mahta.fetrat
/
HomoFast-eSpeak-Persian
Watch 1
Star 0
Fork 0
Code Issues 0 Pull Requests 0 Releases 0 Wiki Activity
Browse Source

[1.30.05] 

Fix an error where vowels start too loud (too strong attack).

Added command-line option  --compile-debug, which is like --compile but it includes source line information from the *_rules file, which is displayed with the -X option. 

espeakedit: Added  Compile->Compile_dictionary(debug),

Fixed <emphasis> which was broken in the previous version(s).
<emphasis> now reduces the previous stressed syllable to secondary stress.


git-svn-id: https://espeak.svn.sourceforge.net/svnroot/espeak/trunk@119 d46cf337-b52f-0410-862d-fd96e6ae7743
master
jonsd 17 years ago
parent
af575bc1db
commit
dd9fd5f27a
22 changed files with 123 additions and 42 deletions
Split View Show Diff Stats
  1. 1
    0
      dictsource/en_rules
  2. 7
    4
      docs/commands.html
  3. 1
    1
      platforms/windows/windows_sapi/ttsengobj.cpp
  4. 8
    2
      src/compiledata.cpp
  5. 14
    2
      src/compiledict.cpp
  6. 19
    5
      src/dictionary.cpp
  7. 6
    4
      src/espeak.cpp
  8. 5
    1
      src/espeakedit.cpp
  9. 13
    1
      src/intonation.cpp
  10. 1
    0
      src/main.h
  11. 1
    0
      src/menus.cpp
  12. 7
    2
      src/prosodydisplay.cpp
  13. 1
    1
      src/readclause.cpp
  14. 8
    6
      src/speak.cpp
  15. 3
    3
      src/speak_lib.cpp
  16. 7
    1
      src/speak_lib.h
  17. 1
    1
      src/speak_riscos.cpp
  18. 1
    1
      src/spect.cpp
  19. 2
    2
      src/synthdata.cpp
  20. 13
    4
      src/translate.cpp
  21. 2
    1
      src/translate.h
  22. 2
    0
      src/vowelchart.cpp

+ 1
- 0
dictsource/en_rules View File

@@ -4242,6 +4242,7 @@
rea (C% r,i:@
_) rea (dC r,i:@
rea (gen rI2eI
_) re (au@P2 ,ri:
re (ck rE
reckon rEk@n
reco (g rEk@

+ 7
- 4
docs/commands.html View File

@@ -3,7 +3,6 @@

<head>
<title>eSpeak Speech Synthesizer</title>
<meta name="GENERATOR" content="Quanta Plus">
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
</head>
<body>
@@ -151,16 +150,20 @@ shown on stdout.
<strong>--stdout</strong><br>
<dd>Writes the speech output to stdout as it is produced, rather than speaking it. The data starts with a WAV file header which indicates the sample rate and format of the data. The length field is set to zero because the length of the data is unknown when the header is produced.
<p>
<dt><strong>--compile[=&lt;voice name&gt;]</strong><br>
<dt><strong>--compile [=&lt;voice name&gt;]</strong><br>
<dd>
Compile the pronunciation rule and dictionary lookup data from their source files in the current directory. The Voice determines which language's files are compiled. For example, if it's an English voice, then <em>en_rules</em>, <em>en_list</em>, and <em>en_extra</em> (if present), are compiled to replace <em>en_dict</em> in the <em>speak-data</em> directory. If no Voice is specified then the default Voice is used.
<p>
<dt><strong>--punct[="&lt;characters&gt;"]</strong><br>
<dt><strong>--compile-debug [=&lt;voice name&gt;]</strong><br>
<dd>
The same as <strong>--compile</strong>, but source line numbers from the *_rules file are included. These are included in the rules trace when the <strong>-X</strong> option is used.
<p>
<dt><strong>--punct [="&lt;characters&gt;"]</strong><br>
<dd>
Speaks the names of punctuation characters when they are encountered in the text. If &lt;characters&gt; are given, then only those listed punctuation characters are spoken, eg. <code> --punct=".,;?"</code>
<p>
<dt>
<strong>--voices[=&lt;language code&gt;]</strong><br>
<strong>--voices [=&lt;language code&gt;]</strong><br>
<dd>Lists the available voices.<br>
If =&lt;language code&gt; is present then only those voices which are suitable for that language are listed.<br>
<code>--voices=variant</code> lists the voice variants (voice modifiers).

