gutenbergr: Search and download public domain texts from Project Gutenberg
David Robinson, Myfanwy Johnston
Source:vignettes/intro.Rmd
intro.RmdThe gutenbergr package helps you download and process public domain works from the Project Gutenberg collection. This includes both tools for downloading books (and stripping header/footer information), and a complete dataset of Project Gutenberg metadata that can be used to find words of interest. Includes:
- A function
gutenberg_download()that downloads one or more works from Project Gutenberg by ID: e.g.,gutenberg_download(84)downloads the text of Frankenstein. - Metadata for all Project Gutenberg works as R datasets, so that they
can be searched and filtered:
-
gutenberg_metadatacontains information about each work, pairing Gutenberg ID with title, author, language, etc -
gutenberg_authorscontains information about each author, such as aliases and birth/death year -
gutenberg_subjectscontains pairings of works with Library of Congress subjects and topics
-
Project Gutenberg Metadata
This package contains metadata for all Project Gutenberg works as R datasets, so that you can search and filter for particular works before downloading.
The dataset gutenberg_metadata contains information
about each work, pairing Gutenberg ID with title, author, language,
etc:
gutenberg_metadata
#> # A tibble: 77,649 × 8
#> gutenberg_id title author gutenberg_author_id language gutenberg_bookshelf
#> <int> <chr> <chr> <int> <chr> <chr>
#> 1 1 "The De… Jeffe… 1638 en Politics/American …
#> 2 2 "The Un… Unite… 1 en Politics/American …
#> 3 3 "John F… Kenne… 1666 en Browsing: History …
#> 4 4 "Lincol… Linco… 3 en US Civil War/Brows…
#> 5 5 "The Un… Unite… 1 en United States/Poli…
#> 6 6 "Give M… Henry… 4 en American Revolutio…
#> 7 7 "The Ma… NA NA en Browsing: History …
#> 8 8 "Abraha… Linco… 3 en US Civil War/Brows…
#> 9 9 "Abraha… Linco… 3 en US Civil War/Brows…
#> 10 10 "The Ki… NA NA en Banned Books List …
#> # ℹ 77,639 more rows
#> # ℹ 2 more variables: rights <chr>, has_text <lgl>For example, you could find the Gutenberg ID(s) of Jane Austen’s Persuasion by doing:
gutenberg_metadata |>
filter(title == "Persuasion")
#> # A tibble: 3 × 8
#> gutenberg_id title author gutenberg_author_id language gutenberg_bookshelf
#> <int> <chr> <chr> <int> <chr> <chr>
#> 1 105 Persuasi… Auste… 68 en Browsing: Culture/…
#> 2 22963 Persuasi… Auste… 68 en Browsing: Culture/…
#> 3 36777 Persuasi… Auste… 68 fr FR Littérature
#> # ℹ 2 more variables: rights <chr>, has_text <lgl>In many analyses, you may want to filter just for English works,
avoid duplicates, and include only books that have text that can be
downloaded. The gutenberg_works() function does this
pre-filtering:
gutenberg_works()
#> # A tibble: 60,573 × 8
#> gutenberg_id title author gutenberg_author_id language gutenberg_bookshelf
#> <int> <chr> <chr> <int> <chr> <chr>
#> 1 1 "The De… Jeffe… 1638 en Politics/American …
#> 2 2 "The Un… Unite… 1 en Politics/American …
#> 3 3 "John F… Kenne… 1666 en Browsing: History …
#> 4 4 "Lincol… Linco… 3 en US Civil War/Brows…
#> 5 5 "The Un… Unite… 1 en United States/Poli…
#> 6 6 "Give M… Henry… 4 en American Revolutio…
#> 7 7 "The Ma… NA NA en Browsing: History …
#> 8 8 "Abraha… Linco… 3 en US Civil War/Brows…
#> 9 9 "Abraha… Linco… 3 en US Civil War/Brows…
#> 10 10 "The Ki… NA NA en Banned Books List …
#> # ℹ 60,563 more rows
#> # ℹ 2 more variables: rights <chr>, has_text <lgl>It also allows you to perform filtering as an argument:
gutenberg_works(author == "Austen, Jane")
#> # A tibble: 13 × 8
#> gutenberg_id title author gutenberg_author_id language gutenberg_bookshelf
#> <int> <chr> <chr> <int> <chr> <chr>
#> 1 105 "Persua… Auste… 68 en "Browsing: Culture…
#> 2 121 "Northa… Auste… 68 en "Gothic Fiction/Br…
#> 3 141 "Mansfi… Auste… 68 en "Browsing: Culture…
#> 4 158 "Emma" Auste… 68 en "Browsing: Culture…
#> 5 161 "Sense … Auste… 68 en "Browsing: Culture…
#> 6 946 "Lady S… Auste… 68 en "Browsing: Culture…
#> 7 1212 "Love a… Auste… 68 en "Browsing: Culture…
#> 8 1342 "Pride … Auste… 68 en "Best Books Ever L…
#> 9 31100 "The Co… Auste… 68 en "Browsing: Culture…
#> 10 37431 "Pride … Auste… 68 en ""
#> 11 42078 "The Le… Auste… 68 en ""
#> 12 63569 "The Wa… Auste… 68 en "Browsing: Culture…
#> 13 74233 "Fragme… Auste… 68 en ""
#> # ℹ 2 more variables: rights <chr>, has_text <lgl>
# or with a regular expression
gutenberg_works(str_detect(author, "Austen"))
#> # A tibble: 23 × 8
#> gutenberg_id title author gutenberg_author_id language gutenberg_bookshelf
#> <int> <chr> <chr> <int> <chr> <chr>
#> 1 105 Persuas… Auste… 68 en Browsing: Culture/…
#> 2 121 Northan… Auste… 68 en Gothic Fiction/Bro…
#> 3 141 Mansfie… Auste… 68 en Browsing: Culture/…
#> 4 158 Emma Auste… 68 en Browsing: Culture/…
#> 5 161 Sense a… Auste… 68 en Browsing: Culture/…
#> 6 946 Lady Su… Auste… 68 en Browsing: Culture/…
#> 7 1212 Love an… Auste… 68 en Browsing: Culture/…
#> 8 1342 Pride a… Auste… 68 en Best Books Ever Li…
#> 9 17797 Memoir … Auste… 7603 en Browsing: Biograph…
#> 10 22536 Jane Au… Auste… 25392 en Browsing: Biograph…
#> # ℹ 13 more rows
#> # ℹ 2 more variables: rights <chr>, has_text <lgl>The meta-data currently in the package was last updated on 14 September 2024.
