Comments

when executing mkdfa.pl  mysample

i got  Warning:     Class accepted flag overflow

is it a serious error ? 

i check the code and find max num of Class is sizeof(CLASSFLAGS) * 8;

in mkfa.h

#define unsigned int   CLASSFLAGS;

i have more than 256 Class. what could i do ?

Best regards

albert

When I run mkdfa.pl on the supplied sample files, I get the following.  I'm on Windows XP and logged in as an administrator.  The "test path" worked as instructed (except "julian" isn't there; it is "julius" in version 4).  I tried dos2unix and unix2dos with no change in the behavior.  I installed Julius 4.1.5 instead of 3.5.2.

# mkdfa.pl sample
sample.grammar has 3 rules
sample.voca    has 6 categories and 18 words
---
Error:       Can't open grammar file "/tmp/g1916.grammar"
Error: cannot open "sample.dfa.tmp"
---
no .dfa or .dict file generated

The cygwin/tmp folder exists, but is empty.  It creates a "sample.term" file, but nothing else.

Any suggestions on what I should try next?

I have made the sample.voca and sample.grammar files:

.voca:

% NS_B
<s>    sil

% NS_E
</s>    sil

% CALL_V
PHONE    f ow n
CALL    k ao l

% DIAL_V
DIAL    d ay l

% NAME_N
STEVE    s t iy v
YOUNG    y ah ng

% DIGIT
FIVE    f ay v
FOUR    f ow r
NINE    n ay n
EIGHT    ey t
OH     ow
ONE    w ah n
SEVEN    s eh v ih n
SIX     s ih k s
THREE    th r iy
TWO    t uw
ZERO    z iy r ow

.grammar:

S: NS_B SENT NS_E
SENT: CALL_V NAME_N
SENT: DIAL_V DIGIT

But when I run mkdfa.pl sample I get the following error:

sample.grammar has 3 rules

sample.voca has 6 categories and 18 words

---

Now parsing grammar file

"rror: Lexical mistake "

---

no .dfa or .dict file generated

i have the similar problem. My voca is:

% NS_B

<s> sil

% NS_E

</s> sil

% CALL_V

GỌI g ọi 

% DIAL_V

BẤM b ấ m

% NAME_N

KHÁNH kh á nh

HÀ h à

THẦY th ầ y

ĐẠT đ ạ t

LINH l i nh

% DIGIT

KHÔNG kh ô ng

MỘT m ộ t

HAI h ai

BA b a

BỐN b ố n

NĂM n ă m

SÁU s á u 

BẢY b ả y

TÁM t á m

CHÍN ch í n

MƯỜI m ườ i

hi,

  i followed the steps till here and when i ran mkdfa.pl then it shows the following,

i7922737@w32328:~/voxforge/auto$ perl mkdfa.pl sample
sample.grammar has 3 rules
sample.voca    has 6 categories and 18 words
---
---
no .dfa or .dict file generated

i checked both my sample files and everything seems to be ok... i am on linux machine.. could anyone tell me the solution please.

Thank you

There is an error that must solved:

1- make sure u leave no "enter"  in grammer file .

2- the grammer file is not the one is used in here, so u have to complete it , for me I used 

S : NS_B CALL NAME_LOOP NS_E

NAME_LOOP: NAME_LOOP NAME

NAME_LOOP: NAME

S : NS_B DIAL DIGIT_LOOP NS_E

DIGIT_LOOP: DIGIT_LOOP DIGIT

DIGIT_LOOP: DIGIT

with no enter in between lines and the error is solved

this is a beginner's solution to  a problem i faced with mkdfa.pl sample

hi

When I tried to run this command : HSLab any_name

I got this result

C:\htk\bin.win32>HSLab any_name

  ERROR [+6870]  MakeXGraf: Not compiled with X11 support: use HGraf.X.c

 FATAL ERROR - Terminating program HSLab

HGraf

+6870 X11 error

Ensure that the DISPLAY variable is set and that the X11 window system is configured

correctly.