+ 1
- 1
platforms/windows/windows_sapi/ttsengobj.cpp View File

@@ -767,7 +767,7 @@ int FAR PASCAL CompileDictionary(const char *voice, const char *path_log)
sprintf(fname,"%s/",path_install);
espeak_SetVoiceByName(voice);
espeak_CompileDictionary(fname,f_log3);
espeak_CompileDictionary(fname,f_log3,0);
fclose(f_log3);
return(0);

+ 8
- 2
src/compiledata.cpp View File

@@ -77,7 +77,7 @@ extern int Read4Bytes(FILE *f);
extern void MakeVowelLists(void);
extern void FindPhonemesUsed(void);
extern void DrawEnvelopes();
extern int CompileDictionary(const char *dsource, const char *dict_name, FILE *log, char *fname);
extern int CompileDictionary(const char *dsource, const char *dict_name, FILE *log, char *fname, int flags);

static int markers_used[8];

@@ -1952,10 +1952,16 @@ void Compile::Report(void)

fprintf(f_report," [%s] %s",WordToString(prev_mnemonic = list[ix]->ph_mnemonic), phoneme_tab_list2[prev_table = list[ix]->ph_table].name);
fputc('\n',f_report);
}

for(ix=0; ix<n; ix++)
{
free(list[ix]);
list[ix] = NULL;
}

free(list);
list = NULL;
fclose(f_report);
}

@@ -2144,7 +2150,7 @@ wxString CompileAllDictionaries()

LoadVoice(voicename,0);

if((err = CompileDictionary(path_dsource, dictname,log,NULL)) > 0)
if((err = CompileDictionary(path_dsource, dictname,log,NULL,0)) > 0)
{
report = report + dictstr + wxString::Format(_T(" %d, "),err);
errors += err;

+ 14
- 2
src/compiledict.cpp View File

@@ -46,6 +46,7 @@ static int transpose_offset; // transpose character range for LookupDictList()
static int transpose_min;
static int transpose_max;
static int text_mode = 0;
static int debug_flag = 0;

int hash_counts[N_HASH_DICT];
char *hash_chains[N_HASH_DICT];
@@ -943,6 +944,15 @@ char *compile_rule(char *input)
}
strcpy(&output[len],rule_match);
len += strlen(rule_match);

if(debug_flag)
{
output[len] = RULE_LINENUM;
output[len+1] = (linenum % 255) + 1;
output[len+2] = (linenum / 255) + 1;
len+=3;
}

if(rule_cond[0] != 0)
{
ix = -1;
@@ -1437,9 +1447,10 @@ static int compile_dictrules(FILE *f_in, FILE *f_out, char *fname_temp)



int CompileDictionary(const char *dsource, const char *dict_name, FILE *log, char *fname_err)
{//==========================================================================================
int CompileDictionary(const char *dsource, const char *dict_name, FILE *log, char *fname_err, int flags)
{//=====================================================================================================
// fname: space to write the filename in case of error
// flags: bit 0: include source line number information, for debug purposes.

FILE *f_in;
FILE *f_out;
@@ -1451,6 +1462,7 @@ int CompileDictionary(const char *dsource, const char *dict_name, FILE *log, cha
char path[sizeof(path_home)+40]; // path_dsource+20

error_count = 0;
debug_flag = flags & 1;

if(dsource == NULL)
dsource = "";

+ 19
- 5
src/dictionary.cpp View File

@@ -1455,6 +1455,7 @@ char *Translator::DecodeRule(const char *group, char *rule)
int match_type;
int finished=0;
int value;
int linenum=0;
int flags;
int suffix_char;
int condition_num=0;
@@ -1476,7 +1477,7 @@ char *Translator::DecodeRule(const char *group, char *rule)
{
rb = *rule++;

if(rb <= 5)
if(rb <= RULE_LINENUM)
{
switch(rb)
{
@@ -1501,6 +1502,11 @@ char *Translator::DecodeRule(const char *group, char *rule)
/* conditional rule, next byte gives condition number */
condition_num = *rule++;
break;
case RULE_LINENUM:
value = (rule[1] & 0xff) - 1;
linenum = (rule[0] & 0xff) - 1 + (value * 255);
rule+=2;
break;
}
continue;
}
@@ -1557,10 +1563,15 @@ char *Translator::DecodeRule(const char *group, char *rule)
*p = 0;

