Search code examples
rpdftext-mining

Read multiple pdf files at once and extract sentences that contain a keyword using R


Let's assume that I have few pdf files stored in a directory and I want to read all those pdf files at one and extract all the sentences that contain a specific keyword (in this case 'provisions') instead of manually opening each file and looking for that keyword.

I have tried reading the files but how can I make R go through each pdf file to search for that keyword and output those sentences? Here's a small piece that I have written:

library(pdftools)
files <- list.files("filepath",pattern = "pdf$", full.names = TRUE)
comb <- lapply(files, pdf_text)

For file reference purpose, the links for pdf files are:

<https://www.supremecourt.gov/opinions/14pdf/13-1314_3ea4.pdf> 
<https://www.supremecourt.gov/opinions/14pdf/14-7955_aplc.pdf>
<https://www.supremecourt.gov/opinions/14pdf/14-46_bqmc.pdf>

I have created a directory and saved the pdf files in it.


Solution

  • An update to the question: I have found the solution which can be achieved with the below code:

    install.packages("textreadr")
    install.packages("tidyverse")
    install.packages("pdfsearch")
    library(textreadr)
    library(tidyverse)
    library(pdfsearch)
    
    dirct <- directory_path
    result <- keyword_directory(dirct, 
                                keyword = 'input_the_keyword_you_want_to_extract',
                                surround_lines = 0, full_names = TRUE)
    head(result$line_text, n = 20)