I would like to work with RestRServe to have a .xlsx file uploaded for processing. I have tried the below using a .csv with success, but some slight modifications for .xlsx with get_file was not fruitful.
ps <- r_bg(function(){
library(RestRserve)
library(readr)
library(xlsx)
app = Application$new(content_type = "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet")
app$add_post(
path = "/echo",
FUN = function(request, response) {
cnt <- request$get_file("xls")
dt <- xlsx::read.xlsx(cnt, sheetIndex = 1, header = TRUE)
response$set_body("some function")
}
)
backend = BackendRserve$new()
backend$start(app, http_port = 65080)
})
What have you tried? According to the documentation request$get_file()
method returns a raw vector - a binary representation of the file. I'm not aware of R packages/functions which allow to read xls/xlsx file directly from the raw vector (probably such functions exist, I just don't know).
Here you can write body to a file and then read it normal way then:
library(RestRserve)
library(readxl)
app = Application$new()
app$add_post(
path = "/xls",
FUN = function(request, response) {
fl = tempfile(fileext = '.xlsx')
xls = request$get_file("xls")
# need to drop attributes as writeBin()
# can't write object with attributes
attributes(xls) = NULL
writeBin(xls, fl)
xls = readxl::read_excel(fl, sheet = 1)
response$set_body("done")
}
)
backend = BackendRserve$new()
backend$start(app, http_port = 65080)
Also mind that content_type
argument is for response encoding, not for request decoding.