Debian Codegraph @

ChangeLog

raw
1 2005-08-28 Milan Zamazal <[email protected]>
2
3 * czech.scm (czech-int-select-contours): Use generalized contour
4 lists instead of pairs.
5 (czech-int-targets): Honor the above change.
6
7 2005-08-26 Milan Zamazal <[email protected]>
8
9 * czech.scm (czech-token-to-words): Handle R-caron in a special
10 way, to work around a Festival bug.
11
12 2005-06-30 Milan Zamazal <[email protected]>
13
14 * czech.scm (czech-basic-lts): If the word is empty or it contains
15 non-readable characters and czech-unknown-symbol-word is empty,
16 return nil.
17 (czech-lts): If the word or its transformation is empty, don't
18 syllabify it.
19
20 2005-04-13 Milan Zamazal <[email protected]>
21
22 * czech.scm (phoneSet): Typo in C's partner fixed; make CH* unvoiced.
23 (czech-adjust-segments): Typo in name reference fixed.
24
25 2005-04-12 Milan Zamazal <[email protected]>
26
27 * czech.scm (czech-adjust-segments): Don't modify unvoiced-voiced
28 if the voiced consonant is V; test for pauses properly.
29 (czech-orthography): Missing declarations of #_ and Vowel+# added.
30 (czech-word): Don't call check-adjust-phonetic-form.
31 (czech-pause): Call check-adjust-phonetic-form.
32
33 2005-04-12 Milan Zamazal <[email protected]>
34
35 * czech.scm (czech-diphones): Removed.
36 (czech-basic-lts): Don't apply czech-diphones rule set.
37 (czech-adjust-phonetic-form): New function.
38 (czech-word): Call it.
39 (phoneset): New feature `postnas'; feature `ctype' removed; `n'
40 value added to the `cvox' feature;
41 (czech-adjust-segments): New function.
42
43 2005-04-12 Milan Zamazal <[email protected]>
44
45 * czech.scm (czech-orthography): Handle Y in a similar way as I.
46
47 2005-04-11 Milan Zamazal <[email protected]>
48
49 * czech-lexicon.scm: Another diphtong removed.
50
51 2005-04-11 Milan Zamazal <[email protected]>
52
53 * czech-mbrola.scm (mbrola_cz2): Set the `gender' voice property.
54
55 2005-04-11 Milan Zamazal <[email protected]>
56
57 * czech.scm (czech-token-to-words): Handle r-caron in a special way, to
58 work around a Festival bug.
59
60 2005-04-07 Milan Zamazal <[email protected]>
61
62 * czech.scm: Phoneme durations changed, taken from voice-czech-ph.
63
64 2005-04-07 Milan Zamazal <[email protected]>
65
66 * README.cs: Some credits added.
67
68 2005-04-06 Milan Zamazal <[email protected]>
69
70 * czech.scm (czech-diphones): r~+# -> r~*+#.
71
72 2005-04-06 Milan Zamazal <[email protected]>
73
74 * czech-lexicon.scm: Don't use the `ou' phone.
75
76 2005-03-14 Milan Zamazal <[email protected]>
77
78 * New diphone set.
79
80 2005-02-17 Milan Zamazal <[email protected]>
81
82 * czech.scm (czech-next-punc, czech-prev-punc): Ensure that nil is
83 never returned.
84
85 2005-02-14 Milan Zamazal <[email protected]>
86
87 * czech.scm (czech-token-end-punc): Removed, no longer needed with
88 the right tokenization method.
89 (czech-phrase-cart-tree): Use token_end_punc instead of
90 czech-token-end-punc.
91
92 2005-02-14 Milan Zamazal <[email protected]>
93
94 * czech.scm (czech-prev-simple-punc): Don't crash on non-existent
95 utterance elements.
96
97 2005-02-14 Milan Zamazal <[email protected]>
98
99 * czech.scm (czech-word-stress-unit): Czech whether sylword has
100 any daughters before calling functions on them.
101
102 2005-02-12 Milan Zamazal <[email protected]>
103
104 * czech.scm (voice-czech-common): Set Token_Method to Token_Any.
105
106 2005-02-12 Milan Zamazal <[email protected]>
107
108 * czech.scm (czech-next-punc, czech-prev-punc): Punctuation
109 detection regexp fixed.
110
111 2005-02-12 Milan Zamazal <[email protected]>
112
113 * czech.scm (czech-identify-stress-units): Apply czech-downcase on
114 item.name before looking it in
115 czech-proper-single-syl-prepositions.
116
117 2005-02-10 Milan Zamazal <[email protected]>
118
119 * czech.scm (czech-next-token-punc, czech-word-stress-unit)
120 (czech-stress-unit-punc, czech-next-punc, czech-prev-punc)
121 (czech-next-simple-punc, czech-prev-simple-punc): New functions.
122 (czech-phrase-cart-tree): Completely reworked, according to
123 phonetic rules.
124 (czech-intonation-units): Small break separates intonation units
125 too.
126 (czech-pos): Put commas before conjunctions.
127 (czech-token-end-punc): New function.
128 (czech-adjust-phrase-breaks): New function.
129 (czech-word): Call it.
130 (czech-duration): Perform pause duration adjustments.
131
132 2005-02-10 Milan Zamazal <[email protected]>
133
134 * czech.scm (czech-phoneme-durations): Pause duration set to 0.1.
135 (czech-duration): Multiply czech-silence-durations by 10.
