Weird Abbreviation Processing Mid-word?


Clement Chou
 

Anyone else using Code Factory's Eloquence and Vocalizer getting
abbreviations incorrectly being processed? For example, look at the
text below.

The original says: Line 6 Marketplace
But Code Factory's Eloquence translates this as: Line March 6th ketplace

Is there a way to delete these dictionary definitions so I just get it
the way it's supposed to be? Thanks.


 

Have you checked your own NVDA dictionaries?   This one is so bizarre that it seems almost impossible for it to be the synth.

But if it is the synth, trying to work around it with dictionary entries would be a stopgap measure.  The synth maker should be notified, as there is no conceivable world where the line you give as original should be rendered in speech as you report.
--

Brian - Windows 10, 64-Bit, Version 21H1, Build 19043  

The ignorance of one voter in a democracy impairs the security of all.

         ~ John F. Kennedy

 


Gene
 

I’d be interested if others have this problem.  It is so odd that it may be caused by something that isn’t a general problem.
 
Gene

-----Original Message-----
Sent: Sunday, October 10, 2021 11:55 AM
Subject: Re: [nvda] Weird Abbreviation Processing Mid-word?
 
Have you checked your own NVDA dictionaries?   This one is so bizarre that it seems almost impossible for it to be the synth.

But if it is the synth, trying to work around it with dictionary entries would be a stopgap measure.  The synth maker should be notified, as there is no conceivable world where the line you give as original should be rendered in speech as you report.
--

Brian - Windows 10, 64-Bit, Version 21H1, Build 19043  

The ignorance of one voter in a democracy impairs the security of all.

         ~ John F. Kennedy

 


Dennis Clark
 

Hello Clement,

I also get the same result using Eloquence from Kurzweil. This has been true for me for the past year or so with NVDA. I haven't a way to cure the problem. Hopefully someone here will have some ideas. Thanks for posting.

Dennis

On 10/10/2021 7:25 AM, Clement Chou wrote:
Anyone else using Code Factory's Eloquence and Vocalizer getting
abbreviations incorrectly being processed? For example, look at the
text below.

The original says: Line 6 Marketplace
But Code Factory's Eloquence translates this as: Line March 6th ketplace

Is there a way to delete these dictionary definitions so I just get it
the way it's supposed to be? Thanks.




 

To be honest while not exactly that issue, I can get weird issues from time to time on here, or other systems I run.

Some persist after a restart, some will persist till I reinstall either the program or windows.

And in a lot of cases they never make it online.

Or they can be completely unrelated.

I once had an issue on a system I was working on which looked like mangled drivers, a virus and a windows issue.

The sound drivers were fixed, a windows issue could be fixed if it was that bad with a reload.

But in the end it boiled down to a defective wifi access card.

Most of the issues went away with its replacement.

So yeah check everything including the buttons on the case I guess.




On 11/10/2021 8:13 am, Gene wrote:
I’d be interested if others have this problem.  It is so odd that it may be caused by something that isn’t a general problem.
 
Gene
-----Original Message-----
Sent: Sunday, October 10, 2021 11:55 AM
Subject: Re: [nvda] Weird Abbreviation Processing Mid-word?
 
Have you checked your own NVDA dictionaries?   This one is so bizarre that it seems almost impossible for it to be the synth.

But if it is the synth, trying to work around it with dictionary entries would be a stopgap measure.  The synth maker should be notified, as there is no conceivable world where the line you give as original should be rendered in speech as you report.
--

Brian - Windows 10, 64-Bit, Version 21H1, Build 19043  

The ignorance of one voter in a democracy impairs the security of all.

         ~ John F. Kennedy

 


 

Well have you tried the code factory eloquence sapi or whatever?


I know they do keep it more or less updated a little more often.

On 11/10/2021 9:12 am, Dennis Clark wrote:
Hello Clement,

I also get the same result using Eloquence from Kurzweil. This has been true for me for the past year or so with NVDA. I haven't a way to cure the problem. Hopefully someone here will have some ideas. Thanks for posting.

Dennis




On 10/10/2021 7:25 AM, Clement Chou wrote:
Anyone else using Code Factory's Eloquence and Vocalizer getting
abbreviations incorrectly being processed? For example, look at the
text below.

