IdentifierCache provides two tools
- A voice command grammar that allows the user to speak in Programming Notation with relative ease. For example you could say “pascal loop counter” to generate the text ‘LoopCounter’
- A Visual Studio add-in to save the most recently used and created identifiers. So you could say “instert ID 1”, to insert the topmost identifier in the cache.
IdentifierCache relies on Dragon NaturallySpeaking for its speech recognition engine, NatLink as the command grammar executor, and Visual Studio .Net to host the add-in. IdentifierCache itself is written in two languages, the command grammar is a NatLink script, so is written in python, the add-in is written in VB.Net
I didn't have any time to spend working out but fixes or improvements, so I've released it as open source to SourceForge. Get it at http://sourceforge.net/projects/identifiercache
Update:
An install of the software can be found on the project homepage at http://www.speechwiki.org/SR/IdentifierCache.html and the source code can be obtained through the CVS interface.
Richard Sprague has been dropping hints about this for some time, and now Robert Brown has made the announcement that the Longhorn speech API has been included into the Avalon/indigo Beta1
I'm downloading it as I speak.
The latest release of NatLink/Vocola is available. NatLink enables command and control of Dragon NaturallySpeaking by loading the Python runtime as a global application module in a COM extention. It presents a relatively simple façade to some IUnknown only COM interfaces.
Vocola uses Perl to generate Python programs which use NatLink to allow a simple command language to extend the capabilities of Dragon NaturallySpeaking.
My own small contributions have been to package it all up as an MSI and most recently to add a nice .net WinForms interface for errors.
And on Wednesday I'll be giving a presentation on the practical use of Vocola at the NYPC Speech Sig.
The speech recognition wiki has moved from speech.squidnet.com to speechwiki.org. I changed the domain name because the old root domain had nothing to do with speech-recognition and I also took the opportunity to move to a new hosting company which I'm expecting will provide faster responsiveness and more reliable uptime.
Existing links that point to the old site should automatically direct to the new site. If you discover any errors or mistakes feel free to contact me or leave a message on the feedback page.
I've also created a new topic page to help explain the features of the wiki .