Crossref clients - R, Python, Ruby
Scott Chamberlain (@recology_/@ropensci)
UC Berkeley / rOpenSci
ORCID: 0000-0003-1444-9135
recology.info/talks/cr-tech
LICENSE: CC-BY 4.0
Why programmatic clients?
Why programmatic clients?
Easier, when N > a few
Why programmatic clients?
Facilitate reproducibility
Why programmatic clients?
Perhaps less error prone
Why programmatic clients?
Scalable
Why programmatic clients?
All science should be done programatically
Low level ~ build a foundation
Flickr
Prevent wheel re-building
Flickr
Users? <my guesses>
| Language | Academic | Journalist | Govt. | App/Web Devs |
|---|---|---|---|---|
| R | X | ish | X | -- |
| Python | X | -- | -- | X |
| Ruby | -- | -- | -- | X |
| Command Line | ish | -- | -- | X |
| Javascriptscienceai | ish | ? | ? | most def |
Use cases
Text mining
Get/Mine supplementary material (e.g.)
Metadata: references, citations, author disambiguation
Extract identifiers (e.g., taxonomic IDs) in full text
Content negotiation - get refs in diff. formats
Crosswalk identifiers (e.g., DOI -> Pubmed ID)
Who knows? if you build it they will come
what i've been doing
R- ropensci/rcrossrefRuby/cli- sckott/serranoPython- sckott/habanero
Feedback? Bugs? report at above locations
to the clients...
https://pypi.python.org/pypi/habanerohttps://rubygems.org/gems/serrano
https://cran.rstudio.com/web/packages/rcrossref/
Text mining
R - fulltext
ropensci/fulltext
cran/fulltext
Text mining
Ruby - textminer
sckott/textminer
rubygems.org/textminer
Client API similar-ish across clients
Functions in each client match Crossref search API routes
/works/members/funders/journals/types/licenses
E.g.: Search members
R - cr_members(98)Ruby - Serrano.members(98)Python - cr.members(98)cli - serrano members 98
Other methods
Get DOI minting agencyGet random DOIsContent negotiationGet full textGet citation count