p = output;
if(linenum > 0)
{
sprintf(p,"%5d:\t",linenum);
p += 7;
}
if(condition_num > 0)
{
sprintf(output,"?%d ",condition_num);
p = &output[strlen(output)];
sprintf(p,"?%d ",condition_num);
p = &p[strlen(p)];
}
if((ix = strlen(buf_pre)) > 0)
{
@@ -1727,7 +1738,7 @@ void Translator::MatchRule(char *word[], const char *group, char *rule, MatchRec
{
rb = *rule++;

if(rb <= 5)
if(rb <= RULE_LINENUM)
{
switch(rb)
{
@@ -1779,6 +1790,9 @@ void Translator::MatchRule(char *word[], const char *group, char *rule, MatchRec
if(!failed)
match.points++; // add one point for a matched conditional rule
break;
case RULE_LINENUM:
rule+=2;
break;
}
continue;
}
@@ -2173,7 +2187,7 @@ void Translator::MatchRule(char *word[], const char *group, char *rule, MatchRec
if(group_length > 1)
pts += 35; // to account for an extra letter matching
DecodePhonemes(match.phonemes,decoded_phonemes);
fprintf(f_trans,"%3d %s [%s]\n",pts,DecodeRule(group,rule_start),decoded_phonemes);
fprintf(f_trans,"%3d\t%s [%s]\n",pts,DecodeRule(group,rule_start),decoded_phonemes);
}
#endif


+ 6
- 4
src/espeak.cpp View File

@@ -274,10 +274,11 @@ int main (int argc, char **argv)
We distinguish them by their indices. */
{"help", no_argument, 0, 'h'},
{"stdin", no_argument, 0, 0x100},
{"stdout", no_argument, 0, 0x101},
{"compile-debug", optional_argument, 0, 0x101},
{"compile", optional_argument, 0, 0x102},
{"punct", optional_argument, 0, 0x103},
{"voices", optional_argument, 0, 0x104},
{"stdout", no_argument, 0, 0x105},
{0, 0, 0, 0}
};

@@ -402,14 +403,15 @@ int main (int argc, char **argv)
flag_stdin = 1;
break;

case 0x101: // --stdout
case 0x105: // --stdout
option_waveout = 1;
strcpy(wavefile,"stdout");
break;

case 0x101: // --compile-debug
case 0x102: // --compile
strncpy0(voicename,optarg,sizeof(voicename));
flag_compile = 1;
flag_compile = c;
quiet = 1;
break;

@@ -466,7 +468,7 @@ int main (int argc, char **argv)
if(flag_compile)
{
// This must be done after the voice is set
espeak_CompileDictionary("",stderr);
espeak_CompileDictionary("", stderr, flag_compile & 0x1);
exit(0);
}


+ 5
- 1
src/espeakedit.cpp View File

@@ -186,6 +186,7 @@ BEGIN_EVENT_TABLE(MyFrame, wxMDIParentFrame)
EVT_MENU(MENU_PATH4, MyFrame::OnOptions)
EVT_MENU(MENU_COMPILE_PH, MyFrame::OnTools)
EVT_MENU(MENU_COMPILE_DICT, MyFrame::OnTools)
EVT_MENU(MENU_COMPILE_DICT_DEBUG, MyFrame::OnTools)
EVT_MENU(MENU_COMPILE_MBROLA, MyFrame::OnTools)
EVT_MENU(MENU_CLOSE_ALL, MyFrame::OnQuit)
EVT_MENU(MENU_QUIT, MyFrame::OnQuit)
@@ -431,6 +432,7 @@ void MyFrame::OnTools(wxCommandEvent& event)
{//=========================================
int err;
FILE *log;
int debug_flag=0;
char fname_log[sizeof(path_dsource)+12];
char err_fname[sizeof(path_home)+15];

