Grep html code between html tags containing a keyword in R

Question

Within a file, I would like to use grep or maybe use the package qdapRegex's rm_between function to extract a whole section of html code containing a keyword, lets say "discount rate" for this example. Specifically, I want results that look like this code snippet:

<P>This is a paragraph containing the words discount rate including other things.</P>

and

<TABLE width="400">
  <tr>
    <th>Month</th>
    <th>Savings</th>
  </tr>
  <tr>
    <td>Discount Rate</td>
    <td>10.0%</td>
  </tr>
  <tr>
    <td>February</td>
    <td>$80</td>
  </tr>
</TABLE>

1. The trick here is it must find discount rate first and then pull out the rest.
2. It is always going to be between <P> and </P> or <TABLE and </TABLE> and no other html tags.

A good sample .txt file for this can be found here:

https://www.sec.gov/Archives/edgar/data/66740/0000897101-04-000425.txt

Answer 1

You can consider the file as html and explore it as if you were scraping it with rvest:

library(rvest)
library(stringr)

# Extract the html from the file
html = read_html('~/Downloads/0000897101-04-000425.txt')

# Get all the 'p' nodes (you can do the same for 'table')
p_nodes <- html %>% html_nodes('p')

# Get the text from each node
p_nodes_text <- p_nodes %>% html_text()

# Find the nodes that have the term you are looking for
match_indeces <- str_detect(p_nodes_text, fixed('discount rate', ignore_case = TRUE))

# Keep only the nodes with matches
# Notice that I remove the first match because rvest adds a 
# 'p' node to the whole file, since it is a text file
match_p_nodes <- p_nodes[match_indeces][-1]

# If you want to see the results, you can print them like this
# (or you could send them to a file)
for(i in 1:length(match_p_nodes)) {
  cat(paste0('Node #', i, ': ', as.character(match_p_nodes[i]), '\n\n'))
}

For the <table> tags, you would not remove the first match:

table_nodes <- html %>% html_nodes('table')
table_nodes_text <- table_nodes %>% html_text()
match_indeces_table <- str_detect(table_nodes_text, fixed('discount rate', ignore_case = TRUE))
match_table_nodes <- table_nodes[match_indeces_table]

for(i in 1:length(match_table_nodes)) {
  cat(paste0('Node #', i, ': ', as.character(match_table_nodes[i]), '\n\n'))
}