Can regex match underlining or italics?

[Christoph Ruehlemann]

Hi all,

a simple question: can underlined text or italicized text be matched in regex? For example, if you have this vector:

example <- c("hello", "that's it!", "what?", "what the hell")

which regex would find either element 2 (italicized) and/or element 4 (underlined)?

Best

Chris

--

Albert-Ludwigs-Universität Freiburg

Projekt-Leiter DFG-Projekt "Analyse multimodaler Interaktion im Geschichtenerzählen"

http://paul.igl.uni-freiburg.de/ruehlemann/?Home

ἰχθύς

[Earl Brown]

I don't think R preserves the bold, italics or underline on text once imported into an R session. However, here's a way to do what you want if the formatted text is in a .docx file (like the attached .docx file). First, manually change the .docx extension to .zip, and then unzip it (cf. thread here). When you unzip it, a new directory named "word" will be created, and within it there are several XML files, one of which is named "document.xml". The follow R code should return the paragraphs with bold, italicized or underlined text (after changing the pathway to "document.xml" on your hard drive):

library(xml2)

library(stringr)

doc <- read_html("/Users/ekb5/Downloads/word/document.xml")

paragraphs <- unlist(str_extract_all(as.character(doc), "<p.*?/p>"))

extract_formatted_text <- function(txt) {

  # Return paragraphs with bold, italicized, or underlined text

  if (str_detect(txt, "</[biu]>")) {

    output <- str_replace_all(txt, "<.*?>", "")

  } else {

    output <- NA

  }

  return(output)

}

# keep paragraphs with some bold, italicized, or underlined text

result <- unlist(lapply(paragraphs, function(x) extract_formatted_text(x)))

result <- result[!is.na(result)]

print(result)

Cheers!