convert-delicious-to-webloc

#!/usr/bin/env python3

import os
import re
import requests
import sys
import xml.dom.minidom

from xml.dom.minidom import Node

PREAMBLE='''<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
<plist version="1.0">
<dict>
	<key>URL</key>
'''
POSTAMBLE='''
</dict>
</plist>
'''

def tidy_filename(filenameToTidy):
    newFilenameToTidy = re.sub('[^\w\-_\. ]', '_', filenameToTidy)
    return newFilenameToTidy

filename = sys.argv[1]
targetdirectory = sys.argv[2]

dom = xml.dom.minidom.parse(filename)

posts = dom.getElementsByTagName('post')

for post in posts:
    description = post.attributes['description'].value
    tags = post.attributes['tag'].value.split()
    url = post.attributes['href'].value
    if(not ('from' in tags and 'twitter' in tags)):
        assert(description != '' and description is not None)
        assert(url != '' and url is not None)
        try:
            request = requests.get(url)
            if request.status_code != 200:
                print("URL " + url + " returns status code " + str(request.status_code) + ", ignoring.", file=sys.stderr)
            else:
                for tag in tags:
                    targetdir = os.path.join(targetdirectory, tag)
                    os.makedirs(targetdir, exist_ok=True)
                    contents = PREAMBLE + '\t<string>' + url + '</string>' + POSTAMBLE
                    new_filename = os.path.join(targetdir, tidy_filename(description) + '.webloc')
                    with open(new_filename, 'w') as f:
                        f.write(contents)
        except requests.exceptions.RequestException as e:
            print("URL " + url + " failed with error " + str(e) + ", ignoring.", file=sys.stderr)