summaryrefslogtreecommitdiffstats
path: root/apps/talk.c
blob: a13dd60ec39fc5c2fdece7f544bd1d5411f0aaef (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
/***************************************************************************
 *             __________               __   ___.
 *   Open      \______   \ ____   ____ |  | _\_ |__   _______  ___
 *   Source     |       _//  _ \_/ ___\|  |/ /| __ \ /  _ \  \/  /
 *   Jukebox    |    |   (  <_> )  \___|    < | \_\ (  <_> > <  <
 *   Firmware   |____|_  /\____/ \___  >__|_ \|___  /\____/__/\_ \
 *                     \/            \/     \/    \/            \/
 * $Id$
 *
 * Copyright (C) 2004 Jrg Hohensohn
 *
 * This module collects the Talkbox and voice UI functions.
 * (Talkbox reads directory names from mp3 clips called thumbnails,
 *  the voice UI lets menus and screens "talk" from a voicefont in memory.
 *
 * All files in this archive are subject to the GNU General Public License.
 * See the file COPYING in the source tree root for full license agreement.
 *
 * This software is distributed on an "AS IS" basis, WITHOUT WARRANTY OF ANY
 * KIND, either express or implied.
 *
 ****************************************************************************/

#include <stdio.h>
#include <stddef.h>
#include "file.h"
#include "buffer.h"
#include "system.h"
#include "settings.h"
#include "mp3_playback.h"
#include "mpeg.h"
#include "lang.h"
#include "talk.h"
#include "id3.h"
extern void bitswap(unsigned char *data, int length); /* no header for this */

/***************** Constants *****************/

#define QUEUE_SIZE 50
const char* dir_thumbnail_name = ".dirname.tbx";


/***************** Data types *****************/

struct clip_entry /* one entry of the index table */
{
    int offset; /* offset from start of voicefont file */
    int size; /* size of the clip */
};

struct voicefont /* file format of our "voicefont" */
{
    int version; /* version of the voicefont */
    int table;   /* offset to index table, (=header size) */
    int id1_max; /* number of "normal" clips contained in above index */
    int id2_max; /* number of "voice only" clips contained in above index */
    struct clip_entry index[]; /* followed by the index tables */
    /* and finally the bitswapped mp3 clips, not visible here */
};


struct queue_entry /* one entry of the internal queue */
{
    unsigned char* buf;
    int len;
};



/***************** Globals *****************/

static unsigned char* p_thumbnail; /* buffer for thumbnail */
static long size_for_thumbnail; /* leftover buffer size for it */
static struct voicefont* p_voicefont; /* loaded voicefont */
static bool has_voicefont; /* a voicefont file is present */
static bool is_playing; /* we're currently playing */
static struct queue_entry queue[QUEUE_SIZE]; /* queue of scheduled clips */
static int queue_write; /* write index of queue, by application */
static int queue_read; /* read index of queue, by ISR context */

/***************** Private prototypes *****************/

static int load_voicefont(void);
static void mp3_callback(unsigned char** start, int* size);
static int shutup(void);
static int queue_clip(unsigned char* buf, int size, bool enqueue);
static int open_voicefile(void);


/***************** Private implementation *****************/

static int open_voicefile(void)
{
    char buf[64];
    char* p_lang = "english"; /* default */

    if ( global_settings.lang_file[0] &&
         global_settings.lang_file[0] != 0xff ) 
    {   /* try to open the voice file of the selected language */
        p_lang = global_settings.lang_file;
    }

    snprintf(buf, sizeof(buf), ROCKBOX_DIR LANG_DIR "/%s.voice", p_lang);

    return open(buf, O_RDONLY);
}



static int load_voicefont(void)
{
    int fd;
    int size;

    p_voicefont = NULL; /* indicate no voicefont if we fail below */

    fd = open_voicefile();
    if (fd < 0) /* failed to open */
    {
        p_voicefont = NULL; /* indicate no voicefont */
        has_voicefont = false; /* don't try again */
        return 0;
    }

    size = read(fd, mp3buf, mp3end - mp3buf);
    if (size > 1000
        && ((struct voicefont*)mp3buf)->table
           == offsetof(struct voicefont, index))
    {
        p_voicefont = (struct voicefont*)mp3buf;

