How to webscrape secured pages in R (https links) (using readHTMLTable from XML package)?

Question

There are good answers on SO about how to use readHTMLTable from the XML package and I did that with regular http pages, however I am not able to solve my problem with https

Answer 1

The new package httr provides a wrapper around RCurl to make it easier to scrape all kinds of pages.

Still, this page gave me a fair amount of trouble. The following works, but no doubt there are easier ways of doing it.

library("httr")
library("XML")

# Define certicificate file
cafile <- system.file("CurlSSL", "cacert.pem", package = "RCurl")

# Read page
page <- GET(
  "https://ned.nih.gov/", 
  path="search/ViewDetails.aspx", 
  query="NIHID=0010121048",
  config(cainfo = cafile)
)

# Use regex to extract the desired table
x <- text_content(page)
tab <- sub('.*(<table class="grid".*?>.*</table>).*', '\\1', x)

# Parse the table
readHTMLTable(tab)

The results:

$ctl00_ContentPlaceHolder_dvPerson
                V1                                      V2
1      Legal Name:                    Dr Francis S Collins
2  Preferred Name:                      Dr Francis Collins
3          E-mail:                 francis.collins@nih.gov
4        Location: BG 1 RM 1261 CENTER DRBETHESDA MD 20814
5       Mail Stop:                                       Â
6           Phone:                            301-496-2433
7             Fax:                                       Â
8              IC:             OD (Office of the Director)
9    Organization:            Office of the Director (HNA)
10 Classification:                                Employee
11            TTY:                                       Â

---

Get httr here: http://cran.r-project.org/web/packages/httr/index.html

---

EDIT: Useful page with FAQ about the RCurl package: http://www.omegahat.org/RCurl/FAQ.html

Answer 2

This is the function I have to deal with this problem. Detects if https in url and uses httr if it is.

readHTMLTable2=function(url, which=NULL, ...){
 require(httr)
 require(XML)
 if(str_detect(url,"https")){
    page <- GET(url, user_agent("httr-soccer-ranking"))
    doc = htmlParse(text_content(page))
    if(is.null(which)){
      tmp=readHTMLTable(doc, ...)
      }else{
        tableNodes = getNodeSet(doc, "//table")
        tab=tableNodes[[which]]
        tmp=readHTMLTable(tab, ...) 
      }
  }else{
    tmp=readHTMLTable(url, which=which, ...) 
  }
  return(tmp)
}

Answer 3

Using Andrie's great way to get past the https

a way to get at the data without readHTMLTable is also below.

A table in HTML may have an ID. In this case the table has one nice one and the XPath in getNodeSet function does it nicely.

# Define certicificate file
cafile <- system.file("CurlSSL", "cacert.pem", package = "RCurl")
# Read page
page <- GET(
  "https://ned.nih.gov/", 
  path="search/ViewDetails.aspx", 
  query="NIHID=0010121048",
  config(cainfo = cafile, ssl.verifypeer = FALSE)
)

h = htmlParse(page)
ns <- getNodeSet(h, "//table[@id = 'ctl00_ContentPlaceHolder_dvPerson']")
ns

I still need to extract the IDs behind the hyperlinks.

for example instead of collen baros as manager, I need to get to the ID 0010080638

Manager:Colleen Barros