Webscraping in R where input is required - r

Question

I have used the rvest package in R to scrape unique URLs before.
However, I am now stuck with a particular website. The URL stays static and I need to select the following dropdowns now and scrape the resulting table that appears.
Will be helpful if someone can guide me on what direction to take with websites like these? Is R even capable of doing this?
Edit: I have researched and it seems RSelenium can handle such tasks. Unfortunately, I have no exposure to it. Can someone recommend an example/blog/material online on using Selenium specifically for clicking and scraping for someone as noob as I am?

Answer 1

I have made a blog post about an RSelenium example:
https://guillaumepressiat.github.io/blog/2021/04/RSelenium-paginated-tables
this website contains a lot of things about selenium, you will have to plug it to RSelenium api package.(verbs are almost the same in all languages, findElement, etc) https://www.guru99.com/selenium-tutorial.html
But as an example based on your question maybe something like this to begin:
# https://stackoverflow.com/q/67021563/10527496
# java -jar selenium-server-standalone-3.9.1.jar
library(RSelenium)
library(tidyverse)
library(rvest)
library(httr)
remDr <- remoteDriver(
remoteServerAddr = "localhost",
port = 4444L, # change port according to terminal
browserName = "firefox"
)
remDr$open()
# remDr$getStatus()
url <- "https://fcainfoweb.nic.in/reports/Report_Menu_Web.aspx"
remDr$navigate(url)
Sys.sleep(5)
# first : radio buttons
u1 <- remDr$findElement(using = "id", value = 'ctl00_MainContent_Rbl_Rpt_type_0')
u2 <- remDr$findElement(using = "id", value = 'ctl00_MainContent_Rbl_Rpt_type_1')
u3 <- remDr$findElement(using = "id", value = 'ctl00_MainContent_Rbl_Rpt_type_2')
u4 <- remDr$findElement(using = "id", value = 'ctl00_MainContent_Rbl_Rpt_type_3')
dynam <- remDr$mouseMoveToLocation(webElement = u1)
u1$click()
Sys.sleep(5)
# second : Select input
s1 <- remDr$findElement(using = "id", value = 'ctl00_MainContent_Ddl_Rpt_Option0')
# get available choices
s_choices <- read_html(s1$getElementAttribute('innerHTML')[[1]]) %>%
html_nodes('option') %>%
html_attrs() %>%
unlist() %>%
.[3:length(.)] %>%
as.vector()
dynam <- remDr$mouseMoveToLocation(webElement = s1)
s1$click()
s1$sendKeysToElement(sendKeys = list(s_choices[1], key = "enter"))
# s_choices[1] is "Daily Prices"
Sys.sleep(5)
# get date choices
s_date_choices <- remDr$findElement(using = "id", value = "ctl00_MainContent_Txt_FrmDate")
dynam <- remDr$mouseMoveToLocation(webElement = s_date_choices)
s_date_choices$click()
s_date_choices$sendKeysToElement(sendKeys = list('01/01/2021', key = "enter"))
Sys.sleep(5)
s_table <- remDr$findElement(using = "id", value = "Panel1")
# get first tables as an example
results_1 <- read_html(s_table$getElementAttribute('innerHTML')[[1]]) %>%
html_table(fill = TRUE) %>%
.[2:length(.)]
we get a list of three tables as a result:
Making a function from this code to loop on a date vector is possible after that I think (you will have to reload a fresh start page on base URL for each date I suppose).