Re: [apps-discuss] Fun with URLs and regex

[Sam Ruby <[e-mail]>]

On 01/07/2015 04:35 PM, Mark Nottingham wrote:
> I’ve updated my Python script that serves as a translation of ABNF for URIs into regex.
>
> https://gist.github.com/mnot/138549
>
> It now validates the following URI schemes according to their respective specifications:
>    - http
>    - https
>    - file
>    - data
>    - gopher
>    - ws
>    - wss
>    - mailto

My test data:

http://intertwingly.net/stories/2014/10/05/urltestdata.json

A program to test each input:

import uri_validate
import json
import re

f = open('urltestdata.json')
tests = json.load(f)
f.close()

valid = {}
for test in tests:
   instr = test['input']
   valid[instr] = False
   if re.match("^%s$" % uri_validate.URI_reference, instr, re.VERBOSE):
     try:
       scheme_validator = "%s_URI" % instr.split(":", 1)[0].lower()
       validator = getattr(uri_validate, scheme_validator)
       if re.match("^%s#" % validator, instr, re.VERBOSE):
         valid[instr] = True
     except AttributeError:
       valid[instr] = True

print json.dumps(valid, indent=2)