        /* thumbnail buffer is the remaining space behind */
        p_thumbnail = mp3buf + size;
        p_thumbnail += (int)p_thumbnail % 2; /* 16-bit align */
        size_for_thumbnail = mp3end - p_thumbnail;
    }
    else
    {
        has_voicefont = false; /* don't try again */
    }
    close(fd);

    return size;
}


/* called in ISR context if mp3 data got consumed */
static void mp3_callback(unsigned char** start, int* size)
{
    int play_now;

    if (queue[queue_read].len > 0) /* current clip not finished? */
    {   /* feed the next 64K-1 chunk */
        play_now = MIN(queue[queue_read].len, 0xFFFF);
        *start = queue[queue_read].buf;
        *size = play_now;
        queue[queue_read].buf += play_now;
        queue[queue_read].len -= play_now;
        return;
    }
    else /* go to next entry */
    {
        queue_read++;
        if (queue_read >= QUEUE_SIZE)
            queue_read = 0;
    }

    if (queue_read != queue_write) /* queue is not empty? */
    {   /* start next clip */
        play_now = MIN(queue[queue_read].len, 0xFFFF);
        *start = queue[queue_read].buf;
        *size = play_now;
        queue[queue_read].buf += play_now;
        queue[queue_read].len -= play_now;
    }
    else
    {
        *size = 0; /* end of data */
        is_playing = false;
        mp3_play_stop(); /* fixme: should be done by caller */
    }
}

/* stop the playback and the pending clips, but at frame boundary */
static int shutup(void)
{
    unsigned char* pos;
    unsigned char* search;
    unsigned char* end;
    /* one silent bitswapped mp3 frame (22kHz), without bit reservoir */
    static const unsigned char silent_frame[] = {
        0xFF,0xCF,0x08,0x23,0x00,0x00,0x00,0xC0,0x12,0x80,0x01,0x00,0x00,
        0x32,0x82,0xB2,0xA2,0xCC,0x74,0x9C,0xCC,0xAA,0xAA,0xAA,0xAA,0xAA,
    };

    mp3_play_pause(false); /* pause */

    if (!is_playing) /* has ended anyway */
        return 0;
    
    /* search next frame boundary and continue up to there */
    pos = search = mp3_get_pos();
    end = queue[queue_read].buf + queue[queue_read].len;

    /* Find the next frame boundary */
    while (search < end) /* search the remaining data */
    {
        if (*search++ != 0xFF) /* search for frame sync byte */
        {
            continue;
        }
            
        /* look at the (bitswapped) 2nd byte of header candidate */
        if ((*search & 0x07) == 0x07  /* rest of frame sync */
         && (*search & 0x18) != 0x10  /* version != reserved */
         && (*search & 0x60) != 0x00) /* layer != reserved */
        {
            search--; /* back to the sync byte */
            break; /* From looking at the first 2 bytes, this is a header. */
            /* this is not a sufficient condition to find header,
               may give "false alert" (end too early), but a start */
        }
    }

    queue_write = queue_read; /* reset the queue */
    is_playing = false;

    /* play old data until the frame end, to keep the MAS in sync */
    if (search-pos)
        queue_clip(pos, search-pos, true);

    /* If the voice clips contain dependent frames (currently they don't),
       it may be a good idea to insert an independent dummy frame here. */
    queue_clip((unsigned char*)silent_frame, sizeof(silent_frame), true);
    
    return 0;
}


/* schedule a clip, at the end or discard the existing queue */
static int queue_clip(unsigned char* buf, int size, bool enqueue)
{
    if (!enqueue)
        shutup(); /* cut off all the pending stuff */
    
    queue[queue_write].buf = buf;
    queue[queue_write].len = size;
    
