1 2008-02-02 John Darrington <john@darrington.wattle.id.au>
3 * get-data.c (cmd_get_data): Support PSQL type.
5 2007-12-07 Ben Pfaff <blp@gnu.org>
9 * data-parser.c (data_parser_make_active_file): Fix case count
10 argument to casereader_create_sequential, which fixes data reading
12 Provided by John Darrington.
14 2007-12-04 Ben Pfaff <blp@gnu.org>
16 Move DATA LIST parsing into generic infrastructure, and generalize
17 it slightly. Then, use the same infrastructure to implement GET
20 * data-parser.c: New file.
22 * data-parser.h: New file.
24 * data-list.c (struct dls_var_spec): Removed.
25 (ll_to_dls_var_spec): Removed.
26 (enum dls_type): Removed.
27 (struct data_list_pgm): Rename struct data_list_trns. Remove
28 pool, specs, type, record_cnt, delims, skip_records, value_cnt
29 members. Add new `parser' member.
30 (cmd_data_list): Use data-parser infrastructure.
33 (dump_fixed_table): Removed.
34 (dump_free_table): Removed.
36 (read_from_data_list): Removed.
37 (read_from_data_list_fixed): Removed.
38 (read_from_data_list_free): Removed.
39 (read_from_data_list_list): Removed.
40 (data_list_trns_free): Rename arguments for clarity.
41 (data_list_trns_proc): Ditto.
42 (data_list_casereader_read): Removed.
43 (data_list_casereader_destroy): Removed.
44 (data_list_casereader_class): Removed.
46 * get-data.c (cmd_get_data): Support TXT type.
47 (set_type): New function.
48 (parse_get_txt): New function.
50 2007-12-04 Ben Pfaff <blp@gnu.org>
52 * placement-parser.c (parse_column): New function.
53 (parse_column_range): Add `base' argument. Update all callers.
55 2007-12-04 Ben Pfaff <blp@gnu.org>
57 Make GET DATA a separate command, instead of something invoked
60 * automake.mk: Remove get-data.h from sources.
62 * get-data.h: Removed.
64 * get-data.c (parse_get_data_command): Rename cmd_get_data.
66 * get.c (parse_read_command): No longer any need to check for DATA
69 2007-12-04 Ben Pfaff <blp@gnu.org>
71 * src/language/data-io/data-reader.c (struct dfm_reader): New
72 `file_size' member to support dfm_get_percent_read.
73 (dfm_open_reader): Initialize file_size.
74 (dfm_get_percent_read): New function.
76 2007-11-08 Ben Pfaff <blp@gnu.org>
78 Patch #6256: add support for binary, 360 file formats. Reviewed
81 * data-reader.c (struct dfm_reader): New member `block_left'.
82 (dfm_open_reader): Initialize block_left. For FH_MODE_TEXT, open
83 the file in text mode.
84 (read_error): New function.
85 (partial_record): New function.
86 (try_to_read_fully): New function.
87 (enum descriptor_type): New enum.
88 (read_descriptor_word): New function.
89 (corrupt_size): New function.
90 (read_size): New function.
91 (read_file_record): Implement new modes.
92 (read_record): Now take care of tracking line numbers here.
93 (dfm_reader_get_legacy_encoding): New function.
95 * data-writer.c (dfm_put_record): Implement new modes.
96 (dfm_writer_get_legacy_encoding): New function.
98 * file-handle.q: Parse new formats.
99 (cmd_file_handle): Set up new formats.
101 * print.c (struct print_trns): New member `encoding'.
102 (internal_cmd_print): Set encoding.
103 (print_trns_proc): Recode output data if necessary.
104 (flush_records): Recode leader byte.
106 2007-11-03 Ben Pfaff <blp@gnu.org>
108 Allow output files to overwrite input files (bug #21280).
110 * data-list.c (cmd_data_list): Manage file handle reference
113 * data-reader.c (struct dfm_reader): Add `lock' member.
114 (dfm_close_reader): Simplify, as reference counting is now
115 separate from locking.