The original says: Line 6 Marketplace
But Code Factory's Eloquence translates this as: Line March 6th ketplace

Is there a way to delete these dictionary definitions so I just get it
the way it's supposed to be? Thanks.







Gene
 

I’ve been experimenting and the bug occurs in The following circumstance, which makes it clear what the problem is.  If I write a number, I used 1 then space, then write capital Mar and that is all I write, March first is spoken when I read that text on the line.  What is happening in the example given in the original message is that a number, followed by mar, whether more is written or not, results in this misreading.  The abbbreviation or short form is applied whether it occurs in a full word or by itself.  The letter m must be capitalized for it to be treated as an abbreviation for March and the date. 
 
Here is another example. 
1 Mayflower
 
Gene

-----Original Message-----
Sent: Sunday, October 10, 2021 3:12 PM
Subject: Re: [nvda] Weird Abbreviation Processing Mid-word?
 
Hello Clement,

I also get the same result using Eloquence from Kurzweil. This has been
true for me for the past year or so with NVDA. I haven't a way to cure
the problem. Hopefully someone here will have some ideas. Thanks for
posting.

Dennis




On 10/10/2021 7:25 AM, Clement Chou wrote:
> Anyone else using Code Factory's Eloquence and Vocalizer getting
> abbreviations incorrectly being processed? For example, look at the
> text below.
>
> The original says: Line 6 Marketplace
> But Code Factory's Eloquence translates this as: Line March 6th ketplace
>
> Is there a way to delete these dictionary definitions so I just get it
> the way it's supposed to be? Thanks.
>
>
>
>
>




 
It is trying to say “may first, Mayflower.”

I would think that this bug doesn’t occur often when reading.  If igt occurs more than rarely for certain kinds of reading, I’d be curious to know.
 
Gene


 

Gene,

Then that pretty conclusively proves it's a synth bug, and the folks that make that synth should fix it.  I can't think of a single instance, ever, where a month abbreviation does not get separated from the following text by white space (even when a period is used).

Great detective work.
--

Brian - Windows 10, 64-Bit, Version 21H1, Build 19043  

The ignorance of one voter in a democracy impairs the security of all.

         ~ John F. Kennedy

 


Sarah k Alawami
 

No, it read that abbreviation
Incorrectly as well.

-----Original Message-----
From: nvda@nvda.groups.io <nvda@nvda.groups.io> On Behalf Of Shaun Everiss
Sent: Sunday, October 10, 2021 1:29 PM
To: nvda@nvda.groups.io
Subject: Re: [nvda] Weird Abbreviation Processing Mid-word?

Well have you tried the code factory eloquence sapi or whatever?


I know they do keep it more or less updated a little more often.



On 11/10/2021 9:12 am, Dennis Clark wrote:
Hello Clement,

I also get the same result using Eloquence from Kurzweil. This has
been true for me for the past year or so with NVDA. I haven't a way to
cure the problem. Hopefully someone here will have some ideas. Thanks
for posting.

Dennis




On 10/10/2021 7:25 AM, Clement Chou wrote:
Anyone else using Code Factory's Eloquence and Vocalizer getting
abbreviations incorrectly being processed? For example, look at the
text below.

The original says: Line 6 Marketplace But Code Factory's Eloquence
translates this as: Line March 6th ketplace

Is there a way to delete these dictionary definitions so I just get
it the way it's supposed to be? Thanks.








Sarah k Alawami
 

Now, to see if they really  care. Has anyone contacted them about this oversite?

 

From: nvda@nvda.groups.io <nvda@nvda.groups.io> On Behalf Of Brian Vogel
Sent: Sunday, October 10, 2021 2:26 PM
To: nvda@nvda.groups.io
Subject: Re: [nvda] Weird Abbreviation Processing Mid-word?

 

Gene,

Then that pretty conclusively proves it's a synth bug, and the folks that make that synth should fix it.  I can't think of a single instance, ever, where a month abbreviation does not get separated from the following text by white space (even when a period is used).

Great detective work.
--

Brian - Windows 10, 64-Bit, Version 21H1, Build 19043  

The ignorance of one voter in a democracy impairs the security of all.