136
137 2005-02-10 Milan Zamazal <[email protected]>
138
139 * czech.scm (czech-intonation-units): Check for the very last
140 syllable directly in the Syllable relation; check for sentence
141 break there too.
142
143 2005-02-05 Milan Zamazal <[email protected]>
144
145 * czech.scm (czech-silence-duration-factors): Replaced by
146 czech-silence-durations.
147 (czech-phoneme-durations): Pause duration changed to 1.00.
148 (czech-duration): Compute durations in the given range.
149
150 2005-01-07 Milan Zamazal <[email protected]>
151
152 * czech.scm (czech-insert-filling-vowels): New variable.
153 (czech-translate-add-vowels): Honor it.
154
155 2005-01-07 Milan Zamazal <[email protected]>
156
157 * czech.scm (czech-translate-add-vowels): Don't try to insert into
158 the SylStructure relation if it is not present.
159
160 2004-07-19 Milan Zamazal <[email protected]>
161
162 * czech-lexicon.scm (=): Added.
163
164 2004-05-20 Milan Zamazal <[email protected]>
165
166 * czech.scm (czech-pos): Set `pos' of separate punctuation
167 characters to `nil'.
168
169 2004-05-20 Milan Zamazal <[email protected]>
170
171 * czech.scm (czech-intonation-units): Ensure the last syllable
172 always makes a new unit.
173
174 2004-05-03 Milan Zamazal <[email protected]>
175
176 * czech.scm (voice-czech-common): Reset postlex_rules_hooks.
177
178 2004-04-20 Milan Zamazal <[email protected]>
179
180 * czech.scm (voice-czech-common): Let after_synth_hooks is a list.
181
182 2004-04-07 Milan Zamazal <[email protected]>
183
184 * czech.scm (czech-translate-split-diphthongs): Insert the new
185 item into the SylStructure relation too.
186
187 2004-04-04 Milan Zamazal <[email protected]>
188
189 * czech.scm (czech-token-to-words): Don't crash in separator
190 testing when there's no punctuation.
191
192 2004-04-01 Milan Zamazal <[email protected]>
193
194 * czech.scm (czech-translate-add-vowels): Put the inserted
195 segments into the SylStructure too.
196
197 2004-04-01 Milan Zamazal <[email protected]>
198
199 * czech.scm (czech-min, czech-max): New functions.
200 (czech-duration): Don't fail on initial single-syllabic words with
201 more than 12 phonemes.
202
203 2004-03-31 Milan Zamazal <[email protected]>
204
205 * czech.scm (czech-token-to-words): Handle multiple punctuation as
206 a separator.
207
208 2004-03-30 Milan Zamazal <[email protected]>
209
210 * czech.scm (czech-token-to-words): Downcase `name' before
211 checking it in czech-multiword-abbrevs.
212 (czech-multiword-abbrevs): Honor the previous change.
213
214 2004-03-30 Milan Zamazal <[email protected]>
215
216 * czech.scm (czech-duration): Don't crash on initial empty words
217 carrying phrase breaks.
218
219 2004-03-29 Milan Zamazal <[email protected]>
220
221 * czech.scm (czech-lts): Handle empty word correctly.
222 (czech-duration): Don't look at word leafs, look at its
223 granddaughters.
224
225 2004-03-24 Milan Zamazal <[email protected]>
226
227 * czech.scm (czech-duration): When checking first word, start from
228 the Word relation, not from SylStructure.
229
230 2004-03-23 Milan Zamazal <[email protected]>
231
232 * czech.scm (lts.ruleset): Some vowel-vowel stroke insertion
233 disabled.
234
235 2004-03-23 Milan Zamazal <[email protected]>
236
237 * czech.scm (lts.ruleset): Typo in phoneme names fixed.
238 (czech-normalize): Sharp S added.
239
240 2004-03-23 Milan Zamazal <[email protected]>
241
242 * czech-lexicon.scm: `copyright' added.
243
244 * czech.scm (lts.ruleset): i-vowel handling fixed; `copy' removed.
245
246 2004-03-23 Milan Zamazal <[email protected]>
247
248 * czech-mbrola.scm (czech-mbrola-lts): `_' added.
249
250 2004-03-23 Milan Zamazal <[email protected]>
251
252 * czech-lexicon.scm: Typo in e: transcription fixed.
253
254 2004-03-22 Milan Zamazal <[email protected]>
255
256 * czech.scm (lts.ruleset): Vowel-vowel rules added.
257
258 2004-03-22 Milan Zamazal <[email protected]>
259
260 * czech-lexicon.scm: Lexicon entries fixed.
261
262 * dump-pronounciation.scm (dump-pronounciation): Dump
263 pronounciation in a form suitable for the lexicon.
264
265 * czech.scm (czech-multiword-abbrevs): `=' added.
266 (lts.ruleset): New special rules added.
267
268 2004-03-22 Milan Zamazal <[email protected]>
269
270 * czech.scm (czech): `syl' feature of _ and @ set to `-'.
271
272 2004-03-20 Milan Zamazal <[email protected]>
273
274 * gen-all-words.sh: New script.
275
276 2004-03-19 Milan Zamazal <[email protected]>
277
278 * Makefile: New file.
279
280 * czech-lexicon.scm: Everything except lexicon entries moved to
281 czech.scm.
282
283 * czech.scm: Load compiled lexicon.
284 (czech-lexicon-file): New variable.
285
286 2004-03-18 Milan Zamazal <[email protected]>
287
288 * czech-words: Updated.
289 * czech-words-all: Removed.