116 (dfm_open_reader): Lock file.
118 * data-writer.c (struct dfm_writer): Add fh_lock, replace_file
120 (dfm_open_writer): Lock file and prepare for replacement.
121 (dfm_close_writer): Unlock file and replace it.
123 * file-handle.q (cmd_close_file_handle): Use fh_unname.
124 (fh_parse): Don't distinguish existing handles for a given file
125 name from new ones. Manage file handle reference counts.
127 * get.c (parse_read_command): Manage file handle reference counts.
128 (parse_write_command): Ditto.
129 (mtf_close_all_files): Ditto.
131 * inpt-pgm.c (cmd_reread): Manage file handle reference counts.
133 * print-space.c (cmd_print_space): Manage file handle reference
136 * print.c (internal_cmd_print): Manage file handle reference
139 2007-11-03 John Darrington <john@darrington.wattle.id.au>
141 * get.c: Add GET DATA command variant.
143 * get-data.c get-data.h (new files): Added support for
144 GET DATA /TYPE='gnm' command.
146 2007-09-23 Ben Pfaff <blp@gnu.org>
148 Bug #21111. Reviewed by John Darrington.
150 * data-list.c (data_list_trns_proc): Properly set retval when END
151 subcommand is in use.
152 (cmd_data_list): Don't allow END subcommand to be used with DATA
155 2007-09-12 Ben Pfaff <blp@gnu.org>
157 * get.c (get_translate_case): Change input case parameter from
158 const struct ccase * to struct ccase *, to match change in
159 casereader and casewriter translators. Destroy input case, to fix
162 2007-08-12 Ben Pfaff <blp@gnu.org>
164 * get.c (parse_read_command): Compact the values in the target
165 dictionary, to save space.
167 2007-08-12 Ben Pfaff <blp@gnu.org>
169 * get.c (struct case_map): Move into new file src/data/case-map.c.
170 (start_case_map): Ditto, and rename case_map_prepare_dict.
171 (finish_case_map): Ditto, and rename case_map_from_dict.
172 (map_case): Ditto, and rename case_map_execute.
173 (destroy_case_map): Ditto, and rename case_map_destroy.
174 (case_map_get_value_cnt): Ditto.
176 2007-08-12 Ben Pfaff <blp@gnu.org>
178 * get.c (case_map_get_value_cnt): New function.
180 2007-07-25 Ben Pfaff <blp@gnu.org>
183 * data-list.c (read_from_data_list_fixed): Handle multi-record
186 2007-07-11 Ben Pfaff <blp@gnu.org>
188 * get.c (map_case): Create destination case instead of leaving it
189 undefined. Fixes bug #20285.
190 Reviewed by John Darrington.
192 2007-06-06 Ben Pfaff <blp@gnu.org>
194 * get.c: Essentially rewrite MATCH FILES to support FIRST and
197 2007-06-06 Ben Pfaff <blp@gnu.org>
199 Adapt case sources, sinks, and clients of procedure code to the
202 * data-list.c: Make DATA LIST into a casereader.
204 * get.c: Change GET, IMPORT, SAVE, EXPORT to use casereaders,
207 * inpt-pgm.c: Use caseinit code. Turn INPUT PROGRAM into a
210 * list.q: Adapt to new procedure code.
212 2007-05-06 Ben Pfaff <blp@gnu.org>
214 Abstract the documents within a dictionary a little better.
215 Thanks to John Darrington for suggestion, initial version, and
218 * get.c (mtf_merge_dictionary): Simplify creating merged document.
220 * sys-file-info.c (display_documents): Use new
221 dict_get_document_line_cnt and dict_get_document_line functions.
223 Thu Feb 1 16:56:02 2007 Ben Pfaff <blp@gnu.org>
225 * file-handle.q (fh_parse): Update to new fh_create_file
228 Sat Dec 16 22:16:18 2006 Ben Pfaff <blp@gnu.org>
230 Make it possible to pull cases from the active file with a
231 function call, instead of requiring indirection through a callback
234 * automake.mk: Removed matrix-data.c.
