lapply 和 read_xml.character

Question

Iam trying to extract data from a website using a custom function:

library(tidyverse)
library(rvest)
url = "https://www.boerse.de/fundamental-analyse/garbage/" # last part does not change outcome, therefore 'garbage'
read_html_tables = function(ISIN){
  content <- read_html(paste0(url,ISIN,"#guv")) %>%
    html_table(dec = ",") %>%
    .[c(5:10)]
  return(content)
}

If I run this function with a given ISIN, eg US88579Y1010, I get the desired result. A list containing 6 tibbles with the data I want. But if I wrap this function into lapply() with a vector containing a few hundred ISIN, I get the following error:

list_of_all <- lapply(X = df[,2], FUN = read_html_tables)

Error: x must be a string of length 1 Called from: read_xml.character(x, encoding = encoding, ..., as_html = TRUE, options = options)

If I call which(length(df[,2]) != 1) (the column where the ISINs are), I get integer(0), so there seems to be no issue with the ISIN column in this dataframe. And since it works with a single ISIN as input, the read_html(paste0(url,ISIN)) part seems to work as well.

I have used a very similar function before and wrapped it into lapply() . The earlier function did basically exactly what this function does, but had to do some searching and combining for the correct URL to pass into the read_html(paste0(url,ISIN)) part (on another website). Iam a bit puzzled, since this error did not occure beforehand. But if it occured and I try to run the earlier function now, I get the same error (which I didn't receive any time before).

Maybe there is a more talented R-programmer out there which can spot the issue?

Edit: Since a reply suggested the ISIN-list is the issue: The first two are US88579Y1010 and US8318652091. Passed individually into the function as well as passing it in a vector ( c(ISIN1, ISIN2) ) and passing the vector to lapply works. But if I point at both ISINs inside the tibble ( df[1:2,2] ) I get the error from above. What am I missing here?

Answer 1

Solution: read_xml.character from read_html() seems to not accept a column from a tibble as valid input. Transfering the tibble to a data.frame and recalculating gives the desired output.