Grabbing supplementary data from articles

[willpearse]

...this is the pull request for issue #61.

Sorry for the delay; 'tis the job season to be jolly, etc...

I've refactored etc. to make everything look a bit more like the rest of the package. Let me know what you think; I won't in any way be offended if you don't want it in. There isn't perfect overlap between what you can get SI for and what you can get text for; I plan to get around to fixing that (...) but I reasoned you would rather have something than nothing!

I've made some changes to the .Rmd files as well, but for some reason I was having a lot of trouble building them (some sort of pandocs problem getting access to a server). I'm a Linux person, so I'm certain this reflects something weird that I've done.

[sckott]

Awesome! Thanks.

On vacation now. Back next monday.

One thing I see so far. Can you use httr instead of rcurl? Don't want to introduce new dependencies.

[willpearse]

Done. I also fixed a weird build problem. Apparently I'm rubbish at RMarkdown...!

Enjoy your holiday - stop checking GitHub :p

[sckott]

looking at this now

[sckott]

please wrap examples in \dontrun{}

[willpearse]

Hopefully this is everything you asked for!

[sckott]

thanks 👍

[sckott]

@willpearse that one test is still failing, can you fix that

[willpearse]

The problem stems from my output from rcrossref::cr_works being different over the last few days. I can just refactor to fix this, but I'm concerned that I'm just confused about how this function works. Why is it that when I run the following:

cr_works(dois="not-a-DOI")

...I get something back? When I first wrote this I'm sure 'nonsense' DOIs returned with no data whatsoever, and the wrong result I get has changed in the last 24 hours.I wrote it, initially, to give two different error messages - one for when I couldn't find the right publisher, and the other for when I found the publisher but couldn't download from them.

Has something shifted on the crossref API, or am I just being stupid (I sense it's the latter!)

[sckott]

@willpearse this is a bug in the Crossref API,

This is the API call from your e.g., http://api.crossref.org/works?query=not-a-DOI which gives that same result.

Here's a related issue CrossRef/rest-api-doc#37 in the repo for their API

Hopefully they'll get that fixed soon.

So I guess a change in your test to account for that would be good.

[willpearse]

Done. Thanks for showing me the bug!

[willpearse]

👍 thanks!

[sckott]

thank you for this!