290
291 2004-03-18 Milan Zamazal <[email protected]>
292
293 * czech.scm (czech-item.has_feat, czech-token.unknown_word_name)
294 (czech-token.separator_word_name, czech-token.garbage_word_name)
295 (czech-token_to_words): Underscores in names replaced by hyphens.
296
297 2004-03-18 Milan Zamazal <[email protected]>
298
299 * czech.scm (czech-int-select-contours): First syllable adjustment
300 of final contours fixed.
301
302 2004-03-18 Milan Zamazal <[email protected]>
303
304 * czech.scm (czech-int-contour-tree): New contourtypes FA and FB.
305 (czech-int-contours): Likewise; one missing FA/3 contour added.
306
307 2004-03-18 Milan Zamazal <[email protected]>
308
309 * czech.scm (czech-int-select-contours): Redundant code removed.
310 (czech-int-contour-tree): Typo in M contour determination fixed;
311 test for preelement fixed.
312
313 2004-03-18 Milan Zamazal <[email protected]>
314
315 * czech.scm (czech-guess-pos): "jak" and derivatives added to
316 `question'; "*pak" question words added; question words removed
317 from `misc'.
318 (czech-pos): Handle "*pak" question words; cosmetic changes.
319
320 2004-03-17 Milan Zamazal <[email protected]>
321
322 * czech.scm (czech-identify-stress-units): Invalid counting in
323 many single-syllabic word processing fixed; single syllabic
324 initial processing fixed; ensure that in many single syllabic
325 chains no single syllabic element remains.
326 (czech-stress-unit-phonemes): Protection against preelement mark
327 added.
328
329 2004-03-17 Milan Zamazal <[email protected]>
330
331 * czech.scm (czech-token_to_words): Read "[ckm]m" as a measure
332 length only if the previous token is a number.
333
334 2004-03-17 Milan Zamazal <[email protected]>
335
336 * czech.scm (czech-pos-in-phrase-from, czech-pos-in-phrase-to):
337 New functions.
338 (czech-pos-last-in-phrase?): Use them.
339 (czech-item.feat*?): New function.
340 (czech-phrase-cart-tree): Standard phrase position feature
341 functions that don't work replaced by our own functions; number of
342 allowed in conjunction break guessing neighboring words reduced to
343 1.
344
345 2004-03-17 Milan Zamazal <[email protected]>
346
347 * czech.scm (czech-int-select-contours): Contour values of final
348 single syllabic units fixed.
349 (czech-item.feat?): Don't crash when item is nil.
350
351 2004-03-16 Milan Zamazal <[email protected]>
352
353 * czech.scm (czech-identify-stress-units): Don't append last-unit
354 if it is nil.
355
356 2004-03-16 Milan Zamazal <[email protected]>
357
358 * czech.scm (czech-translate-split-diphthongs): make-item
359 undefined.
360 (czech-translate-add-vowels): make-item replaced by insert-item.
361
362 2004-03-16 Milan Zamazal <[email protected]>
363
364 * czech.scm (czech-int-select-contours): Bug in long contour
365 handling fixed.
366
367 2004-03-16 Milan Zamazal <[email protected]>
368
369 * czech.scm (czech-duration): Compute duration factor from the
370 number of phonemes, not syllables.
371 (czech-duration): Attach special durations to initial
372 single-syllabic words.
373
374 2004-03-16 Milan Zamazal <[email protected]>
375
376 * czech.scm (czech-add-strokes): Pause testing condition fixed.
377
378 2004-03-16 Milan Zamazal <[email protected]>
379
380 * czech-debug.scm (czech-debug-print-relation): Exclude `NB'
381 feature values from the report; don't report empty feature sets.
382 (czech-debug-print-durations): The first occurance of this
383 function renamed to czech-debug-print-durfactors.
384 (czech-debug-print): Honor the previous change.
385 (czech-debug-print-units): Print contours multiplied by 100.
386
387 2004-03-15 Milan Zamazal <[email protected]>
388
389 * czech-debug.scm: New file.
390
391 2004-03-15 Milan Zamazal <[email protected]>
392
393 * czech.scm (czech-identify-stress-units): List handling in
394 initial single-syllabic word processing fixed.
395
396 2004-03-15 Milan Zamazal <[email protected]>
397
398 * czech.scm (czech-word): Don't call czech-add-keystrokes.
399 (czech-pause): Call czech-add-keystrokes.
400 (czech-add-strokes): Don't insert stroke if pause is present at
401 that place.
402
403 2004-03-15 Milan Zamazal <[email protected]>
404
405 * czech.scm (czech-pause-method): Renamed to czech-pause-breaks.
406 (czech-pause): New function.
407 (voice-czech-common): Honor the previous changes.
408
409 2004-03-15 Milan Zamazal <[email protected]>
410
411 * czech.scm (czech-duration): Honor Duration_Stretch.
412
413 2004-03-12 Milan Zamazal <[email protected]>
414
415 * czech.scm (czech-stress-units): When testing for an FF unit,
416 match any punctuation containing an FF termination character.
417
418 2004-03-12 Milan Zamazal <[email protected]>
419
420 * czech.scm (czech-pos-last-in-phrase?): Extended and improved.
421
422 2004-03-12 Milan Zamazal <[email protected]>
423
424 * czech.scm (czech-duration): Set dur_factor on non-syllabic
425 stress units too.
426
427 2004-03-12 Milan Zamazal <[email protected]>
428
429 * czech.scm (czech-phoneme-durations): Pause length increased.
430 (czech-silence-duration-factors): B and SB lengths decreased.