236 * matrix-data.c: Removed.
238 * data-list.c (data_list_source_read): Conform with new
239 case_source_class interface.
240 (data_list_source_destroy): Ditto.
242 * get.c (case_reader_source_class): Ditto.
243 (case_reader_source_destroy): Ditto.
244 (parse_output_proc): Take advantage of new procedure interface.
245 (output_proc): Removed.
246 (struct mtf_file): Add "struct ccase *" member to allow use of new
248 (cmd_match_files): Take advantage of new procedure interface.
249 (mtf_processing_finish): Removed.
250 (mtf_read_nonactive_records): Renamed mtf_read_records. Now reads
251 from every file, without any exception for the active file.
252 (mtf_compare_BY_values): Simplify for new interface.
253 (mtf_processing): Simplify for new interface.
255 * inpt-pgm.c (is_valid_state): New function.
256 (input_program_source_read): Conform with new case_source_class
258 (input_program_source_destroy): Ditto.
259 (end_case_trns_proc): Now just needs to return TRNS_END_CASE.
261 Sat Dec 9 18:43:34 2006 Ben Pfaff <blp@gnu.org>
263 * list.q (cmd_list): Use new var_create, var_destroy functions.
265 Thu Nov 30 21:51:58 2006 Ben Pfaff <blp@gnu.org>
267 * inpt-pgm.c (cmd_reread): Always return error code upon detecting
268 syntax error. Fixes bug #18419. Thanks to John Darrington for
271 Sun Nov 19 09:17:45 2006 Ben Pfaff <blp@gnu.org>
273 * data-list.c (parse_free): Follow documented (but odd) rule that
274 N format is treated as F format for free-field input.
276 * data-reader.c (read_file_record): Drop new-line character from
277 input text lines. This is symmetrical with the recently changed
278 dfm_put_record semantics.
280 Thu Nov 2 20:56:03 2006 Ben Pfaff <blp@gnu.org>
282 Implement SKIP keyword on DATA LIST. Fixes bug #17099.
284 * data-list.c: (struct data_list_pgm) Add `skip_records' members.
285 (cmd_data_list) Set skip_records based on user input.
286 (data_list_source_read) Skip records requested by user.
288 Tue Oct 31 20:04:06 2006 Ben Pfaff <blp@gnu.org>
290 * placement-parser.c: (PRS_TYPE_T) Now that struct fmt_spec uses
291 an enum fmt_type for its type member, we can't depend on the
292 ability to put negative values into that member as out-of-band
293 values, because enum fmt_type might be an unsigned type. So use
294 values around SCHAR_MAX instead, because we know that SCHAR_MAX
295 will fit into any type, signed or unsigned, and there aren't
296 nearly that many format types.
297 (parse_var_placements) Add for_input parameter to specify whether
298 we're parsing input or output formats. Update all callers.
299 (fixed_parse_columns) Ditto.
300 (fixed_parse_fortran) Ditto.
302 Tue Oct 31 18:21:48 2006 Ben Pfaff <blp@gnu.org>
304 * print-space.c (print_space_trns_proc): Let dfm_put_record add
305 the new-line character, to match dfm_put_record change below.
307 Sat Oct 28 11:57:19 2006 Ben Pfaff <blp@gnu.org>
309 * data-writer.c (struct dfm_writer): Removed `bounce' member, and
310 all references to it.
311 (dfm_put_record) Change semantics so that it adds formatting
312 itself, such as new-line characters, instead of putting that
313 responsibility on the caller. Also, pad binary records with
314 spaces instead of zeros, for compatibility.
316 * print.c (struct prt_out_spec) New member `sysmis_as_spaces'.
317 (struct print_trns) Remove `omit_new_lines' and all references,
318 since dfm_put_record() is taking care of that. Add
320 (internal_cmd_print) Allow an empty set of data to print. Set
322 (parse_specs) Allow an empty set of data to print.
