I am in the process of writing a collection of freely-downloadable R scripts for http://asdfree.com/ to help people analyze the complex sample survey data hosted by the UK data service. In addition to providing lots of statistics tutorials for these data sets, I also want to automate the download and importation of this survey data. In order to do that, I need to figure out how to programmatically log into this UK data service website.
I have tried lots of different configurations of RCurl and httr to log in, but I'm making a mistake somewhere and I'm stuck. I have tried inspecting the elements as outlined in this post, but the websites jump around too fast in the browser for me to understand what's going on.
This website does require a login and password, but I believe I'm making a mistake before I even get to the login page.
The starting page should be: https://www.esds.ac.uk/secure/UKDSRegister_start.asp
This page will automatically re-direct your web browser to a long URL that starts with: https://wayf.ukfederation.org.uk/DS002/uk.ds?[blahblahblah]
(1) For some reason, the SSL certificate does not work on this website. Here's the SO question I posted regarding this. The workaround I've used is simply ignoring the SSL:
library(httr)
set_config( config( ssl.verifypeer = 0L ) )
and then my first command on the starting website is:
z <- GET( "https://www.esds.ac.uk/secure/UKDSRegister_start.asp" )
this gives me back a z$url
that looks a lot like the https://wayf.ukfederation.org.uk/DS002/uk.ds?[blahblahblah]
page that my browser also re-directs to.
In the browser, then, you're supposed to type in "uk data archive" and click the continue
button. When I do that, it re-directs me to the web page https://shib.data-archive.ac.uk/idp/Authn/UserPassword
I think this is where I'm stuck because I cannot figure out how to have cURL followlocation
and land on this website. Note: no username/password has been entered yet.
When I use the httr GET
command from the wayf.ukfederation.org.uk page like this:
y <- GET( z$url , query = list( combobox = "https://shib.data-archive.ac.uk/shibboleth-idp" ) )
the y$url
string looks a lot like z$url
(except it's got a combobox= on the end). Is there any way to get through to this uk data archive
authentication page with RCurl or httr?
I can't tell if I'm just overlooking something or if I absolutely must use the SSL certificate described in my previous SO post or what?
(2) At the point I do make it through to that page, I believe the remainder of the code would just be:
values <- list( j_username = "your.username" ,
j_password = "your.password" )
POST( "https://shib.data-archive.ac.uk/idp/Authn/UserPassword" , body = values)
But I guess that page will have to wait...
The relevant data variables returned by the form are action
and origin
, not combobox
. Give action
the value selection
and origin
the value from the relevant entry in combobox
y <- GET( z$url, query = list( action="selection", origin = "https://shib.data-archive.ac.uk/shibboleth-idp") )
> y$url
[1] "https://shib.data-archive.ac.uk:443/idp/Authn/UserPassword"
Edit
It looks as though the handle pool isn't keeping your session alive correctly. You therefore need to pass the handles directly rather than automatically. Also for the POST
command you need to set multipart=FALSE
as this is the default for HTML forms. The R command has a different default as it is mainly designed for uploading files. So:
y <- GET( handle=z$handle, query = list( action="selection", origin = "https://shib.data-archive.ac.uk/shibboleth-idp") )
POST(body=values,multipart=FALSE,handle=y$handle)
Response [https://www.esds.ac.uk/]
Status: 200
Content-type: text/html
...snipped...
<title>
Introduction to ESDS
</title>
<meta name="description" content="Introduction to the ESDS, home page" />
I think one way to address "enter your organization" page goes like this:
library(tidyverse)
library(rvest)
library(stringr)
org <- "your_organization"
user <- "your_username"
password <- "your_password"
signin <- "http://esds.ac.uk/newRegistration/newLogin.asp"
handle_reset(signin)
# get to org page and enter org
p0 <- html_session(signin) %>%
follow_link("Login")
org_link <- html_nodes(p0, "option") %>%
str_subset(org) %>%
str_match('(?<=\\")[^"]*') %>%
as.character()
f0 <- html_form(p0) %>%
first() %>%
set_values(origin = org_link)
fake_submit_button <- list(name = "submit-btn",
type = "submit",
value = "Continue",
checked = NULL,
disabled = NULL,
readonly = NULL,
required = FALSE)
attr(fake_submit_button, "class") <- "btn-enabled"
f0[["fields"]][["submit"]] <- fake_submit_button
c0 <- cookies(p0)$value
names(c0) <- cookies(p0)$name
p1 <- submit_form(session = p0, form = f0, config = set_cookies(.cookies = c0))
Unfortunately, that doesn't solve the whole problem—(2) is harder than it looks. I've got more of what I think is a solution posted here: R: use rvest (or httr) to log in to a site requiring cookies. Hopefully someone will help us get the rest of the way.
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With