Downloading books by ID
The function gutenberg_download() downloads one or more
works from Project Gutenberg based on their ID. For example, we earlier
saw that one version of Persuasion has ID 105 (see the URL here), so
gutenberg_download(105) downloads this text.
persuasion <- gutenberg_download(105)
persuasion
#> # A tibble: 8,357 × 2
#> gutenberg_id text
#> <int> <chr>
#> 1 105 "Persuasion"
#> 2 105 ""
#> 3 105 ""
#> 4 105 "by Jane Austen"
#> 5 105 ""
#> 6 105 "(1818)"
#> 7 105 ""
#> 8 105 ""
#> 9 105 ""
#> 10 105 ""
#> # ℹ 8,347 more rowsNotice it is returned as a tbl_df (a type of data frame) including
two variables: gutenberg_id (useful if multiple books are
returned), and a character vector of the text, one row per line.
You can also provide gutenberg_download() a vector of
IDs to download multiple books. For example, to download Renascence,
and Other Poems (book 109) along with
Persuasion, do:
books <- gutenberg_download(c(109, 105), meta_fields = c("title", "author"))
books
#> # A tibble: 9,579 × 4
#> gutenberg_id text title author
#> <int> <chr> <chr> <chr>
#> 1 109 "Renascence and Other Poems" Renascence, and Other Poems Millay…
#> 2 109 "" Renascence, and Other Poems Millay…
#> 3 109 "" Renascence, and Other Poems Millay…
#> 4 109 "by" Renascence, and Other Poems Millay…
#> 5 109 "" Renascence, and Other Poems Millay…
#> 6 109 "Edna St. Vincent Millay" Renascence, and Other Poems Millay…
#> 7 109 "" Renascence, and Other Poems Millay…
#> 8 109 "" Renascence, and Other Poems Millay…
#> 9 109 "" Renascence, and Other Poems Millay…
#> 10 109 "" Renascence, and Other Poems Millay…
#> # ℹ 9,569 more rowsNotice that the meta_fields argument allows us to add
one or more additional fields from the gutenberg_metadata
to the downloaded text, such as title or author.
books |>
count(title)
#> # A tibble: 2 × 2
#> title n
#> <chr> <int>
#> 1 Persuasion 8357
#> 2 Renascence, and Other Poems 1222Other meta-datasets
You may want to select books based on information other than their
title or author, such as their genre or topic.
gutenberg_subjects contains pairings of works with Library
of Congress subjects and topics. “lcc” means Library of Congress
Classification, while “lcsh” means Library of Congress
subject headings:
gutenberg_subjects
#> # A tibble: 249,409 × 3
#> gutenberg_id subject_type subject
#> <int> <chr> <chr>
#> 1 1 lcsh United States -- History -- Revolution, 1775-1783 …
#> 2 1 lcsh United States. Declaration of Independence
#> 3 1 lcc E201
#> 4 1 lcc JK
#> 5 2 lcsh Civil rights -- United States -- Sources
#> 6 2 lcsh United States. Constitution. 1st-10th Amendments
#> 7 2 lcc JK
#> 8 2 lcc KF
#> 9 3 lcsh United States -- Foreign relations -- 1961-1963
#> 10 3 lcsh Presidents -- United States -- Inaugural addresses
#> # ℹ 249,399 more rowsThis is useful for extracting texts from a particular topic or genre,
such as detective stories, or a particular character, such as Sherlock
Holmes. The gutenberg_id column can then be used to
download these texts or to link with other metadata.
