Format functions that can take SCText

[nirvdrum]

I found myself wanting to write rules that use the Function action. But because special characters end up being treated oddly by Dragon (e.g., '.' -> '\dot') I'd like to use the custom SCText class. However all of the format functions are hard-coded to use the Text class directly. Perhaps the formatting functions can be augmented to take an optional tax class that be passed from the Function call as extra keywords.

[nirvdrum]

Actually, is there any harm in always using SCText here? Given the primary use case is programming it seems that may be a sane default.

[poppe1219]

I completely agree, it should be the default.

[poppe1219]

I thought it would be a slamdunk to just switch the Text class with the SCText class. Unfortunately, it requires little more work. I created a feature branch, feature/format_sctext, that contains a test with the SCText class. I will get back to this issue later, as I have a few other things that are more pressing.

[poppe1219]

I managed to get the special characters text working for the format methods that take a dictation text as a parameter. The question now is how the camel case and snake case should act around special characters. For instance, should camel case restart with a lowercase after a special character? Right now it keeps capitalizing each word: "someCamel-Text". And snake case wraps the special characters in underscores: "somesnake-_text".

[nirvdrum]

Fantastic news. I tried swapping them out and quickly hit problems with anything that didn't match your predefined patterns. E.g., anything I added to the vocabulary through Dragon proper. I'll have to try this out.

I guess for special characters like '-', maybe remove them? Is that the only case?

[poppe1219]

I think you will have to elaborate. I don't quite understand what you mean. Can you give an example that doesn't match the predefined patterns? What do you mean by removing special characters like '-'?

[nirvdrum]

Sure. I do a lot with Amazon Web Services, so I added "CloudFront" to my vocabulary, telling Dragon that it sounds like "cloud front." When passed into SCText, it looks something like CloudFront\\Cloud Front (I'm going from memory -- I'll have to check the exact text for you later).

[poppe1219]

Interesting. I didn't realize that every word from the vocabulary that differs in written form and spoken form behaves like this. It should be possible to come up with a method that can filter this out, and use that instead of a predefined list of patterns. That should be a much better, more generic solution.

[poppe1219]

I just pushed a new version of the formatting methods. This time it is a more generic method, that should sort out most special characters and custom words from the vocabulary. There are probably a few special cases that are not handled yet. Backslash is one example of a special case.

[nirvdrum]

So far this has been working out great for me. Thank you for taking care of it.

[poppe1219]

Great! I'm closing the issue. If you find any bugs, just report them as separate issues.