@@ -465,11 +467,13 @@ void MyFrame::OnTools(wxCommandEvent& event)
CompileMbrola();
break;
case MENU_COMPILE_DICT_DEBUG:
debug_flag =1; // and drop through to next case
case MENU_COMPILE_DICT:
sprintf(fname_log,"%s%s",path_dsource,"dict_log");
log = fopen(fname_log,"w");

if((err = CompileDictionary(path_dsource,dictionary_name,log,err_fname)) < 0)
if((err = CompileDictionary(path_dsource,dictionary_name,log,err_fname,debug_flag)) < 0)
{
wxLogError(_T("Can't access file:\n")+wxString(err_fname,wxConvLocal));


+ 13
- 1
src/intonation.cpp View File

@@ -910,7 +910,7 @@ void Translator::CalcPitches(int clause_type)


option = langopts.intonation_group;
if(option > INTONATION_TYPES)
if(option >= INTONATION_TYPES)
option = 0;

group_tone = punct_to_tone[option][clause_type];
@@ -952,6 +952,18 @@ void Translator::CalcPitches(int clause_type)

if(syl->stress == 6)
{
// reduce the stress of the previous stressed syllable
for(ix=st_ix-1; ix>=st_start && ix>=(st_ix-3); ix--)
{
if(syllable_tab[ix].stress == 6)
break;
if(syllable_tab[ix].stress == 4)
{
syllable_tab[ix].stress = 3;
break;
}
}

// are the next primary syllables also emphasized ?
for(ix=st_ix+1; ix<n_st; ix++)
{

+ 1
- 0
src/main.h View File

@@ -133,6 +133,7 @@ enum {
MENU_PROSODY,
MENU_COMPILE_PH,
MENU_COMPILE_DICT,
MENU_COMPILE_DICT_DEBUG,
MENU_COMPILE_MBROLA,

MENU_SPEAK_TRANSLATE,

+ 1
- 0
src/menus.cpp View File

@@ -86,6 +86,7 @@ wxMenuBar *MakeMenu(int type)
data_menu = new wxMenu;
data_menu->Append(MENU_COMPILE_PH, _("Compile &phoneme data"));
data_menu->Append(MENU_COMPILE_DICT, _("Compile &dictionary"));
data_menu->Append(MENU_COMPILE_DICT_DEBUG, _("Compile dictionary (&debug)"));
data_menu->Append(MENU_COMPILE_MBROLA, _("Compile &mbrola phonemes list"));
// OPTIONS MENU

+ 7
- 2
src/prosodydisplay.cpp View File

@@ -120,6 +120,7 @@ void ProsodyDisplay::SelectPh(int index)
{//=====================================
// A phoneme has been selected
PHONEME_LIST *p;
const char *emphasized;
char buf[120];

if(index < 0) return;
@@ -127,8 +128,12 @@ void ProsodyDisplay::SelectPh(int index)
p = &phlist[index];
if((p == NULL) || (p->ph == NULL)) return;

sprintf(buf,"Stress %d Amp %2d LenMod %2d Pitch %3d %3d [env=%d] Flags %.2x ",
p->tone,p->amp,p->length,p->pitch1,p->pitch2,p->env,p->ph->phflags);
emphasized = "";
if(p->tone & 8)
emphasized = "*";

sprintf(buf,"Stress %s%d Amp %2d LenMod %2d Pitch %3d %3d [env=%d] Flags %.2x ",
emphasized,p->tone&0x7,p->amp,p->length,p->pitch1,p->pitch2,p->env,p->ph->phflags);
wxLogStatus(wxString(buf,wxConvLocal));
}


+ 1
- 1
src/readclause.cpp View File

@@ -845,7 +845,7 @@ static void ProcessParamStack(char *outbuf, int &outix)
int value;
char buf[20];
int new_parameters[N_SPEECH_PARAM];
static char cmd_letter[N_SPEECH_PARAM] = {0, 'S','A','P','R', 0, 0, 'F'}; // embedded command letters
static char cmd_letter[N_SPEECH_PARAM] = {0, 'S','A','P','R', 0, 0, 0, 0, 0, 0, 0, 'F'}; // embedded command letters


for(param=0; param<N_SPEECH_PARAM; param++)

