- implement Crossref's REST API updates for the public and polite pool
- change of maintainer
- move
bibtexpackage to Suggests as it has been orphaned, use it conditionally (#209) - change all uses of
tibble::tbl_dftotibble::as_tibble(#206) - additional fields now parsed in the
cr_works()output: short-container-title, references-count, is-referenced-by-count, language, content-domain, and update-to (#208)
- docs page vignette has names (#197)
query.titlefield query is no longer supported by Crossref, removed from package (#198)- fix some links in readme (#199)
- improve documentation for how to do deep paging (new section "Deep paging"), and how to use the cursor specifically (#190)
- fix error when progress used and when limit param not used (#190)
- encoding fix for the Crossref RStudio Addin (#194) (#201)
cr_funders(),cr_journals(),cr_licenses(),cr_members(),cr_prefixes(),cr_types(),cr_works()gain ability to show a progress bar when using deep pagination when usingworks=TRUE(#186) (#188)
- fix a test fixture that had non-ascii characters (#187)
cr_works()returned a tibble in the$dataslot in each case except for when a single DOI was passed to thedoiparameter. now fixed (#184) thanks @martinjhnhadley
- big Crossref Addin update (#171) all work by @haozhu233
- Async HTTP introduced for
cr_works(),cr_works_(), andcr_citation_count(). See the new parameterasync(logical) in those functions. Forcr_citation_count(), it now accepts more than 1 DOI, and the output has changed from a numeric value to a data.frame (columns:doiandcount). Withasync=TRUEforcr_works()you get a list of data.frame's; while forcr_works_()you get a list of JSON's (#121) (#160) (#182)
- package tests now using
vcrfor HTTP request/response caching (#178) (#179) - in
worksdata, now returningpublished-printandpublished-onlinefields (#181) - add new filters for
/worksroutes (whenworks = TRUE):isbn,reference_visibility,has_content_domain, andhas_domain_restriction(#176) (#177) - add new return element in
/worksroutes (whenworks = TRUE):$referencegives the references cited in the article (these are not the articles citing the target article, sorry) (#176) - improve documentation on the "polite pool". If you supply an email address Crossref will put you in the polite pool (#173) (#175) thanks @poldham
- added example to
cr_abstract()of handling many DOIs while allowing for failures without stopping progress (#174) thanks @zackbatist
- fix internal parsing of funder data in
cr_works()(#180) thanks @nicholasmfraser for the bug report - fix for
cr_citation_count(): when given a bad/invalid/malformed DOI, throw warning and give back anNA(#164) thanks for the report @chreman - Fix for instances in the package where we incorrectly were doing logical comparison's; detected via R_CHECK_LENGTH_1_LOGIC2
- The RStudio Addin now includes ability to search by article metadata, e.g., title (#148) (#152)
- you can now set a different base url for
cr_cn()(#168)
- better behavior for
cr_cn()when bibentry not valid, that is not parseable. before the package we use to parsebibtexwould stop on invalid bibentry data, but now we get around the invalid bits and give back the bibentry (#147) - updated and improved documentation for filters (#161)
- improved description of the
doisparameter (#162) thanks @ms609 - Much improved error behavior for
cr_*functions. We now give errors like404 (client error): /works/blblbl - Resource not found., which includes HTTP status code, major class of error (success, redirect, client, server), the route requested, and the error message (#163)
- fixed bug in
cr_works()in which field queries should have been possible for title and affiliation, but were not. Fixed now. (#149) cr_journals()was not correctly parsing data when more than 1 ISSN given andworksset toTRUE, fixed now (#156)cr_journals()was not correctly parsing data when no ISSN found andworksset toTRUE, fixed now (#150)cr_journals()was not correctly handling queries with multiple ISSN's andworksset toFALSE, fixed now (#151)- fixed bug in
cr_works(), and any othercr_*function that setworks=TRUE. thelicenseslot can have more than 1 result, and we were only giving the first back. fixed the parsing on this to give back all license results (#170)
- Include documentation and internals to support passing a user supplied email to Crossref to get put into higher rate limit pool (#145) thanks @njahn82
- Functions that operate on the
/worksroute gain aselectparameter to select certain fields to return (#146)
- Updated docs for additional options for the
sortparameter (#142) - Updated docs for additional options for doing field queries (the
