Hi imene,

The HTK tools expect the sort order of a pronunciation dictionary to use Perl's sort order.  Here is an example of a script that uses Perl's sort that might help:

perlsort.pl

Ken

--- (Edited on 7/21/2008 3:05 pm [GMT-0400] by kmaclean) ---

 hi ken , and execuse my bad english

I am not understood  what want to say exactly

I used the file perlsort.pl on dictionary BEPP by the command

perlsort.pl beep newbeep

but always the same error

[+1452]: Word “… “out off order in dict beep

you can mor explain me ..pleazzzzzzzzzz 

--- (Edited on 7/22/2008 3:02 pm [GMT-0500] by Visitor) ---

Hi,

Maybe just try removing the "words" that are causing the problem from your version of BEEP.

Ken

--- (Edited on 7/22/2008 5:01 pm [GMT-0400] by kmaclean) ---

hey this is wot i did to make it work. I don't know if its right tho. I used text pad to do it

to solve this problem, first find and replace all "\" with "" (empty)

then find and replace all "^'" with "`" make sure the regular expression is ticked. the " ^ " tells it to look at the begining of the line for  " ' "

finaly sort it accoring to character code

--- (Edited on 8/14/2008 5:04 pm [GMT-0500] by Visitor) ---

Thanks for the update

Ken

--- (Edited on 8/23/2008 1:13 pm [GMT-0400] by kmaclean) ---

hello

thank you for your help
my  dictionnair BEPP fonction now
I  remove all the words which causes the problem

it was not easy because there were many words to remove but (il marche et ça est le plus important )

thank you another time

--- (Edited on 8/24/2008 12:41 pm [GMT-0500] by Visitor) ---

hello,sir. When i use HTKtools, i find the same problem that diction beep does not work. When I compile HDman, It tells me that the word "xx" out of the order. Would you tell me the reason of this problem and how to solve it? By the way, I do not think that delete every word casing error is a good method as there are too many words. Maybe there is other solution.

Thank you !

--- (Edited on 2/25/2010 5:35 am [GMT-0600] by Visitor) ---

>By the way, I do not think that delete every word casing error is a good

>method as there are too many words. Maybe there is other solution.

Sorry, I don't use the Beep dictionary, so I don't know of anything other than what has already been discussed in this thread.  If anyone knows of another approach, please let us know.

thanks,

Ken

--- (Edited on 2/25/2010 6:42 pm [GMT-0500] by kmaclean) ---

please my friend xxx can you explain me more these instructions

1 replace all "^'" with "`" make sure the regular expression is ticked. the " ^ " tells it to look at the begining of the line for  " ' "

(do you mean according to me to replace the '' ' '' markes located at the begining of every line by the '' ` '' marks )

2 sort it accoring to character code

thank you very much

--- (Edited on 7/24/2010 5:19 pm [GMT-0500] by bejimed) ---