friendly wrappers for simple queries

[MLopez-Ibanez]

A query like:

cr_cn(dois = "10.1007/s10479-023-05251-3", format="citeproc-json")

returns lots of data. It would be nice to have friendly wrappers for simple queries that only return the data queried as an R list:

cr_cn(dois = "10.1007/s10479-023-05251-3", format="list", fields=c("title", "author", "year"))

or to check that a doi actually exists:

doi_exists(dois = "10.1007/s10479-023-05251-3")

or to check that it matches some given data:

doi_match(dois = "10.1007/s10479-023-05251-3", title = "A title", year = "1997", exact=FALSE)

or to check that a paper is retracted or not (https://www.crossref.org/documentation/principles-practices/best-practices/versioning/#00327):

is_retracted(dois = "10.1007/s10479-023-05251-3")

Ideally, this should not require having "bibtex" installed:

cr_cn(dois = "10.1007/s10479-023-05251-3", format="bibentry")

returns:

Error: Please install bibtex

[njahn82]

Hi @MLopez-Ibanez

You can use cr_works to get parsed metadata by DOI and select the fields.

library(rcrossref)
rcrossref::cr_works(filter = list(doi="10.1007/s10479-023-05251-3"), 
                    select = c("title", "issued", "author"))
#> $meta
#>   total_results search_terms start_index items_per_page
#> 1             1           NA           0             20
#> 
#> $data
#> # A tibble: 1 × 3
#>   issued     title                                                      author  
#>   <chr>      <chr>                                                      <list>  
#> 1 2023-03-01 RETRACTED ARTICLE: Prescriptive analytics applications in… <tibble>
#> 
#> $facets
#> NULL

^{Created on 2024-06-12 with reprex v2.1.0}

You can also select columns after getting the data.

library(rcrossref)
req <- rcrossref::cr_works(dois = "10.1007/s10479-023-05251-3") 
req[["data"]][,c("title", "issued", "author")]
#> # A tibble: 1 × 3
#>   title                                                          issued author  
#>   <chr>                                                          <chr>  <list>  
#> 1 RETRACTED ARTICLE: Prescriptive analytics applications in sus… 2023-… <tibble>

^{Created on 2024-06-12 with reprex v2.1.0}

To check if a DOI exists, check the agency.

library(rcrossref)
rcrossref::cr_agency("10.1007/s10479-023-05251-3")
#> $DOI
#> [1] "10.1007/s10479-023-05251-3"
#> 
#> $agency
#> $agency$id
#> [1] "crossref"
#> 
#> $agency$label
#> [1] "Crossref"

^{Created on 2024-06-12 with reprex v2.1.0}

I will look into how to discover retracted articles with Crossref.

I don't think I will support the other suggested data validation functions because they are not supported by the Crossref API.

[MLopez-Ibanez]

You can use cr_works to get parsed metadata by DOI and select the fields.

cr_works does not seem to return the same output as cr_cn

cr_cn(dois="10.4230/DagRep.5.1.96")
#> [1] "@article{https://doi.org/10.4230/dagrep.5.1.96,\n  doi = {10.4230/DAGREP.5.1.96},\n  url = {https://drops.dagstuhl.de/entities/document/10.4230/DagRep.5.1.96},\n  author = {Greco, Salvatore and Klamroth, Kathrin and Knowles, Joshua D. and Rudolph, Günter},\n  keywords = {multiple criteria decision making, evolutionary multiobjective optimization},\n  language = {en},\n  title = {Understanding Complexity in Multiobjective Optimization (Dagstuhl Seminar 15031)},\n  publisher = {Schloss Dagstuhl – Leibniz-Zentrum für Informatik},\n  year = {2015},\n  copyright = {Creative Commons Attribution 3.0 Germany license}\n}\n"
cr_works(filter = list(doi="10.4230/DagRep.5.1.96"))
#> $meta
#>   total_results search_terms start_index items_per_page
#> 1             0           NA           0             20
#> $data
#> # A tibble: 0 × 0
#> $facets
#> NULL

[MLopez-Ibanez]

Basically, I have a large database of bibtex entries: https://iridia-ulb.github.io/references/
I would like to do the following:

• Check that the DOIs added are valid and point to something.
• Check that the DOIs point to the correct paper (by for example matching the title)
• Check that a paper in the database has not been retracted.
• Do the above with the smallest overload on the crossref service possible.