431
432 2004-03-12 Milan Zamazal <[email protected]>
433
434 * czech.scm (czech-phoneset-translation): Long vowels removed.
435
436 2004-03-12 Milan Zamazal <[email protected]>
437
438 * czech.scm (czech-int-targets): Don't ignore entries with
439 contourval 0 present.
440
441 2004-03-12 Milan Zamazal <[email protected]>
442
443 * czech-mbrola.scm (mbrola_cz2): Adjust
444 czech-after-analysis-hooks*; use Param.Set for setting the
445 synthesis method.
446 (czech-mbrola-lts): Diphthongs added.
447 (czech-mbrola-phoneme-durations): Updated to the new phoneset.
448
449 * czech.scm (czech-translate-split-diphthongs): New function.
450 (czech-add-segments): Renamed to czech-translate-add-vowels; run
451 only when the current language is Czech.
452 (czech-phone-adjustment): Renamed to czech-translate-phonemes.
453 (czech-after-analysis-hooks): Honor previous changes; turned into
454 defvar.
455 (czech-phoneme-durations): Diphthong lengths increased.
456
457 2004-03-11 Milan Zamazal <[email protected]>
458
459 * czech-lexicon.scm (lex.add.entry): Pronounciation of angle
460 brackets unified with czech-multiword-abbrevs.
461
462 * czech.scm (czech-token_to_words): Look into lexicon only after
463 punctuation is processed; handle punctuation-only tokens.
464
465 2004-03-10 Milan Zamazal <[email protected]>
466
467 * czech.scm (czech-token_to_words): Assign `sym' pos to single
468 letter word expansions.
469
470 2004-03-10 Milan Zamazal <[email protected]>
471
472 * czech.scm (czech-identify-stress-units): Paren typo fixed;
473 append last-unit to the result; invalid nth_cdr index fixed in
474 final single-syllabic processing.
475
476 2004-03-10 Milan Zamazal <[email protected]>
477
478 * czech.scm (czech-stress-units): Handle multiple character
479 punctuation.
480 (czech-yes-no-question): Likewise.
481
482 2004-03-10 Milan Zamazal <[email protected]>
483
484 * czech.scm (czech-phrase-cart-tree): Put B before left
485 parenthesis; handle multiple character punctuation.
486
487 2004-03-10 Milan Zamazal <[email protected]>
488
489 * czech.scm (czech-phrase-cart-tree): Dash handling fixed.
490
491 2004-03-10 Milan Zamazal <[email protected]>
492
493 * czech.scm (czech-token.punctuation): `-' added.
494 (czech-token_to_words): Don't handle dashes; don't throw away
495 dashes inside words.
496
497 2004-03-10 Milan Zamazal <[email protected]>
498
499 * czech-lexicon.scm: Various kinds of braces added for `punc'.
500
501 2004-03-09 Milan Zamazal <[email protected]>
502
503 * czech.scm (czech-int-targets): Apply the new contour value
504 conversion also when contourval is a list and fix the order of
505 values in such a case.
506 (czech-int-select-contours): Coding typo in parenthesis fixed for
507 long contours.
508
509 2004-03-08 Milan Zamazal <[email protected]>
510
511 * czech.scm (czech-int-targets): Use int_general_params, not
512 czech-int-simple-params; value computation fixed.
513 (czech-int-contours): Values converted to from +-1 to +-0.
514
515 2004-03-05 Milan Zamazal <[email protected]>
516
517 * czech.scm (czech-add-segments): Don't perform schwa insertion.
518
519 2004-03-05 Milan Zamazal <[email protected]>
520
521 * czech.scm (czech-guess-pos): New words added to question.
522
523 2004-03-05 Milan Zamazal <[email protected]>
524
525 * czech.scm (czech-identify-stress-units, czech-add-strokes): Use
526 POS information stored in word features.
527
528 2004-03-05 Milan Zamazal <[email protected]>
529
530 * czech.scm (czech-pos): Check all words against czech-guess-pos.
531
532 2004-03-05 Milan Zamazal <[email protected]>
533
534 * czech.scm (czech-downcase, czech-word-pos?): New functions.
535 (czech-identify-stress-units, czech-pos, czech-add-strokes): Use
536 them.
537
538 2004-03-05 Milan Zamazal <[email protected]>
539
540 * czech.scm (czech-int-contour-tree): New variable.
541 (czech-int-select-contours): Use it; preelements contour handling
542 fixed.
543
544 2004-03-04 Milan Zamazal <[email protected]>
545
546 * czech.scm (czech-randomize): New variable.
547 (czech-rand): Honor it.
548
549 2004-03-04 Milan Zamazal <[email protected]>
550
551 * czech.scm (czech-stress-units): Put last unit under type F
552 instead of FF if it is not followed by punctuation.
553
554 2004-03-04 Milan Zamazal <[email protected]>
555
556 * czech.scm (czech-duration-random-factor): New variable.
557 (czech-duration): Randomize durations.
558
559 2004-03-04 Milan Zamazal <[email protected]>
560
561 * czech.scm (czech-int-simple-params): f0_std set to 10.
562 (czech-int-targets): Honor f0_std in some way.
563
564 2004-03-04 Milan Zamazal <[email protected]>
565
566 * czech.scm (czech-yes-no-question): Implementated.
567 (czech-guess-pos): `question' added.
568
569 2004-03-04 Milan Zamazal <[email protected]>
570
571 * czech.scm (czech-int-countours): New variable.
572 (czech-intonation-targets): Completely rewritten and renamed to
573 czech-int-targets.