323 (parse_variable_argument) Only add space with PRINT command. Set
325 (print_trns_proc) Indent records if include_prefix is set, for
326 compatibility. Output SYSMIS as spaces if sysmis_as_spaces is
327 set. Put "1" in first column if PRINT EJECT is used with an
328 external output file.
329 (flush_records) Ditto.
331 Sat Oct 28 16:19:57 WST 2006 John Darrington <john@darrington.wattle.id.au>
333 * data-reader.c: Eliminated references to extern variable getl_buf
335 Sat Aug 5 08:25:07 2006 Ben Pfaff <blp@gnu.org>
337 Fix bug #17329 in REREAD parsing, reported by John Darrington.
339 * inpt-pgm.c (cmd_reread): Fix file handle parsing.
341 Mon Jul 31 10:32:31 2006 Ben Pfaff <blp@gnu.org>
343 * print.c (parse_specs): Allow a comma between specifications.
345 Sun Jul 16 19:57:10 2006 Ben Pfaff <blp@gnu.org>
347 * automake.mk: (src_language_data_io_libdata_io_a_SOURCE) Add
348 print-space.c, placement-parser.c, placement-parser.h.
350 * data-list.c: Basically rewrote the whole thing. Broke out a lot
351 of code into placement-parser.c. Code is much cleaner now.
353 * placement-parser.c: New file.
355 * placement-parser.h: New file.
357 * print.c: Basically rewrote the whole thing. Broke out PRINT
358 SPACE into print-space.c. Code is much cleaner now.
360 * print-space.c: New file.
362 Sat Jul 1 17:39:40 2006 Ben Pfaff <blp@gnu.org>
364 Fix bug #11612, "q2c documentation does not agree with code".
366 * list.q: Audit use of q2c "+" prefixes that indicate that a
367 command may appear multiple times.
369 Sat Jul 1 20:44:22 2006 Ben Pfaff <blp@gnu.org>
371 Fix bug #15786: System File Creation crashes if directoy is
374 * get.c (parse_write_command): Check that the any_writer open
377 Tue Jun 27 22:44:28 2006 Ben Pfaff <blp@gnu.org>
379 Fix regression in command name completion reported by John
380 Darrington. Now completion is again state-dependent and occurs
381 only on the first line of a command.
383 * inpt-pgm.c: (cmd_input_program) Reading of first token in
384 command moved into cmd_parse.
386 Fri Jun 9 13:56:00 2006 Ben Pfaff <blp@gnu.org>
388 Reform string library.
390 * matrix-data.c (context): Use dynamic string.
391 (another_token) Deal with changed dfm_get_record() interface.
395 * data-list.c (struct data_list_pgm) Delete delims, delim_cnt
396 members, replacing them by struct string delims. Update all
397 references to use struct string functions.
398 (cut_field) Change interface to avoid needing "end_blank", by
399 getting the data-reader to remember that state for us. Change
400 internals to use substring. Update both callers.
402 * data-reader.c (read_file_record): Use ds_read_stream().
403 (dfm_get_record) Change interface to return substring. Updated
405 (dfm_expand_tabs) Use ds_find_char(). Now maintain position
406 relative to end-of-line. Use ds_swap().
407 (dfm_reread_record) Don't limit position by line length.
408 (dfm_column_start) Make parameter const.
409 (dfm_columns_past_end) New function.
410 (dfm_get_column) New function.
412 Thu May 25 18:26:26 WST 2006 John Darrington <john@darrington.wattle.id.au>
414 * print.c (print_trns_free): Made the code agree with the comment, by
415 not freeing PRT. Has the side effect that the command no longer
416 crashes on invalid syntax.
418 Tue May 9 20:55:46 2006 Ben Pfaff <blp@gnu.org>
420 * get.c (cmd_match_files): Fix memory leak replacing default_dict.
422 Sat May 6 22:25:09 2006 Ben Pfaff <blp@gnu.org>
426 * list.q (write_fallback_headers): (write_fallback_headers)
427 Properly record width of leader and pass it to write_varname().
429 Sat May 6 19:03:13 2006 Ben Pfaff <blp@gnu.org>
431 * get.c: (mtf_merge_dictionary) Fix value label memory leak.
