Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Ghost extra numbers while reading tables (.doc, .xls and .htm ) #1444

Closed
nvaccessAuto opened this issue Apr 6, 2011 · 8 comments
Closed
Labels
bug close/duplicate component/i18n existing localisations or internationalisation

Comments

@nvaccessAuto
Copy link

Reported by Daniel Langeron on 2011-04-06 21:37
Environment:

  • Office Word 2003 SP3
  • FireFox 3.6.13
  • NVDA 2011.1 Portable version (USB key)

Defect:

  • Misreading 3-digit numbers in a table (. doc or. htm)

Retails:

Using "eSpeak" synthesizer and English voice, within a table and applying the key commands (Ctrl + Alt + Arrow Right, Left, Up, Down or 9 of the numeric keypad disabled) leads to a correct reading of 3-digit numbers.

But, with "eSpeak" synthesizer and French voice, 3-digit numbers are misread. For example numbers 001, 123 and 123 456 are respectively vocalized in French "Mille" 001, "Mille" 123 and "Million" 123 456 (not: "Un Million" 123 456)

Read attached file " Misreading_Table.doc" for checking.

Additional notes:

1/ Ghost extra "Mille" and "Million" are not visible in the corresponding speech viewer.

2/ Reading the same file, these numbers are read correctly when mouse moves from cell to cell.

3/ Reading the same file with "sVox Pico" and English voice OR "sVox Pico" and French voice leads to the same default of ghost extra numbers (Thousand, Million for English voice and Mille, Million for French voice)

4/ Using Internet, when reading .htm table, e.g. bank statements of an account, leads to the same default of ghost extra numbers for the same configurations of synthesizers and voices.

@nvaccessAuto
Copy link
Author

Attachment Misreading_Table.doc added by Daniel Langeron on 2011-04-06 21:45
Description:
Word 2003 document

@nvaccessAuto
Copy link
Author

Comment 1 by Daniel Langeron on 2011-04-10 20:57
I fear that this defect (well-vocalized reading, but INCORRECT VALUES announcement) makes NVDA unusable by people using eSpeak and the French voice.

Simply put, reading his own bank statement accessed via the Internet, how to distinguish between, say, 123 Euros and 1 123 Euros ?

If the fix is slow to come I fear that French language users are discouraged to use NVDA.

Therefore I request that priority be changed from “minor” to “major”.

Thanks,
Daniel

@nvaccessAuto
Copy link
Author

Comment 2 by msuch (in reply to comment 1) on 2011-04-16 06:23
Replying to Daniel Langeron:

I fear that this defect (well-vocalized reading, but INCORRECT VALUES announcement) makes NVDA unusable by people using eSpeak and the French voice.

Simply put, reading his own bank statement accessed via the Internet, how to distinguish between, say, 123 Euros and 1 123 Euros ?

If the fix is slow to come I fear that French language users are discouraged to use NVDA.

Therefore I request that priority be changed from “minor” to “major”.

Thanks,

Daniel

Well, this is a problem with the way numbers can be written in french.
A space can be used as a thousandseparator.
The problem occurs when, in a spreadshit, a cell contains a 3 digits number.
Imagine that cell C2 contains 123
NVDA will send something like
C2 123
to the synthesizer which will be interpreted
c 2123
on a french synthesizer.
The solution would be to introduice a short pause between the cell name and the content which may be synthesizer depending and so difficult to implement.
You could also append a space at the end of the cell name so the synthesizer would receive 2 spaces between the cell name and the content which would break the interpretation and would give
c2 123

@nvaccessAuto
Copy link
Author

Comment 3 by jteh on 2011-04-16 09:11
Similar to #555, except in that case, there was a letter before the first number.

@nvaccessAuto
Copy link
Author

Comment 4 by msuch (in reply to comment 3) on 2011-04-16 09:22
Replying to jteh:

Similar to #555, except in that case, there was a letter before the first number.

Yess, 555 and 1444 are the same problem.

@nvaccessAuto
Copy link
Author

Comment 5 by Daniel Langeron on 2011-04-21 21:19
I understand that the ticket # 1444 evokes the same problem than the ticket # 555.
But the ticket # 555, was finally "closed" 8 weeks ago by Bernd with just the following explanation: “Closing the ticket as fixed because the problem exists no longer"
In the history of the ticket # 555 it is also said that it is a duplication of the ticket # 320.
And finally the last comment added to the ticket # 320 by Jteh 4 months ago is "there is no solution yet"

Does that mean that the ticket # 1444 will also be closed with the explanation "The problem no longer exists" or "there is no solution yet" ? I hope not !

At that point we must note that the ticket # 320 has been opened 2 years ago and yet I can attest that the problem still exists into Excel, Word and HTM tables for, at least, French and German eSpeak users !

In the NVDA 2011.2.roadmap, solving the problem of ticket # 149 ("Say Without all unnatural pauses") is "assigned" This ticket also involves the way in which NVDA communicates with the synthesizer.
My understanding is that all these tickets show that changes should be made to how NVDA transmits its information to the synthesizer so that it can correctly vocalize the text.

Would it not the right time to process all at once tickets # 149, # 320 and # 1444 (and perhaps others ) as involving the same class of problem: What is the best way for NVDA to forward the text to the synthesizer ?

Changes:
Changed title from "Ghost extra numbers while reading tables (.doc and .htm )" to "Ghost extra numbers while reading tables (.doc, .xls and .htm )"

@nvaccessAuto
Copy link
Author

Comment 6 by jteh on 2011-04-21 23:58

@nvaccessAuto
Copy link
Author

Comment 7 by Bernd on 2011-04-22 04:51
I'll reopen #555 and enhance it with french voice and other tables. Sorry for my misstake.
Closing as duplicate of #555
Changes:
Added labels: duplicate
State: closed

@nvaccessAuto nvaccessAuto added close/duplicate bug component/i18n existing localisations or internationalisation labels Nov 10, 2015
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
bug close/duplicate component/i18n existing localisations or internationalisation
Projects
None yet
Development

No branches or pull requests

1 participant