marc21r

{marc21r} is an R package for examining and analyzing MARC21 files, a set of standards used to describe library materials such as books and journals. Currently, {marc21r} reads only MARC21 XML.

Installation

You can install the development version of marc21r from GitHub with:

# install.packages("devtools")
devtools::install_github("dojobo/marc21r")

Example

The basic use case is to load a MARC XML file as a tibble. Each tag (001, 245, etc.) becomes a column whose name starts with t, e.g., t001, t245. Once so loaded, the dataset can be manipulated and analyzed using standard tidyverse tools.

library(marc21r)
library(dplyr)
#> 
#> Attaching package: 'dplyr'
#> The following objects are masked from 'package:stats':
#> 
#>     filter, lag
#> The following objects are masked from 'package:base':
#> 
#>     intersect, setdiff, setequal, union
collection <- read_marcxml("collection.xml")
collection %>% 
  select(t245, t260, t500)
#> # A tibble: 2 × 3
#>   t245                                     t260                            t500 
#>   <chr>                                    <chr>                           <chr>
#> 1 The Great Ray Charles [sound recording]. New York, N.Y. : Atlantic, [19… Brie…
#> 2 The White House [computer file].         Washington, D.C. : White House… Titl…

For better readability, tag_names() can be applied to a MARC dataframe:

collection %>% 
  select(t245, t260, t500) %>% 
  tag_names(style = "short")
#> # A tibble: 2 × 3
#>   t245_title_stmt                          t260_pub_dist                 t500_…¹
#>   <chr>                                    <chr>                         <chr>  
#> 1 The Great Ray Charles [sound recording]. New York, N.Y. : Atlantic, [… Brief …
#> 2 The White House [computer file].         Washington, D.C. : White Hou… Title …
#> # … with abbreviated variable name ¹t500_general_note

Current status

This project is alpha/experimental. It “works” as a proof of concept with one or two example files, but I expect there are gaps. Post problems you encounter or feature requests on the Issues page. I am also happy to accept pull requests!