How to get a hostname from a URL in Python

Pulling the host name out of a URL keeps logging, routing, allowlist checks, and endpoint comparisons focused on the network destination instead of the full string with credentials, ports, paths, queries, or fragments.

Python exposes URL parsing through urllib.parse. Current Python documentation says urlsplit() should generally be used instead of urlparse() unless path parameters must be split separately, and the returned SplitResult.hostname attribute exposes the normalized host while netloc keeps the original authority section.

A hostname is available only when the input includes a network location. Absolute URLs work directly, and scheme-less network URLs still need a leading // or Python will treat the value as a path. The masked examples below keep the shape of a credentialed API endpoint, a scheme-less service URL, and a hostless mail target, while urlsplit() still requires explicit scheme and host checks when the result affects security decisions.

1. Open a terminal application and start the Python 3 interpreter.

   $ python3
   Python 3.14.3
   >>>

2. Import urlsplit from the standard urllib.parse module.

   >>> from urllib.parse import urlsplit

3. Split the full URL string into its structured components.

   >>> parsed = urlsplit('https://svc_reader:REDACTED@API-01.Example.NET:8443/v1/health?trace=1#status')
   >>> parsed
   SplitResult(scheme='https', netloc='svc_reader:REDACTED@API-01.Example.NET:8443', path='/v1/health', query='trace=1', fragment='status')

   Current Python documentation says urlsplit() should generally be used instead of urlparse() unless the extra path-parameter split from urlparse() is required.

4. Read the hostname attribute when only the host name is needed.

   >>> parsed.hostname
   'api-01.example.net'

   hostname strips credentials and the port number, and normalizes the host name to lowercase.

5. Read netloc only when the raw authority section is still needed for another part of the program.

   >>> parsed.netloc
   'svc_reader:REDACTED@API-01.Example.NET:8443'

   Use netloc for the original authority string and hostname for host-only matching, logging, or filtering.

6. Keep the leading // on a scheme-less network URL so Python still recognizes the host component.

   >>> urlsplit('//edge-api.example.net/v1/health').hostname
   'edge-api.example.net'
   >>> urlsplit('edge-api.example.net/v1/health').hostname
   None

   Without the leading //, Python parses the value as a path and no hostname is returned.

7. Save the extraction in a short script when the same check needs to run from the shell or another tool.

   get-host-name-from-url.py

   #!/usr/bin/env python3
    
   from urllib.parse import urlsplit
   import sys
    
   if len(sys.argv) != 2:
       print(f"Usage: {sys.argv[0]} <url>")
       raise SystemExit(1)
    
   url = sys.argv[1]
   parsed = urlsplit(url)
    
   if parsed.hostname is None:
       print(f"No host name found in URL: {url}")
       raise SystemExit(1)
    
   print(parsed.hostname)

   The explicit None check keeps hostless URLs such as mailto:alerts@example.net from being treated as successful matches.

8. Run the script with a full URL and confirm that it prints only the host name.

   $ python3 get-host-name-from-url.py 'https://svc_reader:REDACTED@API-01.Example.NET:8443/v1/health?trace=1#status'
   api-01.example.net

9. Test a hostless URL and confirm that the script fails with a clear message and a non-zero exit status.

   $ python3 get-host-name-from-url.py 'mailto:alerts@example.net'
   No host name found in URL: mailto:alerts@example.net
   $ echo $?
   1

   The non-zero exit status makes the failure path easier to detect from shell scripts and automation.