         ~ John F. Kennedy

 


 

On Sun, Oct 10, 2021 at 05:28 PM, Sarah k Alawami wrote:
Now, to see if they really  care.
-
And they very well may not.  But, as I've said many times about many pieces of software:  They can't fix what has never been identified (or stumbled upon by them - which is a form of identification) as a problem.

And, sometimes, even if "they" care, quite a bit, fix time frames are directly dependent on both the complexity of the work required and where a given fix falls in their hierarchy of work for both new features and fixes.  Some stuff that's really simple to fix is done very quickly while some stuff that appears to be very simple to fix really isn't, or there are tons of higher priority issues/new features.
--

Brian - Windows 10, 64-Bit, Version 21H1, Build 19043  

The ignorance of one voter in a democracy impairs the security of all.

         ~ John F. Kennedy

 


Gene
 

Thank you. 
 
As a point of interest, I tested the problem with Openbook 9, which uses Eloquence and it exists there as well. 
 
Gene

-----Original Message-----
Sent: Sunday, October 10, 2021 4:26 PM
Subject: Re: [nvda] Weird Abbreviation Processing Mid-word?
 
Gene,

Then that pretty conclusively proves it's a synth bug, and the folks that make that synth should fix it.  I can't think of a single instance, ever, where a month abbreviation does not get separated from the following text by white space (even when a period is used).

Great detective work.
--

Brian - Windows 10, 64-Bit, Version 21H1, Build 19043  

The ignorance of one voter in a democracy impairs the security of all.

         ~ John F. Kennedy

 


Sarah k Alawami
 

Has anyone contacted them about this? If so what was their response if any?

 

 

From: nvda@nvda.groups.io <nvda@nvda.groups.io> On Behalf Of Brian Vogel
Sent: Sunday, October 10, 2021 3:09 PM
To: nvda@nvda.groups.io
Subject: Re: [nvda] Weird Abbreviation Processing Mid-word?

 

On Sun, Oct 10, 2021 at 05:28 PM, Sarah k Alawami wrote:

Now, to see if they really  care.

-
And they very well may not.  But, as I've said many times about many pieces of software:  They can't fix what has never been identified (or stumbled upon by them - which is a form of identification) as a problem.

And, sometimes, even if "they" care, quite a bit, fix time frames are directly dependent on both the complexity of the work required and where a given fix falls in their hierarchy of work for both new features and fixes.  Some stuff that's really simple to fix is done very quickly while some stuff that appears to be very simple to fix really isn't, or there are tons of higher priority issues/new features.
--

Brian - Windows 10, 64-Bit, Version 21H1, Build 19043  

The ignorance of one voter in a democracy impairs the security of all.

         ~ John F. Kennedy

 


Clement Chou
 

Pretty sure I haven't been the first one to experience this, so others
probably have. I will contact them as well, just in case, but I'm not
holding my breath. It's ironic how alternative versions of Eloquence
don't have this issue. Thanks for the detective work Jean. I'm not
surprised that it's a problem with this version of eloquence..
codefactory makes a lot of assumptions about how people want things
read. lol

On 10/10/21, Sarah k Alawami <marrie12@gmail.com> wrote:
Has anyone contacted them about this? If so what was their response if any?





From: nvda@nvda.groups.io <nvda@nvda.groups.io> On Behalf Of Brian Vogel
Sent: Sunday, October 10, 2021 3:09 PM
To: nvda@nvda.groups.io
Subject: Re: [nvda] Weird Abbreviation Processing Mid-word?



On Sun, Oct 10, 2021 at 05:28 PM, Sarah k Alawami wrote:

Now, to see if they really care.

-
And they very well may not. But, as I've said many times about many pieces
of software: They can't fix what has never been identified (or stumbled
upon by them - which is a form of identification) as a problem.

And, sometimes, even if "they" care, quite a bit, fix time frames are
directly dependent on both the complexity of the work required and where a
given fix falls in their hierarchy of work for both new features and fixes.
Some stuff that's really simple to fix is done very quickly while some stuff
that appears to be very simple to fix really isn't, or there are tons of
higher priority issues/new features.
--

Brian - Windows 10, 64-Bit, Version 21H1, Build 19043

The ignorance of one voter in a democracy impairs the security of all.