574 (czech-int-select-contours): New function.
575 (voice-czech-common): Set it as Int_Method.
576 (czech-stress-units): Handle question types and F*-1 positions.
577 (czech-yes-no-question): New function.
578 (czech-syllable-kernels): New function.
579 (czech-stress-unit-phonemes): Return items, not their names.
580 (czech-unit-syllable-count, czech-stress-unit-phonemes): Accept
581 units themselves, not only lists.
582 (czech-duration): Take advantage of the previous change.
583 (defPhoneSet): New phone feature `syl'.
584 (czech-syllabic-vocals, czech-syllabic-consonants): Removed.
585 (czech-item.feat?): New function.
586 (czech-rand-range): New variable.
587 (czech-rand): New function.
588 (czech-random-choice): Use it.
589
590 2004-03-02 Milan Zamazal <[email protected]>
591
592 * czech-mbrola.scm (mbrola_cz2): Set czech-after-analysis-hooks*.
593
594 * czech.scm (czech-after-analysis-hooks): New variable.
595 (czech-reset-parameters): Set new variable
596 czech-after-analysis-hooks*.
597 (voice-czech-common): Use czech-after-analysis-hooks*.
598
599 2004-03-02 Milan Zamazal <[email protected]>
600
601 * czech.scm (voice-czech-common): Call czech-add-segments in
602 after_analysis_hooks, not postlex_rules_hooks.
603 (czech-add-segments): Adjust `end' features.
604 (czech-phoneme-durations): Don't consider segment adjustments.
605 (czech-add-segments, czech-phone-adjustment): Return utterance.
606
607 2004-02-28 Milan Zamazal <[email protected]>
608
609 * czech.scm (czech-duration-cart-tree): Removed.
610 (czech-duration): New function.
611 (voice-czech-common): Set it as the duration method.
612 (czech-silence-duration-factors)
613 (czech-stress-duration-factors): New variables.
614 (czech-stress-units): Create IntStress relation; don't put
615 intonation units to StressUnit, put segments there.
616
617 2004-02-28 Milan Zamazal <[email protected]>
618
619 * czech.scm (czech-add-strokes): New function.
620 (czech-word): Call it.
621 (czech-add-segments): Don't insert strokes.
622
623 2004-02-28 Milan Zamazal <[email protected]>
624
625 * czech.scm (czech-word): czech-add-segments moved to
626 postlex_rules_hooks.
627
628 2004-02-28 Milan Zamazal <[email protected]>
629
630 * czech.scm (czech-pos): Final word in phrase detection fixed and
631 other fixes.
632 (czech-pos-last-in-phrase?): New function.
633
634 2004-02-28 Milan Zamazal <[email protected]>
635
636 * czech.scm (czech-phone-adjustment): Typo in string-equal fixed.
637
638 2004-02-28 Milan Zamazal <[email protected]>
639
640 * czech.scm (czech-stress-units): Paren typo in unit insertion
641 fixed; final punctuation check fixed.
642 (czech-syllabic-vocals): Diphthongs added.
643 (czech-identify-stress-units): Test for units* end in the while
644 loop condition; use cycle instead of recursion in merge-n;
645 preelement placement fixed.
646 (czech-word): Debugging code removed.
647
648 2004-02-26 Milan Zamazal <[email protected]>
649
650 * czech-lexicon.scm (czech-multiword-abbrevs): New variable.
651
652 * czech.scm (czech-word): Renamed to czech-add-segments.
653 (czech-word, czech-stress-units, czech-intonation-units)
654 (czech-identify-stress-units: New functions.
655 (czech-token_to_words): Handle multiword abbreviations.
656 (czech-non-syllabic-prepositions)
657 (czech-proper-1-syllabic-prepositions, czech-special-final-words):
658 New variables.
659 (czech-guess-pos): "ku" added.
660 (czech-syllabic-vocals, czech-syllabic-consonants): New variables.
661 (czech-syllable-count, czech-unit-syllable-count)
662 (czech-stress-unit-phonemes): New functions.
663 (czech-random-choice): New function.
664
665 2004-02-26 Milan Zamazal <[email protected]>
666
667 * czech.scm (czech-phrase-cart-tree): Don't put breaks before
668 vowels.
669 (czech-word): Insert _ between vowels on word boundaries.
670
671 2004-02-26 Milan Zamazal <[email protected]>
672
673 * czech.scm (lts.ruleset): New phoneme _.
674 (czech-phoneme-durations): _ added.
675 (czech-phoneset-translation): Translate _ to #.
676
677 2004-02-26 Milan Zamazal <[email protected]>
678
679 * czech.scm (lts.ruleset): Unused symbol SchwaCons removed.
680
681 2004-02-26 Milan Zamazal <[email protected]>
682
683 (lex.add.entry): Entry syllabification for the "unknown" word
684 fixed.
685
686 2004-02-26 Milan Zamazal <[email protected]>
687
688 * czech.scm (czech-non-pause-words, czech-phrasify): Removed.
689 (voice-czech-common): Set Phrasify_Method to Classic_Phrasify.
690
691 2004-02-25 Milan Zamazal <[email protected]>
692
693 * czech.scm: Use Param.get and Param.set instead of Parameter.get
694 and Parameter.set.
695
696
697 2004-02-25 Milan Zamazal <[email protected]>
698
699 * czech.scm (czech-phrase-cart-tree): Put B before some no-comma
700 conjunctions.
701
702 2004-02-25 Milan Zamazal <[email protected]>
703
704 * czech.scm (czech-phrasify): New function.
