curl

curl(1) is a command-line tool for fetching internet (chiefly web or FTP) content.

Contents

  1. curl
    1. Tips
      1. Suppressing Progress Bar
      2. Downloading a File
      3. Checking Availability

Tips

Suppressing Progress Bar

curl --no-progress-meter https://example.com

Downloading a File

curl --output path/to/file --fail https://example.com

In the case of a server error (HTTP 4xx), an error page is often returned by the server. The --fail option suppresses this page from output and causes curl(1) to exit with code 22.

Checking Availability

code=$(curl --silent --fail --head --write-out '%{http_code}' https://example.com)
case $code in
  0)
    echo "Success! Server returned HTTP code ${code}"
    ;;
  22)
    echo "Error! Server returned HTTP code ${code}"
    ;;
  6)
    echo "Error! Server not found"
    ;;
  60)
    echo "Error! SSL/TLS certificate cannot be verified"
    ;;
  *)
    echo "Error!"
    ;;
esac

The --head option causes curl(1) to only fetch only headers. The --silent option suppresses all output, including the progress bar or any returned pages.

The --write-out option takes a format string that will be populated with available variables.

Other useful variables:

Name

Contains

content_type

Content type of the document

http_version

Version of HTTP that was used

size_header

Number of bytes in downloaded headers

size_download

Number of bytes in downloaded body and data (i.e. not headers)

time_namelookup

Number of seconds from start until name was resolved

time_connect

Number of seconds from start until TCP connection

time_appconnect

Number of seconds from start until SSL handshake

time_pretransfer

Number of seconds from start until all pre-transfer steps are complete

time_starttransfer

Number of seconds from start until first transferred byte is received

time_total

Number of seconds

speed_download

Average number of bytes per second

ssl_verify_result

Result of certificate verification (0 if successful)

CategoryRicottone