~ John F. Kennedy











Jackie
 

I certainly have similar problems w/Eloquence from Kurzweil, so I
don't think it's specific to CodeFactory. & yeah--it's a pig!

On 10/10/21, Clement Chou <chou.clement@gmail.com> wrote:
Pretty sure I haven't been the first one to experience this, so others
probably have. I will contact them as well, just in case, but I'm not
holding my breath. It's ironic how alternative versions of Eloquence
don't have this issue. Thanks for the detective work Jean. I'm not
surprised that it's a problem with this version of eloquence..
codefactory makes a lot of assumptions about how people want things
read. lol

On 10/10/21, Sarah k Alawami <marrie12@gmail.com> wrote:
Has anyone contacted them about this? If so what was their response if
any?





From: nvda@nvda.groups.io <nvda@nvda.groups.io> On Behalf Of Brian Vogel
Sent: Sunday, October 10, 2021 3:09 PM
To: nvda@nvda.groups.io
Subject: Re: [nvda] Weird Abbreviation Processing Mid-word?



On Sun, Oct 10, 2021 at 05:28 PM, Sarah k Alawami wrote:

Now, to see if they really care.

-
And they very well may not. But, as I've said many times about many
pieces
of software: They can't fix what has never been identified (or stumbled
upon by them - which is a form of identification) as a problem.

And, sometimes, even if "they" care, quite a bit, fix time frames are
directly dependent on both the complexity of the work required and where
a
given fix falls in their hierarchy of work for both new features and
fixes.
Some stuff that's really simple to fix is done very quickly while some
stuff
that appears to be very simple to fix really isn't, or there are tons of
higher priority issues/new features.
--

Brian - Windows 10, 64-Bit, Version 21H1, Build 19043

The ignorance of one voter in a democracy impairs the security of all.

~ John F. Kennedy














--
Subscribe to a WordPress for Newbies Mailing List by sending a message to:
wp4newbs-request@freelists.org with 'subscribe' in the Subject field OR by
visiting the list page at http://www.freelists.org/list/wp4newbs
& check out my sites at www.brightstarsweb.com & www.mysitesbeenhacked.com


Gene
 

It isn’t code fac gtory and I dubt the behavior of other synthesizers is due to Code Factory or largely not.  As I said, the problem occurs in Openbook, which uses Eloquence.  I tested with a very old version of JAWS and the problem doesn’t exist there.  I almost never use JAWS so I havedn’t kept it current.  But my test raises the question of whether the behavior occurs in currengt JAWS versions.  In other wordss, was it introduced into Eloquence in a later version or has JAWS compensated for the behavior.
 
Whatever the case, you might as well let Code Factory know.  You may want to write to whomever owns Eloquence now but it is such old software that I doubt igt is bbeing further developed or corrected.
 
Gene

-----Original Message-----
Sent: Sunday, October 10, 2021 7:54 PM
Subject: Re: [nvda] Weird Abbreviation Processing Mid-word?
 
Pretty sure I haven't been the first one to experience this, so others
probably have. I will contact them as well, just in case, but I'm not
holding my breath. It's ironic how alternative versions of Eloquence
don't have this issue. Thanks for the detective work Jean. I'm not
surprised that it's a problem with this version of eloquence..
codefactory makes a lot of assumptions about how people want things
read. lol

On 10/10/21, Sarah k Alawami <marrie12@...> wrote:
> Has anyone contacted them about this? If so what was their response if any?
>
>
>
>
>
> From: nvda@nvda.groups.io <nvda@nvda.groups.io> On Behalf Of Brian Vogel
> Sent: Sunday, October 10, 2021 3:09 PM
> To: nvda@nvda.groups.io
> Subject: Re: [nvda] Weird Abbreviation Processing Mid-word?
>
>
>
> On Sun, Oct 10, 2021 at 05:28 PM, Sarah k Alawami wrote:
>
> Now, to see if they really  care.
>
> -
> And they very well may not.  But, as I've said many times about many pieces
> of software:  They can't fix what has never been identified (or stumbled
> upon by them - which is a form of identification) as a problem.
>
> And, sometimes, even if "they" care, quite a bit, fix time frames are
> directly dependent on both the complexity of the work required and where a
> given fix falls in their hierarchy of work for both new features and fixes.
> Some stuff that's really simple to fix is done very quickly while some stuff
> that appears to be very simple to fix really isn't, or there are tons of
> higher priority issues/new features.
> --
>
> Brian - Windows 10, 64-Bit, Version 21H1, Build 19043
>
> The ignorance of one voter in a democracy impairs the security of all.
>
>          ~ John F. Kennedy
>
>
>
>
>
>
>
>
>
>
>