    /* FixMe: make this IRQ-safe */
        
    if (!is_playing)
    {   /* queue empty, we have to do the initial start */
        int size_now = MIN(size, 0xFFFF); /* DMA can do no more */
        is_playing = true;
        mp3_play_data(buf, size_now, mp3_callback);
        mp3_play_pause(true); /* kickoff audio */
        queue[queue_write].buf += size_now;
        queue[queue_write].len -= size_now;
    }

    queue_write++;
    if (queue_write >= QUEUE_SIZE)
        queue_write = 0;

    return 0;
}


/***************** Public implementation *****************/

void talk_init(void)
{
    int fd;

    fd = open_voicefile();
    if (fd >= 0) /* success */
    {
        close(fd);
        has_voicefont = true;
    }
    else
    {
        has_voicefont = false; /* no voice file available */
    }
    
    talk_buffer_steal(); /* abuse this for most of our inits */
    queue_write = queue_read = 0; 
}


/* somebody else claims the mp3 buffer, e.g. for regular play/record */
int talk_buffer_steal(void)
{
    p_voicefont = NULL; /* indicate no voicefont (trashed) */
    p_thumbnail = mp3buf; /*  whole space for thumbnail */
    size_for_thumbnail = mp3end - mp3buf;
    return 0;
}


/* play a voice ID from voicefont */
int talk_id(int id, bool enqueue)
{
    int clipsize;
    unsigned char* clipbuf;
    int unit;

    if (mpeg_status()) /* busy, buffer in use */
        return -1; 

    if (p_voicefont == NULL && has_voicefont)
        load_voicefont(); /* reload needed */

    if (p_voicefont == NULL) /* still no voices? */
        return -1;

    if (id == -1) /* -1 is an indication for silence */
        return -1;

    /* check if this is a special ID, with a value */
    unit = ((unsigned)id) >> UNIT_SHIFT;
    if (unit)
    {   /* sign-extend the value */
        //splash(200, true,"unit=%d", unit);
        id = (unsigned)id << (32-UNIT_SHIFT);
        id >>= (32-UNIT_SHIFT);
        talk_value(id, unit, enqueue); /* speak it */
        return 0; /* and stop, end of special case */
    }

    if (id > VOICEONLY_DELIMITER)
    {   /* voice-only entries use the second part of the table */
        id -= VOICEONLY_DELIMITER + 1;
        if (id >= p_voicefont->id2_max)
            return -1; /* must be newer than we have */
        id += p_voicefont->id1_max; /* table 2 is behind table 1 */
    }
    else
    {   /* normal use of the first table */
        if (id >= p_voicefont->id1_max)
            return -1; /* must be newer than we have */
    }
    
    clipsize = p_voicefont->index[id].size;
    if (clipsize == 0) /* clip not included in voicefont */
        return -1;

    clipbuf = mp3buf + p_voicefont->index[id].offset;

    queue_clip(clipbuf, clipsize, enqueue);

    return 0;
}


/* play a thumbnail from file */
int talk_file(char* filename, bool enqueue)
{
    int fd;
    int size;
    struct mp3entry info;

    if (mpeg_status()) /* busy, buffer in use */
        return -1; 

    if (p_thumbnail == NULL || size_for_thumbnail <= 0)
        return -1;

    if(mp3info(&info, filename)) /* use this to find real start */
    {   
        return 0; /* failed to open, or invalid */
    }

    fd = open(filename, O_RDONLY);
    if (fd < 0) /* failed to open */
    {
        return 0;
    }

    lseek(fd, info.first_frame_offset, SEEK_SET); /* behind ID data */

    size = read(fd, p_thumbnail, size_for_thumbnail);
    close(fd);

