TMT reads hidden files

[claudewillan]

If a user chooses as "input" a folder which has previously, for example, held a script worked-on in R Studio, then a hidden file is left over in the folder, which makes for some screwy outputs. Might it be possible to default ignore hidden files?

Edit: Mac OS 10.10.5, 2013 Air, R 3.4.2, R Studio 1.1.383.

[senderle]

I've been quietly gnashing my teeth about this for months.

One small question: here, hidden files would just be files starting with a ., yes? Nothing more complicated?

On Windows it will be a little different, but I'm afraid this may not be something I can extend to Windows now. (I can still ignore files starting with . but not files marked as "hidden" in the way Windows does it, which is mysterious to me.)

[senderle]

Sigh -- just now coming back to this after a few months. The problem is that mallet reads all the files in the given folder, without giving a really straightforward way to specify just some files and not others. Since there's no easy way to directly control how mallet behaves in this way, this is going to be kind of hard to fix. I'd have to work directly with Mallet's internal API, which the Topic Modeling Tool was not written (by its previous creators) to do. And I don't have time to rewrite that part! Argh...

Unless I'm mistaken...

@claudewillan do you have a way of telling mallet to ignore hidden files? If so, please share it with me!! I'm sorry to have missed it, in that case.