+ 8
- 6
src/speak.cpp View File

@@ -354,10 +354,11 @@ int main (int argc, char **argv)
We distinguish them by their indices. */
{"help", no_argument, 0, 'h'},
{"stdin", no_argument, 0, 0x100},
{"stdout", no_argument, 0, 0x101},
{"compile-debug", optional_argument, 0, 0x101},
{"compile", optional_argument, 0, 0x102},
{"punct", optional_argument, 0, 0x103},
{"voices", optional_argument, 0, 0x104},
{"stdout", no_argument, 0, 0x105},
{0, 0, 0, 0}
};

@@ -531,15 +532,16 @@ int main (int argc, char **argv)
flag_stdin = 1;
break;

case 0x101: // --stdout
case 0x105: // --stdout
option_waveout = 1;
strcpy(wavefile,"stdout");
break;

case 0x101: // --compile-debug
case 0x102: // --compile
if(optarg2 != NULL)
strncpy0(voicename,optarg2,sizeof(voicename));
flag_compile = 1;
flag_compile = c;
break;

case 0x103: // --punct
@@ -574,16 +576,16 @@ int main (int argc, char **argv)
strcpy(path_dsource,path_home);
path_dsource[strlen(path_home)-11] = 0; // remove "espeak-data" from the end
strcat(path_dsource,"dictsource\\");
CompileDictionary(path_dsource,dictionary_name,NULL,NULL);
CompileDictionary(path_dsource,dictionary_name,NULL,NULL, flag_compile & 0x1);
#else
#ifdef PLATFORM_WINDOWS
char path_dsource[sizeof(path_home)+20];
strcpy(path_dsource,path_home);
path_dsource[strlen(path_home)-11] = 0; // remove "espeak-data" from the end
strcat(path_dsource,"dictsource\\");
CompileDictionary(path_dsource,dictionary_name,NULL,NULL);
CompileDictionary(path_dsource,dictionary_name,NULL,NULL, flag_compile & 0x1);
#else
CompileDictionary(NULL,dictionary_name,NULL,NULL);
CompileDictionary(NULL,dictionary_name,NULL,NULL, flag_compile & 0x1);
#endif
#endif
exit(0);

+ 3
- 3
src/speak_lib.cpp View File

@@ -1027,10 +1027,10 @@ ESPEAK_API void espeak_SetPhonemeTrace(int value, FILE *stream)
} // end of espeak_SetPhonemes


ESPEAK_API void espeak_CompileDictionary(const char *path, FILE *log)
{//==================================================================
ESPEAK_API void espeak_CompileDictionary(const char *path, FILE *log, int flags)
{//=============================================================================
ENTER("espeak_CompileDictionary");
CompileDictionary(path,dictionary_name,log,NULL);
CompileDictionary(path, dictionary_name, log, NULL, flags);
} // end of espeak_CompileDirectory



+ 7
- 1
src/speak_lib.h View File

@@ -35,6 +35,9 @@ Revision 2
Revision 3
Added espeakWORDGAP to espeak_PARAMETER

Revision 4
Added flags parameter to espeak_CompileDictionary()

*/
/********************/
/* Initialization */
@@ -442,13 +445,16 @@ void espeak_SetPhonemeTrace(int value, FILE *stream);
#ifdef __cplusplus
extern "C"
#endif
void espeak_CompileDictionary(const char *path, FILE *log);
void espeak_CompileDictionary(const char *path, FILE *log, int flags);
/* Compile pronunciation dictionary for a language which corresponds to the currently
selected voice. The required voice should be selected before calling this function.

path: The directory which contains the language's '_rules' and '_list' files.
'path' should end with a path separator character ('/').
log: Stream for error reports and statistics information. If log=NULL then stderr will be used.

flags: Bit 0: include source line information for debug purposes (This is displayed with the
-X command line option).
*/
/***********************/
/* Voice Selection */

+ 1
- 1
src/speak_riscos.cpp View File

@@ -776,7 +776,7 @@ void command_line(char *arg_string, int wait)
strncpy0(command,param_string(&p),sizeof(command));
if(memcmp(command,"compile=",8)==0)
{
CompileDictionary(NULL,&command[8],NULL,NULL);
CompileDictionary(NULL,&command[8],NULL,NULL,0);
return;
}
else

+ 1
- 1
src/spect.cpp View File

@@ -431,7 +431,7 @@ double SpectFrame::GetRms(int seq_amplitude)
{
total += ((htab[h] * htab[h]) >> 10);
}
rms = sqrt(total) / 14.5;
rms = sqrt(total) / 7.25;
// DrawPeaks(NULL,0,0,amp);
return(rms);
}