    /* ToDo: find audio, skip ID headers and trailers */

    if (size)
    {
        bitswap(p_thumbnail, size);
        queue_clip(p_thumbnail, size, enqueue);
    }

    return size;
}


/* say a numeric value, this word ordering works for english,
   but not necessarily for other languages (e.g. german) */
int talk_number(int n, bool enqueue)
{
    int level = 0; /* mille count */
    int mil = 1000000000; /* highest possible "-illion" */

    if (mpeg_status()) /* busy, buffer in use */
        return -1; 

    if (!enqueue)
        shutup(); /* cut off all the pending stuff */
    
    if (n==0)
    {   /* special case */
        talk_id(VOICE_ZERO, true);
        return 0;
    }
    
    if (n<0)
    {
        talk_id(VOICE_MINUS, true);
        n = -n;
    }
    
    while (n)
    {
        int segment = n / mil; /* extract in groups of 3 digits */
        n -= segment * mil; /* remove the used digits from number */
        mil /= 1000; /* digit place for next round */

        if (segment)
        {
            int hundreds = segment / 100;
            int ones = segment % 100;

            if (hundreds)
            {
                talk_id(VOICE_ZERO + hundreds, true);
                talk_id(VOICE_HUNDRED, true);
            }

            /* combination indexing */
            if (ones > 20)
            {
               int tens = ones/10 + 18;
               talk_id(VOICE_ZERO + tens, true);
               ones %= 10;
            }

            /* direct indexing */
            if (ones)
                talk_id(VOICE_ZERO + ones, true);
 
            /* add billion, million, thousand */
            if (mil)
                talk_id(VOICE_BILLION + level, true);
        }
        level++;
    }

    return 0;
}

int talk_value(int n, int unit, bool enqueue)
{
    int unit_id;
    const int unit_voiced[] = 
    {   /* lookup table for the voice ID of the units */
        -1, -1, -1, /* regular ID, int, signed */
        VOICE_MILLISECONDS, /* here come the "real" units */
        VOICE_SECONDS, 
        VOICE_MINUTES, 
        VOICE_HOURS, 
        VOICE_KHZ, 
        VOICE_DB, 
        VOICE_PERCENT, 
        VOICE_MEGABYTE, 
        VOICE_GIGABYTE,
        VOICE_MILLIAMPHOURS,
        VOICE_PIXEL,
        VOICE_PER_SEC,
        VOICE_HERTZ,
    };

    if (mpeg_status()) /* busy, buffer in use */
        return -1; 

    if (unit < 0 || unit >= UNIT_LAST)
        unit_id = -1;
    else
        unit_id = unit_voiced[unit];

    if ((n==1 || n==-1) /* singular? */
        && unit_id >= VOICE_SECONDS && unit_id <= VOICE_HOURS)
    {
        unit_id--; /* use the singular for those units which have */
    }

    /* special case with a "plus" before */
    if (n > 0 && (unit == UNIT_SIGNED || unit == UNIT_DB))
    {
        talk_id(VOICE_PLUS, enqueue);
        enqueue = true;
    }

    talk_number(n, enqueue); /* say the number */
    talk_id(unit_id, true); /* say the unit, if any */
        
    return 0;
}

/* spell a string */
int talk_spell(char* spell, bool enqueue) 
{
    char c; /* currently processed char */
    
    if (mpeg_status()) /* busy, buffer in use */
        return -1; 

    if (!enqueue)
        shutup(); /* cut off all the pending stuff */
    
    while ((c = *spell++) != '\0')
    {
        /* if this grows into too many cases, I should use a table */
        if (c >= 'A' && c <= 'Z')
            talk_id(VOICE_CHAR_A + c - 'A', true);
        else if (c >= 'a' && c <= 'z')
            talk_id(VOICE_CHAR_A + c - 'a', true);
        else if (c >= '0' && c <= '9')
            talk_id(VOICE_ZERO + c - '0', true);
        else if (c == '-')
            talk_id(VOICE_MINUS, true);
        else if (c == '+')
            talk_id(VOICE_PLUS, true);
        else if (c == '.')
            talk_id(VOICE_POINT, true);
    }

    return 0;
}