What does the result mean:

C:\htk\bin.win32>HDMan -m -w wlist.txt -n monophones1 -l dlog newdict.txt beep/b

eep-2.0

  ERROR [+1452]  ReadDictProns: word ASTRALLY out of order in dict beep/beep-2.0

 FATAL ERROR - Terminating program HDMan

E:\My work\ASR\julius\code\julius-3.5.2-quickstart-windows\bin>mkdfa.pl sample
sample.grammar has 6 rules
sample.voca    has 6 categories and 23 words
---
'/mkfa' is not recognized as an internal or external command,
operable program or batch file.
---
no .dfa or .dict file generated

ken,

i want to create grammar in windows, i didn't install Julian, i just download and copy to my drive C in windows

and my problem when i compile my grammar is not find folder dfa_minimize, i know dfa_minimize not in one folder with mkdfa folder.

my scipt in cmd is

C:\julius-4.1.4\julius-4.1.4\gramtools\mkdfa>perl mkdfa.pl.in sample

sample.grammar has 2 rules

sample.voca    has 5 categories and 32 words

---

Warning: dfa_minimize not found in the same place as mkdfa.pl

Warning: no minimization performed

'mkfa' is not recognized as an internal or external command,

operable program or batch file.

---

no .dfa or .dict file generated

can you help me?

As many of you already may know, given a .grammar and a .voca we can see some examples of utterances the system will be able to recognize, thanks to the generate tool.

While there's no prob with tabs in the .voca file, I experienced some problem when you use tab instead of spaces inside the .grammar file.

Just to let you know that if you have .grammar a

S       :     NS_B     SOMETHING SOMETHING2 NS_E
...

and you expect

<s> SOMETHING SOMETHING2 </s>

but get output like

SOMETHING <s> SOMETHING2 </s>

or whatsoever case of misplaced silence tag <s> and </s>, the problem may be because of the tabs between NS_B and something and something2. Or this is what i experienced.

hope will help.

M

Can I use eSpeak to translate words into phonemes?

  i followed the steps but it says

            can not open "sample.grammar' at /julius/bin/mkdfa.pl line 66

check if there were unrecognized chatacters in files "sample.grammar" and "sample.voca" by type "vi *" uder linux.

I've got the same problems.But now solved. 

for   Lexical mistake

u don't have to put "Enter" between BNF grammar 

If you get this error when trying to compile your grammar and voca files: 

$ mkdfa.pl sample
sample.grammar has 3 rules
sample.voca    has 6 categories and 18 words
---
Now parsing grammar file
"rror:       Lexical mistake "
---
no .dfa or .dict file generated

use the Cygwin "dos2unix" command in the same directory as your grammar and voca files, for example:

$ cd temp

$ ls
sample.grammar  sample.voca

$ dos2unix *.*
sample.grammar: done.
sample.term: done.
sample.voca: done.

$ mkdfa.pl sample
sample.grammar has 3 rules
sample.voca    has 6 categories and 18 words
---
Now parsing grammar file
Now modifying grammar to minimize states[-1]
Now parsing vocabulary file
Now making nondeterministic finite automaton[6/6]
Now making deterministic finite automaton[6/6]
Now making triplet list[6/6]
---
generated: sample.dfa sample.term sample.dict

$ ls
sample.dfa  sample.dict  sample.grammar  sample.term  sample.voca

Ken 

hi,

I got the problem about phones.In the first step , I shoud use some phonemes to buildup a word,but how do I know which i shoud use and how to use this phonemes.

thanx~

Hi,

I have not been able to get the expected result after compilation of the grammer file.I do not know what the problem is as I followed the steps carefully.

It gives this:

sample.grammar has 3 rules
sample.voca    has 6 categories and 18 words

 Now parsing grammar file
"rror:       Lexical mistake "
---
no .dfa or .dict file generated

Help will be appreciated. 

The content of sample.grammar is not included in the tutorial of Step 1 Task Grammar.