705 (czech-pause-method): Long phrase breaking moved to
706 czech-phrasify.
707 (voice-czech-common): Set Phrasify_Method.
708
709 2004-02-25 Milan Zamazal <[email protected]>
710
711 * czech.scm (lts.ruleset): Vowel duplication and schwa insertion
712 retracted.
713 (czech-word): New function.
714 (voice-czech-common): Set Word_Method parameter.
715
716 2004-02-25 Milan Zamazal <[email protected]>
717
718 * czech-lexicon.scm: Moved to new syllabification.
719
720 2004-02-24 Milan Zamazal <[email protected]>
721
722 * czech.scm (czech-phoneme-durations): New phonemes included;
723 phoneme lengths adjusted considering diphone construction.
724
725 2004-02-24 Milan Zamazal <[email protected]>
726
727 * czech.scm (lts.ruleset): Completely reworked to match the new
728 diphone set.
729
730 2004-02-24 Milan Zamazal <[email protected]>
731
732 * czech.scm (czech-syllabify): Removed.
733 (czech-syllabify-phstress): Make the word as a single syllable.
734
735 2004-02-23 Milan Zamazal <[email protected]>
736
737 * czech.scm (czech-phoneset-translation): New variable.
738 (czech-reset-parameters): Use it as default.
739
740 2004-02-23 Milan Zamazal <[email protected]>
741
742 * czech.scm (defPhoneSet): Phonemes from literature added; phoneme
743 properties changed.
744
745 2004-02-23 Milan Zamazal <[email protected]>
746
747 * czech.scm (czech-accent-cart-tree): Return NONE unconditionally.
748 (czech-syllabify-phstress): Don't generate starting accent.
749
750 2004-02-23 Milan Zamazal <[email protected]>
751
752 * README.Czech.cs: New file.
753
754 2004-02-10 Milan Zamazal <[email protected]>
755
756 * czech.scm (czech-proclaim-voice): Introduce voice options,
757 support dialect and gender settings.
758
759 2004-02-03 Milan Zamazal <[email protected]>
760
761 * czech.scm (lts.ruleset): Infinite loop bug fixed; special words
762 moved to lexicon.
763
764 2004-02-03 Milan Zamazal <[email protected]>
765
766 * czech-lexicon.scm: New file.
767
768 2004-02-02 Milan Zamazal <[email protected]>
769
770 * czech-mbrola.scm (mbrola_cz2): Set czech-volume-scale*.
771
772 2004-01-30 Milan Zamazal <[email protected]>
773
774 * czech.scm (czech-int-lr-params): Removed, use
775 czech-int-simple-params instead.
776 (czech-reset-parameters, voice-czech-common): Honor the previous
777 change.
778 (czech-int-simple-params): Values changed to reasonable numbers.
779 (voice-czech-common): Switch to general intonation method.
780 (czech-intonation-targets): New function.
781
782 2004-01-30 Milan Zamazal <[email protected]>
783
784 * czech.scm (czech-duration-cart-tree): Use other means than the
785 undocumented feature syl_break for determining clause
786 initials/finals; factor of stressed initials/finals reduced to
787 1.3.
788
789 2004-01-29 Milan Zamazal <[email protected]>
790
791 * czech.scm (czech-syllabify-phstress, czech-syllabify): New
792 functions.
793 (czech-lts): Use it instead of lex.syllabify.phstress.
794
795 2004-01-29 Milan Zamazal <[email protected]>
796
797 * czech.scm (czech-int-lr-params): Changed to respect the English
798 intonation tree parameters and an experimental Czech voice.
799
800 2004-01-29 Milan Zamazal <[email protected]>
801
802 * czech.scm (czech-duration-tree): New variable.
803 (voice-czech-common): Duration method changed to Tree_ZScores.
804
805 2004-01-29 Milan Zamazal <[email protected]>
806
807 * czech.scm (czech-volume-scale, czech-volume-scale*): New variables.
808 (czech-adjust-volume): New function.
809 (czech-reset-parameters): Reset czech-volume-scale*.
810 (voice-czech-common): Set after_synth_hooks.
811
812 2004-01-29 Milan Zamazal <[email protected]>
813
814 * czech.scm (czech-non-pause-words): New function.
815 (czech-pause-method): Insert pauses into long non-breaking
816 phrases.
817
818 2004-01-29 Milan Zamazal <[email protected]>
819
820 * czech.scm (defPhoneSet, PhoneSet.silences)
821 (czech-phoneme-durations): ## removed.
822 (czech-phrase-cart-tree): Use defvar instead of set; SB
823 introduced.
824 (czech-phone-adjustment): ## support removed.
825 (czech-pause-method): Handle SB instead of generating ##.
826
827 2004-01-28 Milan Zamazal <[email protected]>
828
829 * czech.scm (czech-f2b-f0-lr-end): Renamed to czech-f0-lr-end.
830
831 2004-01-15 Milan Zamazal <[email protected]>
832
833 * czech-unisyn.scm: Completely rewritten.
834 * czech-mbrola.scm (czech-mbrola-init, czech-mbrola-description):
835 Removed.
836 * czech.scm (czech-description, czech-parameter): Removed.
837 (czech-phoneset-translation*): New variable.
838 (czech-reset-parameters): Reset synthesis-method and
839 czech-phoneset-translation.
840 (czech-phone-adjustment, voice-czech-common): Honor the previous
841 changes.
842 (voice-czech-common): Don't initialize synthesizer.