+ 2
- 2
src/synthdata.cpp View File

@@ -35,8 +35,8 @@
#include "translate.h"
#include "wave.h"

const char *version_string = "1.30.04 07.Jan.08";
const int version_phdata = 0x013000;
const char *version_string = "1.30.05 07.Jan.08";
const int version_phdata = 0x013005;

int option_device_number = -1;


+ 13
- 4
src/translate.cpp View File

@@ -700,7 +700,7 @@ if((wmark > 0) && (wmark < 8))
}
}

if((wflags & FLAG_ALL_UPPER) && (word_length > 1) && (clause_lower_count > 3) && iswalpha(first_char))
if((wflags & FLAG_ALL_UPPER) && (word_length > 1)&& iswalpha(first_char))
{
if((option_tone_flags & OPTION_EMPHASIZE_ALLCAPS) && !(dictionary_flags[0] & FLAG_ABBREV))
{
@@ -708,7 +708,7 @@ if((wmark > 0) && (wmark < 8))
emphasize_allcaps = FLAG_EMPHASIZED;
}
else
if(!found && !(dictionary_flags[0] & FLAG_SKIPWORDS) && (word_length<4) && (clause_upper_count <= clause_lower_count))
if(!found && !(dictionary_flags[0] & FLAG_SKIPWORDS) && (word_length<4) && (clause_lower_count > 3) && (clause_upper_count <= clause_lower_count))
{
// An upper case word in a lower case clause. This could be an abbreviation.
spell_word = 1;
@@ -1023,12 +1023,21 @@ strcpy(phonemes2,phonemes);
strcat(word_phonemes,end_phonemes);
}

if(wflags & FLAG_LAST_WORD)
{
// don't use $brk pause before the last word of a sentence
// (but allow it for emphasis, see below
dictionary_flags[0] &= ~FLAG_PAUSE1;
}

if(wflags & FLAG_EMPHASIZED)
{
// A word is indicated in the source text as stressed
// Give it stress level 6 (for the intonation module)
ChangeWordStress(this,word_phonemes,6);
dictionary_flags[0] |= FLAG_PAUSE1; // precede by short pause

// if(!(wflags & FLAG_LAST_WORD)) // ?? omit pre-pause if it's the last word in the sentence?
dictionary_flags[0] |= FLAG_PAUSE1; // precede by short pause
}
else
if(wtab[dictionary_skipwords].flags & FLAG_LAST_WORD)
@@ -1335,7 +1344,7 @@ int Translator::TranslateWord2(char *word, WORD_TAB *wtab, int pre_pause, int ne
}
}

if(!(word_flags & FLAG_LAST_WORD) && !(word_flags & FLAG_HYPHEN))
if(!(word_flags & FLAG_HYPHEN))
{
if(flags & FLAG_PAUSE1)
{

+ 2
- 1
src/translate.h View File

@@ -120,6 +120,7 @@
#define RULE_CONDITION 5 // followed by condition number (byte)
#define RULE_GROUP_START 6
#define RULE_GROUP_END 7
#define RULE_LINENUM 8 // next 2 bytes give a line number, for debugging purposes

#define RULE_SPACE 32 // ascii space
#define RULE_SYLLABLE 9
@@ -549,7 +550,7 @@ extern int (* phoneme_callback)(const char *);
extern void SetLengthMods(Translator *tr, int value);

Translator *SelectTranslator(const char *name);
int CompileDictionary(const char *dsource, const char *dict_name, FILE *log, char *err_name);
int CompileDictionary(const char *dsource, const char *dict_name, FILE *log, char *err_name,int flags);
void LoadConfig(void);
int PhonemeCode(unsigned int mnem);
void ChangeWordStress(Translator *tr, char *word, int new_stress);

+ 2
- 0
src/vowelchart.cpp View File

@@ -388,6 +388,8 @@ void FindPhonemesUsed(void)
{
if(*p == RULE_CONDITION)
p+=2;
if(*p == RULE_LINENUM)
p+=3;
if(*p == RULE_GROUP_END)
{
p++;

Write Preview
Loading…
Cancel
Save