cURL - command line tool for URL

“cURL” to transfer data in URL syntax

“cURL” is a powerful command line tool and library for transferring data with URLs
(since 1998) – quoted in https://curl.se , we commonly use it to send the HTTP GET and POST requests especially in terminal mode. We will have some basic usage to share here.

Get the http response code

It’s very common to use “curl” to collect the http response code in shell environment, e.g. you may need to verify if remote URLs are working fine (http code = 200).

# the "\n" means to print the newline at the end of http_code
# "-s" and "-o /dev/null" just to omit the length output

$ curl -w '%{http_code}\n' -s -o /dev/null "https://google.com"
301

$ curl -w '%{http_code}\n' -s -o /dev/null "https://www.yahoo.com"
200

Collect the response time

As you can see in previous example, we use %{http_code} to get the http response code, we have other parameters to be used when we want to collect the latency, e.g.

$ curl -s -o /dev/null -w '\nEstablish Connection: %{time_connect}s\nTTFB: %{time_starttransfer}s\nTotal: %{time_total}s\n' "https://www.yahoo.com"

Establish Connection: 0.010960s
TTFB: 0.244146s
Total: 0.883493s

You can also to create a TEXT file, e.g. curl.txt and then let “curl” to use it to get formatted output.

$ cat curl.txt
     time_namelookup:  %{time_namelookup}s\n
        TCP_Connect:   %{time_connect}\n
        SSL_Connect:   %{time_appconnect}\n
        time_connect:  %{time_connect}s\n
     time_appconnect:  %{time_appconnect}s\n
    time_pretransfer:  %{time_pretransfer}s\n
       time_redirect:  %{time_redirect}s\n
  time_starttransfer:  %{time_starttransfer}s\n
                     ----------\n
          time_total:  %{time_total}s\n

# and then run the "curl" command
$ curl -w '@curl.txt' -s -o /dev/null "https://www.yahoo.com"
     time_namelookup:  0.009075s
        TCP_Connect:   0.012072
        SSL_Connect:   0.042414
        time_connect:  0.012072s
     time_appconnect:  0.042414s
    time_pretransfer:  0.042584s
       time_redirect:  0.000000s
  time_starttransfer:  0.242770s
                     ----------
          time_total:  0.886892s

Check “curl” manual page to see what other available variables you can use. That’s a lot..:)

Send multiple requests in one shot

You can repeat the commands by 5 times, or you can try to use :

$ curl -w '%{http_code}\n' -s -o /dev/null "https://search.yahoo.com/search?p=apple"

# add "?n=[1-5]" in the end of curl command:
$ curl -w '%{http_code}\n' -s -o /dev/null "https://search.yahoo.com/search?p=apple?n=[1-5]"
200
200
200
200
200

Use “curl –resolve” to change the actual destination

This is useful when you want to test the original in a particular IP address, let’s say you have 5 hosts behind : www.yahoo.com and you can use the following sample to connect to particular one :

curl --resolve www.yahoo.com:443:10.10.10.1 https://www.yahoo.com/

Get the size of response header and body

As we said, there’re lots of variables available in curl to be use, and “size_header” and “size_download” can be used in this case.

$ curl -s -o /dev/null -w 'Header: %{size_header}\nBody: %{size_download}\n' "https://www.google.com/search?q=apple"
Header: 1100
Body: 139165

cURL – command line tool for URL