843
844 * czech.scm (czech): Starting dz and dz~ fix in the ruleset.
845
846 2004-01-07 Milan Zamazal <[email protected]>
847
848 * czech.scm (czech-reset-parameters): New function.
849 (czech-proclaim-voice): Call it.
850 (czech-guess_pos): Renamed to czech-guess-pos.
851 (czech-phrase_cart_tree): Renamed to czech-phrase-cart-tree.
852 (czech-int_simple_params, czech-int_lr_params)
853 (czech-accent_cart_tree, czech-int-tone-cart-tree)
854 (czech-f0-lr-start, czech-f0-lr-mid, czech-f2b-f0-lr-end)
855 (czech-phoneme-durations): Likewise.
856 (czech-parameter): Use czech-description* instead of
857 czech-description.
858 (voice-czech-common): Use parameter variables.
859
860 * czech-mbrola.scm (czech-mbrola-lts): New ruleset.
861 (mbrola_cz2): Enable it.
862 * czech.scm (czech-basic-lts): New function.
863 (czech-lts-extra-rules): New variable.
864 (czech-lts): Apply czech-lts-extra-rules.
865 (voice-czech-common): Initialize czech-lts-extra-rules.
866
867 2003-12-31 Milan Zamazal <[email protected]>
868
869 * czech.scm (czech-phoneme_durations): `e' and `e:' durations
870 increased.
871
872 2003-12-22 Milan Zamazal <[email protected]>
873
874 * czech.scm (czech-proclaim-voice): Set current voice. Voice name
875 in voice proclamation fixed.
876
877 * czech-mbrola.scm (czech-mbrola-phoneme_durations): New variable.
878 (mbrola_cz2): Set czech-phoneme_durations.
879 * czech.scm (czech-phoneme_durations): Completely changed.
880
881 2003-12-17 Milan Zamazal <[email protected]>
882
883 * czech-unisyn.scm: New file.
884 * czech-mbrola.scm: Require `czech'.
885 (voice_czech_mbrola_cz2): New function.
886 (czech_mbrola_cz2): Voice declared.
887 * czech.scm: Unisyn database declaration etc. moved to
888 czech-unisyn.scm.
889 (voice_czech): Renamed to voice-czech-common.
890 (czech-proclaim-voice): New macro.
891
892 * czech.scm (czech-int_lr_params): Means set to 105.
893
894 2003-12-01 Milan Zamazal <[email protected]>
895
896 * czech.scm (czech): Phoneset made compatible with the
897 radio_phones phoneset.
898 (czech-phoneme_durations): Changed to defvar. i: shortened, u:
899 lenghtened.
900 (lex.add.entry): New words added.
901 (czech-int_tone_cart_tree, czech-int_lr_params): New variables.
902 (voice_czech): Use Intonation_Tree intonation method.
903 (czech-int_simple_params, czech-accent_cart_tree): Changed to
904 defvar.
905 (czech-f0_lr_start, czech-f0_lr_mid, czech-f2b_f0_lr_end): New
906 variables.
907 (voice_czech): Use them.
908
909 2003-11-28 Milan Zamazal <[email protected]>
910
911 * czech.scm (lex.add.entry): m/s and km/h added.
912 (czech-token_to_words): Check for presence of the token name in
913 lexicon at the proper place.
914 (czech-prepend-numprefix, czech-number*): New functions.
915 (czech-token_to_words): Support spaced numbers.
916 (czech-number-from-digits): Don't read thousands etc. when they
917 are zero.
918 (czech-item.has_feat): New function.
919 (czech-prepend-numprefix, czech-token_to_words, czech-pos): Use
920 it.
921 (czech-token_to_words): Time and ratio support added.
922 (czech-number@): New function.
923 (czech-downcase): Renamed to czech-normalize; some German
924 characters added.
925 (czech-char-regexp, czech-chars): New variables.
926 (czech-tokenize-on-nonalphas, czech-token_to_words, czech-pos):
927 Use them.
928 (czech-token_to_words): Missing argument to string-match added.
929 (czech-phrase_cart_tree): Make semicolon cause BB, exclude quote
930 from the rules.
931 (czech-guess_pos): New variable.
932 (voice_czech): Set guess_pos.
933 (czech-phrase_cart_tree): Insert B before some conjunctions.
934 (voice_czech): Int_Method value fixed.
935 (czech-int_simple_params): f0_std reduced to 5.
936 (czech-accent_cart_tree): Completely rewritten.
937 (czech-int_simple_params): f0_mean increased to 100.
938 (lex.add.entry): "GNU" added.
939 (czech-all-same): New function.
940 (czech-token_to_words): Use it, for separator matching fix.
941
942 2003-11-27 Milan Zamazal <[email protected]>
943
944 * czech.scm (czech): proclaim_voice added.
945 (czech-token_to_words, lex.add.entry): Czech monetary support
946 added.
947 (czech-token.punctuation): Dash removed.
948 (czech-token_to_words): Handle dashes.
949 (czech-token_to_words): Try to identify more acronyms; apply
950 czech-downcase on them.
951 (czech-token_to_words): Don't include spaces in number regexps.
952 (czech-token_to_words): Support for numeric ranges added.
953 (czech-pos): Don't set `pos' feature if already set.
954 (czech-token_to_words): Split non-alphanumeric tokens to
955 characters; handle separators and long sequences of
956 non-alphanumeric characters.
957 (czech-token.separator_word_name)
958 (czech-token.garbage_word_name): New variables.