Clement Chou
 

I'm pretty sure it's specific versions of eloquence, Codefactory or
not. I just tried it with IBM TTS, and I'm guessing that has different
dictionaries because that's technically a different synthesizer? No
problems with it there.

On 10/10/21, Gene <gsasner@gmail.com> wrote:
It isn’t code fac gtory and I dubt the behavior of other synthesizers is due
to Code Factory or largely not. As I said, the problem occurs in Openbook,
which uses Eloquence. I tested with a very old version of JAWS and the
problem doesn’t exist there. I almost never use JAWS so I havedn’t kept it
current. But my test raises the question of whether the behavior occurs in
currengt JAWS versions. In other wordss, was it introduced into Eloquence
in a later version or has JAWS compensated for the behavior.

Whatever the case, you might as well let Code Factory know. You may want to
write to whomever owns Eloquence now but it is such old software that I
doubt igt is bbeing further developed or corrected.

Gene
-----Original Message-----
From: Clement Chou
Sent: Sunday, October 10, 2021 7:54 PM
To: nvda@nvda.groups.io
Subject: Re: [nvda] Weird Abbreviation Processing Mid-word?

Pretty sure I haven't been the first one to experience this, so others
probably have. I will contact them as well, just in case, but I'm not
holding my breath. It's ironic how alternative versions of Eloquence
don't have this issue. Thanks for the detective work Jean. I'm not
surprised that it's a problem with this version of eloquence..
codefactory makes a lot of assumptions about how people want things
read. lol

On 10/10/21, Sarah k Alawami <marrie12@gmail.com> wrote:
Has anyone contacted them about this? If so what was their response if
any?





From: nvda@nvda.groups.io <nvda@nvda.groups.io> On Behalf Of Brian Vogel
Sent: Sunday, October 10, 2021 3:09 PM
To: nvda@nvda.groups.io
Subject: Re: [nvda] Weird Abbreviation Processing Mid-word?



On Sun, Oct 10, 2021 at 05:28 PM, Sarah k Alawami wrote:

Now, to see if they really care.

-
And they very well may not. But, as I've said many times about many
pieces
of software: They can't fix what has never been identified (or stumbled
upon by them - which is a form of identification) as a problem.

And, sometimes, even if "they" care, quite a bit, fix time frames are
directly dependent on both the complexity of the work required and where
a
given fix falls in their hierarchy of work for both new features and
fixes.
Some stuff that's really simple to fix is done very quickly while some
stuff
that appears to be very simple to fix really isn't, or there are tons of
higher priority issues/new features.
--

Brian - Windows 10, 64-Bit, Version 21H1, Build 19043

The ignorance of one voter in a democracy impairs the security of all.

~ John F. Kennedy



















Gene
 

It sounds very similar but it is a different synthesizer.
 
Gene

-----Original Message-----
Sent: Sunday, October 10, 2021 9:11 PM
Subject: Re: [nvda] Weird Abbreviation Processing Mid-word?
 
I'm pretty sure it's specific versions of eloquence, Codefactory or
not. I just tried it with IBM TTS, and I'm guessing that has different
dictionaries because that's technically a different synthesizer? No
problems with it there.

