org>
1628 Get rid of our own int32 type in favor of the standard int32_t
1631 * sfm-private.h: (int32 macro) Don't define this anymore. Do
1634 * sys-file-reader.c: Use int32_t instead of int32 throughout.
1636 * sys-file-writer.c: Use int32_t instead of int32 throughout.
1638 Sat Apr 15 19:36:47 2006 Ben Pfaff <blp@gnu.org>
1640 Remove ill-considered file routines that are no longer used.
1642 * filename.c: (fn_open_ext) Removed.
1643 (fn_close_ext) Removed.
1645 * filename.h: (struct file_ext) Removed.
1647 Mon Apr 3 13:22:39 2006 Ben Pfaff <blp@gnu.org>
1649 * variable.c (var_is_valid_name): Move declarations before code
1652 Tue Apr 4 15:28:40 WST 2006 John Darrington <john@darrington.wattle.id.au>
1654 * filename.ch (fn_interp_vars): Fixed small buglet.
1656 Tue Mar 28 13:47:16 WST 2006 John Darrington <john@darrington.wattle.id.au>
1658 * filename.[ch] (fn_interp_vars): Changed the signature and semantics
1659 so as to modify the string inline. Thus makeing it easier to
1660 destroy the results when no longer needed.
1662 2006-03-25 Jason Stover <jhs@math.gcsu.edu>
1664 * category.c (cat_stored_values_destroy): Fixed memory leak.
1666 Fri Mar 24 18:15:41 2006 Ben Pfaff <blp@gnu.org>
1668 Add some missing frees. Thanks to John Darrington for reporting
1671 * any-writer.c (any_writer_close): Free writer.
1673 * any-reader.c (any_reader_close): Free reader.
1675 Mon Mar 20 16:33:53 2006 Ben Pfaff <blp@gnu.org>
1677 * por-file-reader.c: (error) Mark as NO_RETURN.
1679 Sat Mar 11 15:06:07 WST 2006 John Darrington <john@darrington.wattle.id.au>
1681 * settings.c: Changed default value of scompress to true.
1683 Sat Mar 4 13:22:51 2006 Ben Pfaff <blp@gnu.org>
1685 * sfm-private.h: Include variable.h, to get SHORT_NAME_LEN.
1687 * value.h: Remove check on MAX_SHORT_STRING, which I don't think
1690 * variable.h: Move definition of SHORT_NAME_LEN, LONG_NAME_LEN
1691 here from pref.h.orig.
1693 Sat Mar 4 12:50:48 WST 2006 John Darrington <john@darrington.wattle.id.au>
1695 * sys-file-reader.c: Fixed bug reading compressed files.
1697 Thu Mar 2 08:40:33 WST 2006 John Darrington <john@darrington.wattle.id.au>
1699 * Numerous renames. See src/ChangeLog for details.
1701 * Moved files from src directory