Closed gavinsimpson closed 9 years ago
(This is failing as expected given that we're now testing more of the package in the unit tests...)
One of the variable names in the API has changed names somehow, the API documentation lists CollUnitType
as one of the variables returned by the dataset
command, but it's coming through as CollType
now. This seems to be causing almost all of the failures.
Ha, really bad timing, somethings going on with the whole API. . . Even the documented examples online aren't working.
Phew!
Yeah, I couldn't figure it out, and then went to the most basic level, and lo and behold. . .
They must have started doing stuff as I was running my checks because (I think) all the examples ran without error it was just the stuff in test_neotoma()
...
Yes, that's what I think happened. I was thinking it might be useful to actually pass some of the API examples directly so that we can test whether or not it's the API or the functions.
On Fri, May 1, 2015 at 2:14 PM, Gavin Simpson notifications@github.com wrote:
They must have been doing stuff as I was running my checks because (I think) all the examples ran without error it was just the stuff in test_neotoma()...
— Reply to this email directly or view it on GitHub https://github.com/ropensci/neotoma/pull/177#issuecomment-98209384.
Yeah, it's all bonkers. I added one line to get_dataset
(see commit 96c31fd4a8135304bdcda12f70445f41e82ee0b2) and that should fix everything (plus the API side fixes that have just happened.
Try running the tests now, they should all work.
This uncomments all bar one unit test. Running this shows lots of failures that we need to track down.