gutenberg_subjects |>
filter(subject == "Detective and mystery stories")
#> # A tibble: 894 × 3
#> gutenberg_id subject_type subject
#> <int> <chr> <chr>
#> 1 170 lcsh Detective and mystery stories
#> 2 173 lcsh Detective and mystery stories
#> 3 244 lcsh Detective and mystery stories
#> 4 305 lcsh Detective and mystery stories
#> 5 330 lcsh Detective and mystery stories
#> 6 481 lcsh Detective and mystery stories
#> 7 547 lcsh Detective and mystery stories
#> 8 863 lcsh Detective and mystery stories
#> 9 905 lcsh Detective and mystery stories
#> 10 1155 lcsh Detective and mystery stories
#> # ℹ 884 more rows
gutenberg_subjects |>
filter(grepl("Holmes, Sherlock", subject))
#> # A tibble: 55 × 3
#> gutenberg_id subject_type subject
#> <int> <chr> <chr>
#> 1 108 lcsh Holmes, Sherlock (Fictitious character) -- Fiction
#> 2 221 lcsh Holmes, Sherlock (Fictitious character) -- Fiction
#> 3 244 lcsh Holmes, Sherlock (Fictitious character) -- Fiction
#> 4 834 lcsh Holmes, Sherlock (Fictitious character) -- Fiction
#> 5 1661 lcsh Holmes, Sherlock (Fictitious character) -- Fiction
#> 6 2097 lcsh Holmes, Sherlock (Fictitious character) -- Fiction
#> 7 2343 lcsh Holmes, Sherlock (Fictitious character) -- Fiction
#> 8 2344 lcsh Holmes, Sherlock (Fictitious character) -- Fiction
#> 9 2345 lcsh Holmes, Sherlock (Fictitious character) -- Fiction
#> 10 2346 lcsh Holmes, Sherlock (Fictitious character) -- Fiction
#> # ℹ 45 more rowsgutenberg_authors contains information about each
author, such as aliases and birth/death year:
gutenberg_authors
#> # A tibble: 25,514 × 7
#> gutenberg_author_id author alias birthdate deathdate wikipedia aliases
#> <int> <chr> <chr> <int> <int> <chr> <chr>
#> 1 1 United States U.S.… NA NA https://… U.S.A.
#> 2 3 Lincoln, Abr… NA 1809 1865 https://… United…
#> 3 4 Henry, Patri… NA 1736 1799 https://… NA
#> 4 5 Adam, Paul NA 1849 1931 https://… NA
#> 5 7 Carroll, Lew… Dodg… 1832 1898 https://… Dodgso…
#> 6 8 United State… NA NA NA https://… Agency…
#> 7 9 Melville, He… Melv… 1819 1891 https://… Melvil…
#> 8 10 Barrie, J. M… NA 1860 1937 https://… Barrie…
#> 9 11 Church of Je… NA NA NA https://… NA
#> 10 12 Smith, Josep… Smit… 1805 1844 https://… Smith,…
#> # ℹ 25,504 more rowsAnalysis
What’s next after retrieving a book’s text? Well, having the book as a data frame is especially useful for working with the tidytext package for text analysis.
words <- books |>
unnest_tokens(word, text)
words
#> # A tibble: 90,581 × 4
#> gutenberg_id title author word
#> <int> <chr> <chr> <chr>
#> 1 109 Renascence, and Other Poems Millay, Edna St. Vincent renascence
#> 2 109 Renascence, and Other Poems Millay, Edna St. Vincent and
#> 3 109 Renascence, and Other Poems Millay, Edna St. Vincent other
#> 4 109 Renascence, and Other Poems Millay, Edna St. Vincent poems
#> 5 109 Renascence, and Other Poems Millay, Edna St. Vincent by
#> 6 109 Renascence, and Other Poems Millay, Edna St. Vincent edna
#> 7 109 Renascence, and Other Poems Millay, Edna St. Vincent st
#> 8 109 Renascence, and Other Poems Millay, Edna St. Vincent vincent
#> 9 109 Renascence, and Other Poems Millay, Edna St. Vincent millay
#> 10 109 Renascence, and Other Poems Millay, Edna St. Vincent contents
#> # ℹ 90,571 more rows
word_counts <- words |>
anti_join(stop_words, by = "word") |>
count(title, word, sort = TRUE)
word_counts
#> # A tibble: 6,664 × 3
#> title word n
#> <chr> <chr> <int>
#> 1 Persuasion anne 447
#> 2 Persuasion captain 302
#> 3 Persuasion elliot 254
#> 4 Persuasion lady 214
#> 5 Persuasion wentworth 191
#> 6 Persuasion charles 155
#> 7 Persuasion time 152
#> 8 Persuasion sir 149
#> 9 Persuasion miss 125
#> 10 Persuasion walter 123
#> # ℹ 6,654 more rowsYou may also find these resources useful:
- The Natural Language Processing CRAN View suggests many R packages related to text mining, especially around the tm package
- You could match the
wikipediacolumn ingutenberg_authorto Wikipedia content with the WikipediR package or to pageview statistics with the wikipediatrend package - If you’re considering an analysis based on author name, you may find
the humaniformat
(for extraction of first names) and gender (prediction
of gender from first names) packages useful. (Note that humaniformat has
a
format_reversefunction for reversing “Last, First” names).