pathr

The goal of {pathr} is to simplify the process of string manipulations for URLs and file paths. This uses modified functions from the {httr} to create opinionated functions that allow working with URLs and paths easier.

Installation

You can install the development version of {pathr} like so:

devtools::install_github("koderkow/pathr")

URLs

Basic URLs

Sort components of a URL

url <- "https://github.com/KoderKow/pathr"

x <- parse_url(url)

x
#> - url: https://github.com/KoderKow/pathr
#> - scheme: https
#> - hostname: github.com
#> - host_info_name: github
#> - host_info_extension: com
#> - path: KoderKow/pathr

str(x)
#> List of 15
#>  $ url                : chr "https://github.com/KoderKow/pathr"
#>  $ scheme             : chr "https"
#>  $ hostname           : chr "github.com"
#>  $ host_info_name     : chr "github"
#>  $ host_info_extension: chr "com"
#>  $ port               : NULL
#>  $ path               : chr "KoderKow/pathr"
#>  $ file_dir           : NULL
#>  $ file_name          : NULL
#>  $ file_ext           : NULL
#>  $ query              : NULL
#>  $ params             : NULL
#>  $ fragment           : NULL
#>  $ username           : NULL
#>  $ password           : NULL
#>  - attr(*, "class")= chr [1:2] "list" "pathr"

Put it back to together

build_url(x)
#> [1] "https://github.com/KoderKow/pathr"

URLs with a file path

URLs with item paths will be detected

url <- "https://github.com/KoderKow/pathr/example-folder/my-data.csv"

x <- parse_url(url)

x
#> - url: https://github.com/KoderKow/pathr/example-folder/my-data.csv
#> - scheme: https
#> - hostname: github.com
#> - host_info_name: github
#> - host_info_extension: com
#> - path: KoderKow/pathr/example-folder/my-data.csv
#> - file_dir: KoderKow/pathr/example-folder
#> - file_name: my-data
#> - file_ext: csv

str(x)
#> List of 15
#>  $ url                : chr "https://github.com/KoderKow/pathr/example-folder/my-data.csv"
#>  $ scheme             : chr "https"
#>  $ hostname           : chr "github.com"
#>  $ host_info_name     : chr "github"
#>  $ host_info_extension: chr "com"
#>  $ port               : NULL
#>  $ path               : chr "KoderKow/pathr/example-folder/my-data.csv"
#>  $ file_dir           : chr "KoderKow/pathr/example-folder"
#>  $ file_name          : chr "my-data"
#>  $ file_ext           : chr "csv"
#>  $ query              : NULL
#>  $ params             : NULL
#>  $ fragment           : NULL
#>  $ username           : NULL
#>  $ password           : NULL
#>  - attr(*, "class")= chr [1:2] "list" "pathr"

Building the URL ignores the file element

build_url(x)
#> [1] "https://github.com/KoderKow/pathr/example-folder/my-data.csv"

Vector of URLS 

urls <- c(
  "https://github.com/KoderKow/pathr1",
  "https://github.com/KoderKow/pathr2"
)

x <- parse_url(urls)

x
#> [[1]]
#> - url: https://github.com/KoderKow/pathr1
#> - scheme: https
#> - hostname: github.com
#> - host_info_name: github
#> - host_info_extension: com
#> - path: KoderKow/pathr1
#> 
#> [[2]]
#> - url: https://github.com/KoderKow/pathr2
#> - scheme: https
#> - hostname: github.com
#> - host_info_name: github
#> - host_info_extension: com
#> - path: KoderKow/pathr2

str(x)
#> List of 2
#>  $ :List of 15
#>   ..$ url                : chr "https://github.com/KoderKow/pathr1"
#>   ..$ scheme             : chr "https"
#>   ..$ hostname           : chr "github.com"
#>   ..$ host_info_name     : chr "github"
#>   ..$ host_info_extension: chr "com"
#>   ..$ port               : NULL
#>   ..$ path               : chr "KoderKow/pathr1"
#>   ..$ file_dir           : NULL
#>   ..$ file_name          : NULL
#>   ..$ file_ext           : NULL
#>   ..$ query              : NULL
#>   ..$ params             : NULL
#>   ..$ fragment           : NULL
#>   ..$ username           : NULL
#>   ..$ password           : NULL
#>   ..- attr(*, "class")= chr [1:2] "list" "pathr"
#>  $ :List of 15
#>   ..$ url                : chr "https://github.com/KoderKow/pathr2"
#>   ..$ scheme             : chr "https"
#>   ..$ hostname           : chr "github.com"
#>   ..$ host_info_name     : chr "github"
#>   ..$ host_info_extension: chr "com"
#>   ..$ port               : NULL
#>   ..$ path               : chr "KoderKow/pathr2"
#>   ..$ file_dir           : NULL
#>   ..$ file_name          : NULL
#>   ..$ file_ext           : NULL
#>   ..$ query              : NULL
#>   ..$ params             : NULL
#>   ..$ fragment           : NULL
#>   ..$ username           : NULL
#>   ..$ password           : NULL
#>   ..- attr(*, "class")= chr [1:2] "list" "pathr"

Working with a pathr object

Below are three common methods for turning a lists of lists into a data frame. The base R method is built into the {pathr} package due to it needing no dependencies. The other methods, data.table and dplyr, go over the process for those work flows.

Base R 

pathr_to_df(x)
#>                                  url scheme   hostname host_info_name
#> 1 https://github.com/KoderKow/pathr1  https github.com         github
#> 2 https://github.com/KoderKow/pathr2  https github.com         github
#>   host_info_extension port            path file_dir file_name file_ext query
#> 1                 com   NA KoderKow/pathr1       NA        NA       NA    NA
#> 2                 com   NA KoderKow/pathr2       NA        NA       NA    NA
#>   params fragment username password
#> 1     NA       NA       NA       NA
#> 2     NA       NA       NA       NA
data.table 
d <- data.table::rbindlist(x)

#> Warning message:
#> In data.table::rbindlist(x) :
#>   Column 6 ['port'] of item 1 is length 0. This (and 17 others like it) has been filled with NA (NULL for list columns) to make each item uniform.
dplyr

The pathr class does not play nicely with dplyr::bind_rows.

d <- 
  x |> 
  dplyr::bind_rows()
  
#> Error in `dplyr::bind_rows()`:
#> ! Argument 1 must be a data frame or a named atomic vector.
#> Run `rlang::last_error()` to see where the error occurred.

The function remove_pathr() will unclass any pathr object or a list containing nothing but pathr objects. This makes it easy to move the object into Rs default class, list.

d <-
  x |>
  remove_pathr() |>
  dplyr::bind_rows()

Thanks to