flqparameter) (#143) - Filters: New filters added to the package, and some removed
(e.g,
publisher-name). You can see the filters with the functionsfilter_details/filter_names. Beware, some filters error sometimes with the Crossref API - they may not work, but they may, let me know at https://github.com/ropensci/rcrossref/issues or let Crossref know at https://github.com/CrossRef/rest-api-doc/issues (#136) (#139) (#141)
- All text mining functionality moved into a new package:
crminerhttps://github.com/ropensci-archive/crminer . Functions that did text mining stuff now defunct, see?rcrossref-defunct(#122) - All Crossref API requests now using
httpsinstead ofhttp(#133)
- replace
xml2::xml_find_onewithxml2::xml_find_first(#128) thanks @njahn82 - replaced
httrwithcrulfor HTTP requests (#132) - Now using markdown for documentation (#135)
- Added more documentation to
cr_journalsandcr_worksabout what the returned data fieldsbackfile_doisandcurrent_doisreally mean (#105) thanks @SteveViss
- Fix to
cr_prefixesto not fail when no results found (#130) thanks @globbestael - Fixed
cr_worksto allow queries likefacet = license:*to be passed tofacetparameter (was always allowed by Crossref, but we neglected to allow it - previously only allowed a boolean) (#129) - Fixed
cr_fundersandcr_journalsto give back facet data along with other data (#134) - Fix to
cr_*functions to check for a missing content-type headers and instead of failing, we continue anyway and try to parse data as sometimes Crossref doesn't give back a content type header at all (#127)
- Added support for field queries (see
https://github.com/CrossRef/rest-api-doc/blob/master/rest_api.md#field-queries
for information on them). Means that you can query on specific fields rather
than using
queryparameter which queries across all fields (#111) - Now using
rappdirsfor local storage and caching forcr_ft_text(#106)
- Added to man files where appropriate new 10K max value for the
offsetparameter (#126) - Added to pkg level man file new rate limit headers included,
and how users can get to those, via
config=verbose()call (#124) - Better failure modes on input parameters, still more work to do surely (#101)
- sleeping now between tests to avoid making crossref rate limit gate keepers mad (#125)
cr_searchandcr_search_freeare now defunct. They were marked deprecated in previous version, and warned of defunct, and now they are defunct. Similar functionality can be done with e.g.,cr_works()(#102)crossciteis now defunct. The functionality of this function can be achieved withcr_cn()(#82)cr_fundrefis now defunct. Crossref changed their namefundreftofunders, so we've changed our function, seecr_funders()(#83)- parameter
samplemaximum value is now 100, was previously 1000. documentation updated. (#118) - New filters
has-clinical-trial-numberandhas-abstractadded to the package, see?filtersfor help (#120)
- Added an RStudio Addin for searching for citations. See
?rcrossreffor more. Addin authored by Hao Zhu @haozhu233 (#114) - New function
cr_abstract()that tries to get an abstract via XML provided by Crossref - NOTE: an abstract is rarely available though (#116)
- Fixed bug in
cr_cn()where DOIs with no minting agency found were failing because we were previously stopping when no agency found. Now, we just assume Crossref and move on from there. (#117) thanks @dfalster ! - Fix to
cr_r()when number requested > 100. Actual fix is incr_works(). Max for sample used to be 1000, asked this on the Crossref API forum, see CrossRef/rest-api-doc#146 (#115) - Fix to
cr_journals()in internal parsing, was failing in cases whereISSNarray was of length zero
- Improved documentation for
cr_citation_count()to remove PLOS reference as the function isn't only for PLOS works (#108) - Changed use of
dplyr::rbind_all()todplyr::bind_rows()(#113)
- User-agent string is passed with every request to Crossref with
names and versions of this package, and its HTTP dependency packages:
httrandcurl(whichhttrdepends on). Will potentially be useful to Crossref to know how many requests come from this R client (#100)
cr_search()andcr_search_free()use old Crossref web services, so are now marked deprecated, and will throw a deprecation message, but can still be used. They will both be defunct inv0.6of this package (#99)
XMLreplaced withxml2(#98)httr::content()calls: all parse to text then parse content manually. in addition, encoding explicitly set toUTF-8onhttr::content()calls (#98)
- Bug fix to
cr_journals()- fix to parse correctly on some failed requests (#97) thanks @nkorf - Bug fix to
cr_fundref()/cr_funders()- parsing wasn't working correctly in all cases
Skipped v0.4 to v0.5 because of many changes - as described below.