433 Sat May 6 13:51:16 2006 Ben Pfaff <blp@gnu.org>
435 Use a casefile, instead of a case sink, for MATCH FILES output.
436 It's more straightforward, although it has the same effect.
438 * get.c: (struct mtf_proc) Replace `sink' case sink member by
439 `output' casefile member.
440 (cmd_match_files) Work with casefile instead of sink.
441 (mtf_processing) Add case to casefile instead of sink.
443 Sat May 6 10:43:07 2006 Ben Pfaff <blp@gnu.org>
445 Continue reforming procedure execution. In this phase, get rid of
446 the output code for SPLIT FILE groups in procedure.c, which really
447 shouldn't be doing any output. Move it into the individual
448 procedures instead. This also adds some flexibility.
450 * list.q (write_all_headers): Call output_split_file_values().
452 Wed May 3 23:00:17 2006 Ben Pfaff <blp@gnu.org>
454 Continue reforming procedure execution. In this phase, get rid of
455 many global variables, consolidating procedure execution in
456 procedure.c. Encapsulate transformations in new "struct
457 trns_chain". Also, change implementation of N OF CASES, FILTER,
458 and PROCESS IF from special cases to transformations.
460 * data-list.c: (data_list_trns_proc) Return TRNS_END_FILE at end
461 of file. (Why didn't we do this before?)
462 (cmd_match_files) Direct procedure output to null sink.
463 Use discard_variables() instead of indirect version.
465 * inpt-pgm.c: Use transformation chain.
466 (struct input_program_pgm) Add trns_chain member.
467 (cmd_input_program) Initialize trns_chain member and capture
468 transformations with proc_capture_transformations().
469 (input_program_source_read) Use trns_chain_execute().
470 (destroy_input_program) Destroy input chain.
472 Tue May 2 10:39:56 WST 2006 John Darrington <john@darrington.wattle.id.au>
474 * list.q Changed from using fixed length char buffers to struct
475 string so that any length variables can be used.
477 Mon May 1 18:21:19 2006 Ben Pfaff <blp@gnu.org>
479 Further clean up the CMD_* command result codes.
481 * (enum cmd_result_extensions) New. Add CMD_END_INPUT_PROGRAM and
482 CMD_END_CASE result codes.
483 (struct input_program_pgm) Added case_nr, write_case, wc_data
484 members for use by END CASE transformation.
485 (emit_END_CASE) New function.
486 (cmd_input_program) Interpret CMD_END_CASE by outputting an END
487 CASE transformation. If none is output by the input program
488 itself, add one automatically at the end. Change lack of
489 variables from warning to error.
490 (cmd_end_input_program) Return CMD_END_INPUT_PROGRAM instead of
492 (input_program_source_read) No longer any need to special-case END
493 CASE. Handle TRNS_DROP_CASE properly. Initialize new members in
494 inp for use by END CASE transformation.
495 (destroy_input_program) New function.
496 (input_program_source_destroy) Just call destroy_input_program().
497 (cmd_end_case) Just return CMD_END_CASE.
498 (end_case_trns_proc) No longer a stub handled by
499 input_program_source_read(). Actually output the case and
500 increment the case number.
502 Mon May 1 16:06:30 2006 Ben Pfaff <blp@gnu.org>
504 Remove vestiges of REPEATING DATA support.
506 * data-list.c: (struct rpd_num_or_var) Removed.
507 (struct repeating_data_trns) Removed.
508 (cmd_repeating_data) Removed.
509 (find_variable_input_spec) Removed.
510 (parse_num_or_var) Removed.
511 (parse_repeating_data) Removed.
512 (realize_value) Removed.
513 (struct rpd_parse_info) Removed.
514 (rpd_parse_record) Removed.
515 (repeating_data_trns_proc) Removed.
516 (repeating_data_trns_free) Removed.
517 (repeating_data_set_write_case) Removed.
520 * inpt-pgm.c: (input_program_source_read) Don't deal with
523 * data-list.h: Removed.
