fixed a bug where the kitten function errored out on tinytest, when c…

[paulhodor]

…alled with a non-default path; fixed a typo in the help page of kitten

[eddelbuettel]

So I messed up root vs name here? That's certainly possibe.

[eddelbuettel]

The things is ... that it works as is. (Expand for log)

edd@rob:/tmp$ mkdir kitten
edd@rob:/tmp$ cd kitten/
edd@rob:/tmp/kitten$ kitten.r -h\
> ^C
edd@rob:/tmp/kitten$ kitten.r -h
Usage: kitten.r [-t TYPE] [-b] [-p] [-h] [-x] PACKAGE

-t --type TYPE      type of kitten: plain, rcpp, arma, eigen. [default: plain]
-b --bunny          install roxygen2 documentation example and roxygenize (only for plain)
-p --puppy          invoke tinytest::puppy to set up testing (only for plain)
-h --help           show this help text
-x --usage          show help and short example usage 
edd@rob:/tmp/kitten$ kitten.r -p quickcheck
Creating directories ...
Creating DESCRIPTION ...
Creating NAMESPACE ...
Creating Read-and-delete-me ...
Saving functions and data ...
Making help files ...
Done.
Further steps are described in './quickcheck/Read-and-delete-me'.

Adding pkgKitten overrides.
 >> added .gitignore file
 >> added .Rbuildignore file
 >> added tinytest support
Deleted 'Read-and-delete-me'.
Done.

Consider reading the documentation for all the packaging details.
A good start is the 'Writing R Extensions' manual.

And run 'R CMD check'. Run it frequently. And think of those kittens.


Thank you edd, for showing us some PUPPY LOVE <3

    ,-.___,-.
    \_/_ _\_/
      )O_O(
     { (_) }  W00F!
      `-^-'   
edd@rob:/tmp/kitten$ ls -l quickcheck/tests/tinytest.R 
-rw-rw-r-- 1 edd edd 94 Nov 15 12:49 quickcheck/tests/tinytest.R
edd@rob:/tmp/kitten$ ls -l quickcheck/inst/tinytest/test_quickcheck.R 
-rw-rw-r-- 1 edd edd 56 Nov 15 12:49 quickcheck/inst/tinytest/test_quickcheck.R
edd@rob:/tmp/kitten$ 

How did you call it to trigger an error? Maybe I get lucky here via pkgKitten.r from litter ?

[eddelbuettel]

Ah, I think you just answered in #19.

[eddelbuettel]

Could I ask you to add yourself to the ChangeLog file ? Format is pretty standard: name, two spaces, date, two spaces, email and then eight spaces and a * followed by the filename and one or two line change summary.

[eddelbuettel]

Thank you!

[eddelbuettel]

Thanks for bug fix!

[paulhodor]

This is way too much automation for me. I did not realize that changes to my fork get propagated to you. I had made a mistake in the changelog and was working on fixing it. I wasn't aware that you can see all that activity.

[eddelbuettel]

Fair point but I think it makes sense in the context of a pull request which links a partitcular branch of your fork to my repo. Being able to dynamically update (wilder still: I get to write to it too !!) is generally good for PRs. But I understand your surprise.

No harm done though and force-push is good for this. I usually merge as is (whereas I squash in other repos). Ok with you?

[paulhodor]

I added tests for the hello function, per #20. Looks like it went into the same pull request. I guess I should have created a new branch.

[eddelbuettel]

Looks like it went into the same pull request. I guess I should have created a new branch.

(Just playing "git doctor" here: correct. If you want that, you could do the following:

• create a new branch of your current branch, that is basically for safekeeping
• roll this branch back via eg git reset --hard HEAD^ should do, I usually reset to a sha1 as I am paranoid)
• force push the rolled back branch to restore this PR
• wait til I merge, pull, then create a new branch
• cherry pick the third commit here and on the safekeep branch into the new branch, PR)

If it is too much work we just merge this. Your call. It's kinda nice to have two separate PRs but it doesn't matter greatly.

[paulhodor]

Let's leave it as is. But thanks for the explanation. I'll remember it for next time.

[eddelbuettel]

I'll clean this up post merge. It's generally an entry per day, not per commit. So these two will get fused.