Accessing an API with R - r

Question

I am trying to retrieve information from an API, which gives the name of the product from the barcode, through the API.
I am using the httr::GET().
The URL needed for the API contains the barcode itself, but I do not know how to automate the system so it can read the barcode contained in every entry, and plugging it into the url without me copying and pasting the code manually in the script.
one_code <- GET("api.upcdatabase.org/json/aebfed7a26f24a05efd7f77‌​749dc2fcc/…")
result <- content(one_code)
result$description

Answer 1

A couple extra things to consider.
First, the site provides https for the API so it should be used since you're exposing your API key on any network you make requests from otherwise.
Test the core HTTP status code and halt on major HTTP errors (not API errors).
You should also put your API key in something like an environment variable so it never ends up in scripts or GitHub repo commits. Use ~/.Renviron (make a single line entry for UPCDATABASE_API_KEY=your_key and then restart R).
You should handle error and success conditions and consider returning a data frame so you can have all the fields in a tidy, accessible fashion.
Finally, do some basic type conversion prior to returning the values to make return field values easier to use.
library(httr)
library(jsonlite)
library(purrr)
get_upc_code_info <- function(code, api_key=Sys.getenv("UPCDATABASE_API_KEY")) {
URL <- sprintf("https://api.upcdatabase.org/json/%s/%s", api_key, code)
res <- GET(URL)
stop_for_status(res)
res <- content(res, as="text", encoding="UTF-8")
res <- fromJSON(res, flatten=TRUE)
if (res$valid == "true") {
res <- flatten_df(res)
res$valid <- TRUE
res$avg_price <- as.numeric(res$avg_price)
res$rate_up <- as.numeric(res$rate_up)
res$rate_down <- as.numeric(res$rate_down)
return(res)
} else {
message(res$reason)
return(data.frame(number = code, valid = FALSE, stringsAsFactors=FALSE))
}
}
xdf <- get_upc_code_info("0111222333446")
dplyr::glimpse(xdf)
## Observations: 1
## Variables: 8
## $ valid <lgl> TRUE
## $ number <chr> "0111222333446"
## $ itemname <chr> "UPC Database Testing Code"
## $ alias <chr> "Testing Code"
## $ description <chr> "http://upcdatabase.org/code/0111222333446"
## $ avg_price <dbl> 123.45
## $ rate_up <dbl> 14
## $ rate_down <dbl> 3
Similar to what Aurèle suggested, you can use the function to make it easier to get multiple codes. Since this function returns a data frame, you can easily get a larger, complete data frame from individual lookups with purrr::map_df():
codes <- c("0057000006976", "3228881010711", "0817346023170", "44xx4444444")
xdf <- map_df(codes, get_upc_code_info)
dplyr::glimpse(xdf)
## Observations: 4
## Variables: 8
## $ valid <lgl> TRUE, TRUE, TRUE, FALSE
## $ number <chr> "0057000006976", "3228881010711", "0817346023170",...
## $ itemname <chr> "Heinz Original Beans (Pork & Molasses)", "Lip...
## $ alias <chr> "", "", "", NA
## $ description <chr> "", "Boîte de 20 sachets", "", NA
## $ avg_price <dbl> NA, NA, 39.99, NA
## $ rate_up <dbl> 0, 0, 1, NA
## $ rate_down <dbl> 0, 0, 0, NA
Consider putting finishing touches on this, adding a function to POST to the API, possibly make a Shiny app so folks can submit new entries through R and turning it into a package. You might even get extra free credits on the site if you do so.

Answer 2

Store your barcodes in one data structure (list or vector):
barcodes <- c(
"aebfed7a26f24a05efd7f77749dc2fcc",
"xyz1234567f24a05efd7f77749dc2fcc",
"pqr9876543f24a05efd7f77749dc2fcc"
)
Write a function:
scrape <- function(barcode) {
sample=GET(paste0("api.upcdatabase.org/json/", barcode, "/rest/of/the/url"))
result=content(sample)
result$description
}
And apply:
res <- lapply(barcodes, scrape)
The results are stored in a single list, so that they're easier to manipulate.