- Support added for 'deep paging' with the newer Crossref search API. Two new params added to each function:
cursor, which accepts a cursor alphanumeric string or the special*, which indicates that you want to initiate deep paging;cursor_max, which is not in the Crossref API, but just used here in this package to indicate where to stop - otherwise, you'd get all results, even if there was 70 million, for example. A new internalR6class used to make cursor requests easy (#77) - New function
id_converter()to get a PMID from a DOI and vice versa (#49) - Added a Code of Conduct.
- New function
cr_types(), along with its low level equivalentcr_types_()for when you just want a list or json back (#92) - New suite of functions composing a low-level API for the Crossref search API. These functions only do data request, and return json or a list, and do not attempt to parse to a data.frame. The new functions:
cr_funders_(),cr_journals_(),cr_licenses_(),cr_members_(),cr_prefixes_(),cr_types_(),cr_works_(). These functions are a bit faster, and aren't subject to parsing errors in the event of a change in the Crossref API. (#93) - Added new
filter_names()andfilter_details()functions to get information on what filters are available, the expected values, and what they mean.
- Added documentation for new filter types, and added them to list of filters for
filter_names()andfilter_details()(#73) cr_funders()alias added tocr_fundref()(#74)- Made note in docs that funders without IDs don't show up on the
/fundersroute,s incr_funders()(#79) - Made note in docs that
sampleparameter ignored unlessworks=TRUE(#81) - Added note to docs that only what is returned in the API is what is searched when you search the Crossref API - that is, abstracts and full text aren't searched (#91)
cr_cn()now checks that the user supplied content-type is supported for the DOI minting agency associated with the DOI (#88) (thanks @njahn82)- Removed
.progressparameter use internally where it wasn't applicable. sampleparameter dropped fromcr_licenses().cr_works()parsing changed. We now don't attempt to flatten nested arrays, but instead give them back as data.frame's nested within the main data.frame. For example,authoroften has many entries, so we return that as a single column, but indexing to that column gives back a data.frame with a row for each author, and N number of columns. Hopefully this doesn't break too much code downstream :)- Additional text added to the package level man file (
?rcrossref) to explain: what you're actually searching when you search; deprecated and defunct functions; and explanation of high vs. low level API.
- Fix to
cr_members()to warn on error instead of stop during parsing (#68) - Fix to internal parser for
cr_works()to output links data, for full text links (#70) - Minor fix in
cr_cn()example that didn't work (#80) - Fixed parsing of
affiliationdata insideauthorobject in Crossref search API returned data (#84) - Fixed parsing of funder
awardslot in Crossref search API returned data (#90)
crosscite()deprecated, will be removed in a future version of this package (#78)cr_fundref()now has a deprecated message, and will be removed in the next version (#74)
- Added new function
crosscite()to work with the Citeproc service (http://crosscite.org/citeproc/) (#60)
- Fixed problems related to
httrv1(#65) - Import all non-base R functions (#64)
- The agency route was down used by the
cr_agency()function, back up and fixed now (#63)
- New function
extract_pdf()to extract text from pdfs - New function
cr_ft_links()to get links for full text content of an article (#10) - New function
cr_ft_text()to get links for full text content of an article. In addition,cr_ft_pdf(),cr_ft_plain(), andcr_ft_xml()are convenience functions that will get the format pdf, plain text, or xml, respectively. You can of course specify format in thecr_ft_text()function with thetypeparameter (#10) (#42)
- Filled out more tests (#45)
- No longer assign queried doi to the
data.frameincr_works(), which caused failure if a non-Crossref DOI included (#52) pmid2doi()anddoi2pmid()functions removed temporarily as the web service is down temporarily, but will be online again soon from Crossref (#48)
- Fixes for man file examples. (#35)
cr_citation()is deprecated (stil useable, but will be removed in a future version of the package). usecr_cn()instead. (#34)
- released to CRAN