NVDA e-Speak Text-to-Speech Project Update (March 2014)

Under the NVDA project, we have been working in the past year on issues relating to Excel, Power Point and MS Outlook. The issues which have been fixed are:

• 3680 : Read all formula in excel sheet (NVDA+f7).
• 3681 : Read all comments in excel sheet (NVDA+f7).
• Finished implementation for 'viewing modes' supported like Cells-Flat, Row-Cells, Column-Cells, etc.
• Pressing NVDA+f7 will print a dialog box. This has a choice to show cells with comment or formula. Pressing 'Enter' at any selection in the tree will close dialog and select cell or area.
• 3231: NVDA incorrectly speaks previously entered but later deleted characters when the space bar is pressed in PowerPoint.
• #3538: Office 2010, NVDA now reads the selected symbols in the insert|symbols dialog box in Word and Excel 2010.
• #2921: In Excel, NVDA will say 'has comment' if the cell has a comment in it.
• #3043 : In Excel, Ctrl+A doesn't notify the change in selection.
• #1938 : In word 2007 and 2010, NVDA does not automatically speak the error text in the spell checker dialog after the first time the dialog is launched if the focus is already in the error text field.
• Ctrl + Up and Down arrows in Word now skip blank paragraphs.
• #3431: Spell check in Word 2010 now reads the correct context error text instead of reading all bolded text in the context.
• #3290: Bulleted text is now correctly read with Ctrl+Up and Down arrows.
• #1686: Protected documents are now read in Word 2010.
• #649: Status bar in Word Office 2010 is now read.
• #2816: Auto complete suggested contacts in Outlook 2010 are now read.
• 2047: Language detection in Word is available now.

Some of these fixes are available in NVDA and can be downloaded from www.nvaccess.org. The info about these fixes can be checked in what’s new section of the application at the website.

We have also been working on some documentation for the project, primarily FAQs to cater to the community’s needs for users switching from Jaws to NVDA which is available at http://code.google.com/p/saksham-projects/wiki/Tutorials.

The project wiki is available at http://code.google.com/p/saksham-projects/wiki/WikiContrib.

We are also working on the development of e-speak text-to-speech synthesisers for the following languages:

1. Hindi,
2. Bengali,
3. Telugu,
4. Malayalam,
5. Sindhi,
6. Punjabi,
7. Gujrati,
8. Oriya,
9. Asamese and
10. Manipuri.

They are all in different stages of development. We are near to closing our work on Gujrati this month. The issues which we have worked on are:

1. Introduction of Gujarati in e-speak.
2. The following issues were identified by the reviewer and closed:

• Character support for Gujarati language was introduced.
• Introduced rules for Gujarati language.
• Gujarati numbers have been introduced and set according to the feedback received.
• Several other changes made.
• Fixed the R sound which was sounding inappropriate when at the end of a word.Fixed the number sound 200 according to the feedback received.
• Problems with half characters where the half characters come before the same full character have been fixed.
• e-Speak could not join sounds of these two characters: ka, ch, tta, ta.
• The character name of the character [anna] has been modified according to the feedback received.
• Number 2 has been modified when it comes as the third character from the right.
• The name of two characters: [Hnna], [ ણ ], [adda]   [ ડ઼ ] was modified according to the feedback received.
• The phoneme sound of letter [Gnna]  [ જ્ઞ ] has been modified according to feedback received.
• Phoneme [ra] [ર] has been improved because at the end of a Word it was pronounced wrongly.
• When [Va] [વ]  comes at the end of the word the matra was sounding wrong. This has been fixed.
• Incorrect pronnounciations: All words which ended with "Va" like Sarva Maanava Daanava Sarva Kharava   Jakhava, etc., have been fixed.