fulltext is a single interface to many sources of scholarly texts. In practice, this means only ones that are legally useable. We will support sources that require authentication on a case by case basis - that is, if more than just a few people will use it, and it's not too burdensome to include, then we can include that source.
See https://books.ropensci.org/fulltext/ for a longer form manual for using fulltext.
We currently include support for search and full text retrieval for a variety
of publishers. See ft_search() for what we include for search,
and ft_get() for what we include for full text retrieval.
The following are tasks/use cases supported:
search - ft_search()
get texts - ft_get()
get full text links - ft_links()
get abstracts - ft_abstract()
extract text from pdfs - ft_extract()
serialize to different data formats - ft_serialize()
extract certain article sections (e.g., authors) - moved to pubchunks
Beware that DOIs are not searchable via Crossref/Entrez immediately. The delay may be as much as a few days, though should be less than a day. This delay should become shorter as services improve. The point of this is that you man not find a match for a relatively new DOI (e.g., for an article published the same day). We've tried to account for this for some publishers. For example, for Crossref we search Crossref for a match for a DOI, and if none is found we attempt to retrieve the full text from the publisher directly.
Scopus: 20,000 per 7 days. See
https://dev.elsevier.com/api_key_settings.html for rate
limit information. To see what your personal rate limit details are,
request verbose HTTP request output - this will vary on the function
you are using - see the docs for the function. See the response
headers X-RateLimit-Limit, X-RateLimit-Remaining, and
X-RateLimit-Reset (your limit, those requests remaining, and UTC
date/time it will reset)
Microsoft: 10,000 per month, and 1 per second. There are no rate limit headers, sorry :(
PLOS: There are no known rate limits for PLOS, though if you do hit something let us know.
Crossref: From time to time Crossref needs to impose rate limits
to ensure that the free API is usable by all. Any rate limits that are in
effect will be advertised in the X-Rate-Limit-Limit and
X-Rate-Limit-Interval HTTP headers. This boils down to: they allow X
number of requests per some time period. The numbers can change so we
can't give a rate limit that will always be in effect. If you're curious
pass in verbose = TRUE to your function call, and you'll get headers
that will display these rate limits. See also Authentication.
Semantic Scholar: Not documented in their docs, and no response headers given. At time of this writing (2020-07-01) the rate limit is: 100 requests per 5-minutes per IP address. or 20 requests per min. Note that this rate limit may change.
BMC: BMC is integrated into Springer Publishers now,
and that API requires an API key. Get your key by signing up at
https://dev.springer.com/, then you'll get a key. Pass the key to a
named parameter key to bmcopts. Or, save your key in your .Renviron
file as SPRINGER_KEY, and we'll read it in for you, and you don't
have to pass in anything.
Scopus: Scopus requires two things: an API key and your institution must
have access. For the API key, go to https://dev.elsevier.com/index.html,
register for an account, then when you're in your account, create an API key.
Pass in as variable key to scopusopts, or store your key under the name
ELSEVIER_SCOPUS_KEY as an environment variable in .Renviron, and
we'll read it in for you. See Startup for help. For the institution access
go to a browser and see if you have access to the journal(s) you want.
If you don't have access in a browser you probably won't have access via
this package. If you aren't physically at your institution you will likely
need to be on a VPN or similar and eventually require correct proxy settings,
so that your IP address is in the range
that the two publishers are accepting for that institution.
It might be, that the API access seems to work even while
in the wrong IP range or have wrong proxy settings,
but you are not able to see the abstracts, they will be empty.
By using the currect curl options into the calls to ft_search or ft_abstracts even
the most complex proxy including authentication should work. As an example:
opts <- list(key="your-scopus-key") ft_abstract(x = dois, from = "scopus", scopusopts = opts, proxy="proxy-ip-address", proxyport=your-proxy-port, proxyuserpwd="username:password", # often the same as your windows login proxyauth=8) # ntlm - authentication
Elsevier/ScienceDirect: Elsevier ScienceDirect requires two things: an
API key and your institution must have access. For the API key,
go to https://dev.elsevier.com/index.html,
register for an account, then when you're in your account, create an API key
that is allowed to access the TDM API (must accept their TDM policy).
Pass in as variable key to elsevieropts/sciencedirectopts, or store
your key under the name ELSEVIER_TDM_KEY as an environment variable in
.Renviron, and we'll read it in for you. See Startup for help. For
the institution access
go to a browser and see if you have access to the journal(s) you want.
If you don't have access in a browser you probably won't have access via
this package. If you aren't physically at your institution you will likely
need to be on a VPN or similar so that your IP address is in the range
that the publisher is accepting for that institution.
Wiley: Replacing Crossref TDM service as of February 2021, Wiley
now requires you get a Wiley TDM key. Get one at
https://onlinelibrary.wiley.com/library-info/resources/text-and-datamining
Pass in as variable key to wileyopts, or preferably store
your key under the name WILEY_TDM_KEY as an environment variable in
.Renviron, and we'll read it in for you. See Startup for help. Some
notes about Wiley's TDM service:
They always respond initially with a redirect to a server dedicated to the serving of binary resources - fulltext takes care of this
Wiley uses rate-limiting: no more than 3 requests per second. you may get 429 errors if making too requests too rapidly
Microsoft: Get a key by creating an Azure account
then request a key for Academic Knowledge API within
Cognitive Services. Store it as an environment variable in your
.Renviron file - see Startup for help. Pass your
API key into maopts as a named element in a list like
list(key = Sys.getenv('MICROSOFT_ACADEMIC_KEY'))
Crossref: Crossref encourages requests with contact information
(an email address) and will forward you to a dedicated API cluster
for improved performance when you share your email address with them. This
is called the "Polite Pool".
https://github.com/CrossRef/rest-api-doc#good-manners--more-reliable-service
To pass your email address to Crossref via this client, store it
as an environment variable in .Renviron like
crossref_email=name@example.com, or CROSSREF_EMAIL=name@example.com.
Save the file and restart your R session. To stop sharing your email when
using rcrossref simply delete it from your .Renviron file OR to temporarily
not use your email unset it for the session
like Sys.unsetenv('crossref_email'). To be sure your in the polite pool
use curl verbose by e.g., ft_cr_links(doi = "10.5555/515151", verbose = TRUE)
Crossref TDM: TDM = "Text and Data Mining". This used to apply to just two publishers - Wiley and Elsevier - This service officially shut down at the end of 2020. For Elsevier, see the "Elsevier/ScienceDirect" section above. For Wiley, see the "Wiley" section above.
Entrez: NCBI limits users to making only 3 requests per second. But, users
who register for an API key are able to make up to ten requests per second.
Getting a key is simple; register for a "my ncbi" account then click on a
button in the account settings page. Once you have an API key, you can pass it
as the argument api_key to entrezopts in both ft_get() and ft_search().
However, we advise you use environment variables instead as they are more secure.
To do that you can set an environment variable for the current R session like
Sys.setenv(ENTREZ_KEY="yourkey") OR better yet set it in your .Renviron
or equivalent file with an entry like ENTREZ_KEY=yourkey so that it is
used across R sessions.
No authentication needed for PLOS, eLife, arxiv, biorxiv, Euro PMC
Open an issue if you run into trouble with authentication.
Let us know what you think at https://github.com/ropensci/fulltext/issues