959
960 2003-11-26 Milan Zamazal <[email protected]>
961
962 * czech.scm (czech-int_simple_params): New variable.
963 (voice_czech): Use `simple' intonation method.
964 (czech-default-synthesis-init): Initialize all required UniSyn
965 parameters.
966
967 2003-11-10 Milan Zamazal <[email protected]>
968
969 * czech.scm (lex.add.entry): Non-words removed.
970
971 2003-11-06 Milan Zamazal <[email protected]>
972
973 * czech.scm (czech-pos): Put the `sym' type on certain words.
974
975 2003-11-05 Milan Zamazal <[email protected]>
976
977 * czech.scm (lex.add.entry): Pronounciation of standalone
978 one-letter words added.
979
980 2003-10-27 Milan Zamazal <[email protected]>
981
982 * czech.scm (czech-pause_method): New function.
983 (voice_czech): Set it as the pause method.
984 (defPhoneSet): New phone `##'.
985 (PhoneSet.silences): Added.
986 (czech-phoneme_durations): Added.
987
988 2003-10-17 Milan Zamazal <[email protected]>
989
990 * czech.scm (lex.add.entry): "shift" and "control" added.
991
992 2003-10-01 Milan Zamazal <[email protected]>
993
994 * czech.scm (czech-phone-adjustment): Run only if the current
995 language is Czech.
996
997 2003-09-23 Milan Zamazal <[email protected]>
998
999 * czech.scm (lex.add.entry): Braces added.
1000 (lex.add.entry): `punc' parentheses added.
1001 (czech-pos): Handle parentheses as punctuation.
1002
1003 2003-09-22 Milan Zamazal <[email protected]>
1004
1005 * czech.scm (lex.add.entry): Mistakenly used y's replaced by i's.
1006 (czech-pos): If there's no `punctype' feature set on the token,
1007 mark punctuation characters as punctuation.
1008
1009 2003-09-19 Milan Zamazal <[email protected]>
1010
1011 * czech.scm (czech-token.whitespace): New variable.
1012 (voice_czech): Set token.whitespace.
1013
1014 2003-09-18 Milan Zamazal <[email protected]>
1015
1016 * czech.scm (lts.ruleset): Bug fixes, additions, improvements.
1017 (czech-number): Handle introducing + and -; handle commas.
1018 (czech-tokenize-on-nonalphas): Handle numbers too.
1019 (czech-token_to_words): Don't fail on numbers containing
1020 punctuation; handle numbers with +, -, and comma.
1021 (lex.add.entry): Numerical punctuation added; `range' - added.
1022 (czech-number): Use lexicon lookup for + and -.
1023 (czech-token_to_words): `punctype' token feature introduced.
1024 (czech-pos): Consider it.
1025 (czech-number): Use the comma character directly, not its spoken
1026 form.
1027 (czech-token_to_words): Don't accept final `,' or `.' as comma.
1028 Handle dot just after dot in a very special way,
1029 (lex.add.entry): Pronounce ("." num) as dot.
1030 (czech-token_to_words, czech-pos): `punctype' token feature
1031 handling cosmetic changes.
1032 (czech-token_to_words): Don't duplicate after-number dot at end of
1033 string.
1034
1035 2003-09-17 Milan Zamazal <[email protected]>
1036
1037 * czech.scm (czech-token_to_words): Spell numbers starting with 0.
1038 (czech-token_to_words): Exclude R from the list of spell-only
1039 letters.
1040 (czech-token_to_words): Don't split non-syllable words that are in
1041 lexicon. Let starting L or R doesn't make syllable.
1042 (lex.add.entry): Duplicate backslash definition removed, double
1043 quote definition added.
1044
1045 2003-08-29 Milan Zamazal <[email protected]>
1046
1047 * czech.scm (czech-int_simple_params): Removed.
1048 (voice_czech): Don't set czech-int_simple_params. Set
1049 Int_Targets. Set Int_Method to Intonation_Tree.
1050
1051 2003-08-28 Milan Zamazal <[email protected]>
1052
1053 * czech.scm (czech-pos): New function.
1054 (voice_czech): Set it as POS_Method.
1055 (lex.add.entry): Punctuation entries defined properly.
1056 (lex.add.entry): Pauses removed from the lexicon words.
1057 (czech-token_to_words): Don't consider `s' being syllabic.
1058 (lex.add.entry): New words.
1059
1060 2003-08-27 Milan Zamazal <[email protected]>
1061
1062 * czech-mbrola.scm (PhoneSet.silences): Removed.
1063 (provide): Added.
1064
1065 * czech.scm (czech-unknown-symbol-message): Renamed to
1066 czech-unknown-symbol-word. Value changed to a single word value.
1067 (czech-downcase): Don't include punctuation.
1068 (czech-token.punctuation, czech-token.prepunctuation): New
1069 variables.
1070 (voice_czech): Use them.
1071 (lex.set.lts.ruleset): Don't call it.
1072 (lex.add.entry): Parentheses added.
1073 (lex.add.entry): Proper sylabification and accents added.
1074 (czech-phoneme_durations): Typo in e:'s value fixed.
1075 (czech-phoneme_durations): Random vowel duration adjustments.
1076 (czech-token.punctuation): Dash added.
1077 (lex.add.entry): "Emacs" and "copyright" added.
1078 (czech-tokenize-on-nonalphas): New function.
1079 (czech-token_to_words): Handle punctuation characters in tokens.
1080 (czech-token_to_words): Apply recursively on parts separated by
1081 punctuation. Spell non-syllabic words.
1082