| ok {crul} | R Documentation |
check if a url is okay
Description
check if a url is okay
Usage
ok(x, status = 200L, info = TRUE, verb = "head", ua_random = FALSE, ...)
Arguments
x |
either a URL as a character string, or an object of class HttpClient |
status |
(integer) one or more HTTP status codes, must be integers.
default: |
info |
(logical) in the case of an error, do you want a
|
verb |
(character) use "head" (default) or "get" HTTP verb for the request. note that "get" will take longer as it returns a body. however, "verb=get" may be your only option if a url blocks head requests |
ua_random |
(logical) use a random user agent string?
default: |
... |
args passed on to HttpClient |
Details
We internally verify that status is an integer and in the known set of HTTP status codes, and that info is a boolean
You may have to fiddle with the parameters to ok() as well as
curl options to get the "right answer". If you think you are
incorrectly getting FALSE, the first thing to do is to pass in
verbose=TRUE to ok(). That will give you verbose curl output and will
help determine what the issue may be. Here's some different scenarios:
the site blocks head requests: some sites do this, try
verb="get"it will be hard to determine a site that requires this, but it's worth trying a random useragent string, e.g.,
ok(useragent = "foobar")some sites are up and reachable but you could get a 403 Unauthorized error, there's nothing you can do in this case other than having access
its possible to get a weird HTTP status code, e.g., LinkedIn gives a 999 code, they're trying to prevent any programmatic access
A FALSE result may be incorrect depending on the use case. For example,
if you want to know if curl based scraping will work without fiddling with
curl options, then the FALSE is probably correct, but if you want to
fiddle with curl options, then first step would be to send verbose=TRUE
to see whats going on with any redirects and headers. You can set headers,
user agent strings, etc. to get closer to the request you want to know
about. Note that a user agent string is always passed by default, but it
may not be the one you want.
Value
a single boolean, if TRUE the URL is up and okay,
if FALSE it is down; but, see Details
Examples
## Not run:
# 200
ok("https://www.google.com")
# 200
ok("https://hb.opencpu.org/status/200")
# more than one status
ok("https://www.google.com", status = c(200L, 202L))
# 404
ok("https://hb.opencpu.org/status/404")
# doesn't exist
ok("https://stuff.bar")
# doesn't exist
ok("stuff")
# use get verb instead of head
ok("http://animalnexus.ca")
ok("http://animalnexus.ca", verb = "get")
# some urls will require a different useragent string
# they probably regex the useragent string
ok("https://doi.org/10.1093/chemse/bjq042")
ok("https://doi.org/10.1093/chemse/bjq042", verb = "get", useragent = "foobar")
# with random user agent's
## here, use a request hook to print out just the user agent string so
## we can see what user agent string is being sent off
fun_ua <- function(request) {
message(paste0("User-agent: ", request$options$useragent), sep = "\n")
}
z <- crul::HttpClient$new("https://doi.org/10.1093/chemse/bjq042",
hooks = list(request = fun_ua))
z
replicate(5, ok(z, ua_random=TRUE), simplify=FALSE)
## if you set useragent option it will override ua_random=TRUE
ok("https://doi.org/10.1093/chemse/bjq042", useragent="foobar", ua_random=TRUE)
# with HttpClient
z <- crul::HttpClient$new("https://hb.opencpu.org/status/404",
opts = list(verbose = TRUE))
ok(z)
## End(Not run)