Week 7: HTML and JSON

Introduction

I previously studied history and have many books in my collection from that time (and my own personal interest before and after). I selected three books about World War I to be the topic of my assignment. I documented each book’s title, authors, publication year, publisher, and ISBN in a hand written .json and .html file. Then, I used R to read the contents into a dataframe and compared the results.

Data Import

Using the rvest package, I’m able to read in the html file as tables. The first table contains my book data and the second contains table attributes. For our purposes, we just need the first table, so I extract that into a data frame.

html_url <- "https://raw.githubusercontent.com/jocslater-code/DATA607/refs/heads/main/Week7/book.html"
# Read in as table
html_tbl <- read_html(html_url) %>% html_table()

# Extract first table into dataframe
html_df <- html_tbl[[1]]

print(html_df)

# A tibble: 3 × 5
  title                                 authors publication_year publisher isbn 
  <chr>                                 <chr>              <int> <chr>     <chr>
1 A Peace to End All Peace              David …             1989 Henry Ho… 978-…
2 The Sleepwalkers: How Europe Went to… Christ…             2012 HarperCo… 978-…
3 14-18: Understanding the Great War    Stéph…             2002 Hill and… 978-…

Using the jsonlite package, I read in the json file directly into a dataframe.

json_url <- "https://raw.githubusercontent.com/jocslater-code/DATA607/refs/heads/main/Week7/books.json"
# Load json into data frame
json_df <- fromJSON(json_url)

# View the dataframe
print(json_df)

                                             title
1                         A Peace to End All Peace
2 The Sleepwalkers: How Europe Went to War in 1914
3               14-18: Understanding the Great War
                                  authors publication_year          publisher
1                           David Fromkin             1989 Henry Holt and Co.
2                       Christopher Clark             2012      HarperCollins
3 Stéphane Audoin-Rouzeau, Annette Becker             2002      Hill and Wang
            isbn
1 978-0805068849
2 978-0061146657
3 978-0809046430

Analysis

Visually comparing the data frames, we can see that the contents is essentially identical, with a few minor differences. There were some special character issues with the json file. Instead of reading Stéphane Audoin-Rouzeau’s name with the properly accented é, their name is rendered “Stéphane”. The json format also right justified the table while the html format was left justified.

Each format handles the multiple authors differently. The json format casts authors as a list, so that the books with multiple authors just has a list of two elements. The html format concatonates the authors together, separated by a comma.

Next Steps

The special character issues are due to R using my system default encoding, rather than UTF-8. This can cause an misinterpertation of multi-byte characters like é. If I wanted to resolve this, I would be explicit about the system encoding.