On 10/10/21, Gene <gsasner@...> wrote:
> It isn’t code fac gtory and I dubt the behavior of other synthesizers is due
> to Code Factory or largely not.  As I said, the problem occurs in Openbook,
> which uses Eloquence.  I tested with a very old version of JAWS and the
> problem doesn’t exist there.  I almost never use JAWS so I havedn’t kept it
> current.  But my test raises the question of whether the behavior occurs in
> currengt JAWS versions.  In other wordss, was it introduced into Eloquence
> in a later version or has JAWS compensated for the behavior.
>
> Whatever the case, you might as well let Code Factory know.  You may want to
> write to whomever owns Eloquence now but it is such old software that I
> doubt igt is bbeing further developed or corrected.
>
> Gene
> -----Original Message-----
> From: Clement Chou
> Sent: Sunday, October 10, 2021 7:54 PM
> To: nvda@nvda.groups.io
> Subject: Re: [nvda] Weird Abbreviation Processing Mid-word?
>
> Pretty sure I haven't been the first one to experience this, so others
> probably have. I will contact them as well, just in case, but I'm not
> holding my breath. It's ironic how alternative versions of Eloquence
> don't have this issue. Thanks for the detective work Jean. I'm not
> surprised that it's a problem with this version of eloquence..
> codefactory makes a lot of assumptions about how people want things
> read. lol
>
> On 10/10/21, Sarah k Alawami <marrie12@...> wrote:
>> Has anyone contacted them about this? If so what was their response if
>> any?
>>
>>
>>
>>
>>
>> From: nvda@nvda.groups.io <nvda@nvda.groups.io> On Behalf Of Brian Vogel
>> Sent: Sunday, October 10, 2021 3:09 PM
>> To: nvda@nvda.groups.io
>> Subject: Re: [nvda] Weird Abbreviation Processing Mid-word?
>>
>>
>>
>> On Sun, Oct 10, 2021 at 05:28 PM, Sarah k Alawami wrote:
>>
>> Now, to see if they really  care.
>>
>> -
>> And they very well may not.  But, as I've said many times about many
>> pieces
>> of software:  They can't fix what has never been identified (or stumbled
>> upon by them - which is a form of identification) as a problem.
>>
>> And, sometimes, even if "they" care, quite a bit, fix time frames are
>> directly dependent on both the complexity of the work required and where
>> a
>> given fix falls in their hierarchy of work for both new features and
>> fixes.
>> Some stuff that's really simple to fix is done very quickly while some
>> stuff
>> that appears to be very simple to fix really isn't, or there are tons of
>> higher priority issues/new features.
>> --
>>
>> Brian - Windows 10, 64-Bit, Version 21H1, Build 19043
>>
>> The ignorance of one voter in a democracy impairs the security of all.
>>
>>          ~ John F. Kennedy
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>
>
>
>
>
>
>
>
>
>





Mohamed
 

Here are some dictionary entries to work around that. For these entries, when adding them to the NVDA dictionary, set the type to regular expression.
Pattern.
\b(\d+) (Jan|Feb|Mar|Apr|May|Jun|Jul|Aug|Sep|Sept|Oct|Nov|Dec)([a-z]+)

Replacement.
\1  \2\3
And here's an expression to avoid breaking legitimate dates that are formatted this way.
Pattern.
\b(\d+) (January|February|March|April|May|June|July|August|September|October|November|December)\b
Replacement.
\1 \2

On 10/10/2021 10:25 AM, Clement Chou wrote:
Anyone else using Code Factory's Eloquence and Vocalizer getting
abbreviations incorrectly being processed? For example, look at the
text below.

The original says: Line 6 Marketplace
But Code Factory's Eloquence translates this as: Line March 6th ketplace

Is there a way to delete these dictionary definitions so I just get it
the way it's supposed to be? Thanks.




Pele West
 

Hi Everyone

I have had this problem of Number and the Month for ages. Someone on this list, and I apologise that I can't remember who it was, suggested I used regular expressions in the default dictionary to solve it.

I am sorry if I have misunderstood the problem, but thanks, anyway, to the person who suggested it to me.

Here are the entries:

1. For first 3 letters of month without a space

(\d+) (Jan|Feb|Mar|Apr|May|Jun|Jul|Aug|Sept|Oct|Nov|Dec)([a-z]+) \1  \2\3    0    1

2. For first 3 letters of the month followed by space.

(\d+) (Jan |Feb |Mar |Apr |May |Jun |Jul |Aug |Sept |Oct |Nov |Dec )([a-z]+)    \1  \2\3    0    1

Pele West