Skip to contents

Resolve GitHub repo URLs for a working sample, and filter down to the packages for which one was found.

Source: R/sample.R
resolve_repo_urls.Rd

Applies repo_urls_fn to the name column of working_sample (e.g. pypi_repo_urls_many() or npm_repo_urls_many()), then drops rows for which no GitHub repo URL could be resolved. Shared post-processing step for both the PyPI and npm working samples.

Usage

resolve_repo_urls(working_sample = NULL, repo_urls_fn = NULL)

Arguments

working_sample

A tibble with at least name and downloads columns, as returned by build_working_sample().

repo_urls_fn

A function taking a character vector of package names and returning a character vector of the same length, with a resolved GitHub repo URL or NA for each.

Value

A tibble with name, downloads, and repo_url columns, filtered to rows with a non-missing repo_url.

Examples

working_sample <- tibble::tibble (
    name = c ("pkg1", "pkg2", "pkg3"),
    downloads = c (300, 200, 100)
)
fake_repo_urls_fn <- function (names_vec) {
    ifelse (
        names_vec == "pkg2",
        NA_character_,
        paste0 ("https://github.com/org/", names_vec)
    )
}
resolve_repo_urls (working_sample, fake_repo_urls_fn)
#> # A tibble: 2 × 3
#>   name  downloads repo_url                   
#>   <chr>     <dbl> <chr>                      
#> 1 pkg1        300 https://github.com/org/pkg1
#> 2 pkg3        100 https://github.com/org/pkg3