525 * automake.mk (src_language_data_io_libdata_io_a_SOURCES): Removed
528 Mon May 1 15:58:28 2006 Ben Pfaff <blp@gnu.org>
530 Remove vestiges of FILE TYPE support.
532 * data-list.c: (cmd_data_list) Don't check for FILE TYPE.
533 (cmd_repeating_data) Ditto.
535 * automake.mk (src_language_data_io_libdata_io_a_SOURCES): Remove
536 file-type.c, file-type.h.
538 * file-type.c: Removed.
540 * file-type.h: Removed.
542 Wed Apr 26 13:16:28 2006 Ben Pfaff <blp@gnu.org>
544 Improve the way we handle the various parsing "states". Until now
545 we've hard-coded the state transitions in the command definition
546 file, but that's error-prone and, worse, it's redundant--we can
547 figure out what state we're in anyhow. We can cleanly handle
548 INPUT PROGRAM and FILE TYPE with a nested command-processing loop.
550 * data-list.c: (cmd_data_list) Use in_file_type() or
551 in_input_program() in place of case_source_is_class() or
552 case_source_is_complex().
554 * file-type.c: NB: Not really fixed except minimally to compile,
555 because it doesn't work anyway.
556 (in_file_type) New function.
557 (cmd_record_type) No need to check that we're in FILE TYPE.
558 (cmd_end_file_type) Ditto.
559 (var file_type_source_class) Make static.
561 * get.c: (cmd_match_files) Check vfm_source instead of pgm_state.
563 * inpt-pgm.c: (in_input_program) New function.
564 (cmd_input_program) Rewrite to include nested command processing
566 (cmd_end_input_program) Just return CMD_END_SUBLOOP.
567 (var input_program_source_class) Make static.
568 (cmd_end_case) No need to check that we're in INPUT PROGRAM.
569 (cmd_end_file) Ditto.
571 * automake.mk (src_language_data_io_libdata_io_a_SOURCES): Add
572 file-type.h, inpt-pgm.h.
574 * file-type.h: New file.
576 * inpt-pgm.h: New file.
578 Tue Apr 25 13:11:55 2006 Ben Pfaff <blp@gnu.org>
580 * print.c: Don't special-case MS-DOS line terminators.
581 (macro LINE_END_WIDTH) Removed.
582 (alloc_line) Line ends are 1 byte.
583 (print_trns_proc) Just output \n for line end.
585 Sun Apr 23 22:05:58 2006 Ben Pfaff <blp@gnu.org>
587 Continue reforming error message support. In this phase, get rid
588 of message "titles" and put the message text in `struct error'.
589 Now `struct error' encapsulates a message more properly.
591 * data-list.c: (macro RPD_ERR) Removed.
592 (rpd_msg) New function. Updated all references to tmsg() to call
593 this function instead.
595 Sat Apr 15 19:38:13 2006 Ben Pfaff <blp@gnu.org>
597 Remove last users of struct file_ext so we can get rid of it
600 * data-reader.c: (struct dfm_reader) Change file member from
601 struct file_ext to FILE *. Updated all references.
602 (dfm_close_reader) Close file with fn_close() instead of
603 fn_close_ext(). Also, make a copy of the file name from the file
604 handle before closing it, because we can't extract it after we
606 (dfm_open_reader) Open file with fn_open() instead of
609 * data-writer.c: (struct dfm_writer) Change file member
610 struct file_ext to FILE *. Updated all references.
611 (dfm_close_writer) Close file with fn_close() instead of
612 fn_close_ext(). Also, make a copy of the file name from the file
613 handle before closing it, because we can't extract it after we
615 (dfm_open_writer) Open file with fn_open() instead of
618 Sat Apr 15 18:00:32 2006 Ben Pfaff <blp@gnu.org>
620 * data-list.c: Add prototype to suppress warning for
621 cmd_repeating_data().
623 Thu Mar 2 08:40:33 WST 2006 John Darrington <john@darrington.wattle.id.au>
625 * Moved files from src directory