Rselenium - How to scrape data from a webpage with no id or names of any sort

Question

I'm currently trying to scrape Biodiversity data from a specific website (http://www.faunaeur.org/?no_redirect=1). I have managed to get some results but not as automated as I hoped...The First Part is done, which is browsing through the website:

Setting up Rselenium:

library(RSelenium)
download.file("https://github.com/mozilla/geckodriver/releases/download/v0.11.1/geckodriver-v0.11.1-win64.zip",destfile="./gecko.zip")
unzip("./gecko.zip",exdir=".",overwrite=T)
checkForServer(update=T)
selfserv = startServer()
mybrowser1 = remoteDriver(browserName="firefox",extraCapabilities = list(marionette = TRUE))
mybrowser1$open()

Then starting my browsing (that would be an example for the Balearic Islands):

mybrowser1$navigate("http://www.faunaeur.org/distribution.php?current_form=species_list")
mybrowser1$findElement(using="xpath","//select[@name='taxon_rank']/option[@value='7']")$clickElement()    # Class
mybrowser1$findElement(using="xpath","//input[@name='taxon_name']")$sendKeysToElement(list('Oligochaeta'))  # Oligochète
mybrowser1$findElement(using="xpath","//select[@name='region']/option[@value='15']")$clickElement()
mybrowser1$findElement(using="xpath","//input[@name='include_doubtful_presence']")$clickElement()
mybrowser1$findElement(using="xpath","//input[@name='submit2']")$clickElement()

From this point I can download the xls file of the 20 subspecies by using:

mybrowser1$findElement(using = "xpath", "//a[@href='JavaScript:document.export_species_list.submit()']")$clickElement()

But that's not what I want, I don't want to use a "click". Is it possible to download the file from this JavaScript Link directly in my R environment or to scrape the table of the 20 subspecies directly from the source code of the webpage using Rselenium ?

I tried those two solutions but it's an impasse...The biggest problem is that the page is a temporary page or 'result page' and it seems that I can't find in it any @value, @id, @name or @class corresponding to the table I need.

Any clue on a solution which implied an automated way of doing it via R ? I need it in this form because the script have to be run afterwards by people who needs to create themself the results. Thanks in advance !

Answer 1

If you just want the table that is displayed on the website this can be done without Rselenium via httr as follows:

require(rvest)
require(httr)
res <- POST("http://www.faunaeur.org/species_list.php",
            encode = "form", 
            body = list(selected_regions="15",
                        show_what="species list",
                        referring_page="distribution",
                        taxon_rank="7",
                        taxon_name="Oligochaeta",
                        region="15",
                        include_doubtful_presence="yes",
                        submit2="Display Species",
                        show_what="species list",
                        species_or_higher_taxa="species"))
doc <- res %>% read_html
dat <- doc %>% html_table(fill=TRUE, ) %>% .[[9]]
colnames(dat) <- dat[1,]
dat <- dat[-1, ]

Which gives you:

            Family                      Species / subspecies
2  Acanthodrilidae       Microscolex dubius (Fletscher 1887)
3    Enchytraeidae      Enchytraeus buchholzi Vejdovsky 1878
4    Enchytraeidae     Fridericia berninii Dozsa-Farkas 1988
5    Enchytraeidae            Fridericia caprensis Bell 1947
...
21        Naididae           Aulophorus furcatus (Oken 1815)