Bug in debug output: was skipping every other phoneme

[tyomitch]

No description provided.

[discordier]

I agree with you on the change for the debugging.

The change for the rising and falling inflection though we should discuss.
If I get you right, only the naming is the wrong way around.
See Original code here for question mark and here for periods.

Is there any reason you changed the order of the conditions instead of simply swapping the constant name in the AddInflection() calls?

[tyomitch]

The change for the rising and falling inflection though we should discuss.
If I get you right, only the naming is the wrong way around.

Yes, the values 1 and 255 were used correctly, but they were assigned wrong names.
The original SAM docs say: "The period inserts a pause and also causes the pitch to fall. The question-mark also inserts a pause, but it causes the pitch to rise."

See Original code here for question mark and here for periods.

Yep, those two comments in the C code are wrong, too.

Is there any reason you changed the order of the conditions instead of simply swapping the constant name in the AddInflection() calls?

What I did was simply swapping the constant names in the AddInflection() calls, but I can see how the diff display is somewhat confusing. If you switch the display mode from "Unified" to "Split", it becomes a bit less confusing.

[tyomitch]

See Original code here for question mark and here for periods.

Yep, those two comments in the C code are wrong, too.

I have now submitted a PR for it, too: s-macke#13

[discordier]

Released in 0.1.2