# download GPX points from https://geocaching.su/?pn=103 select region and download gpx
file_from_site = "import_points.gpx"
result_for_osmand = "osmand_result.gpx"


import xml.etree.ElementTree as ET
from html.parser import HTMLParser

class HTMLFilter(HTMLParser):
    text = ""
    def handle_data(self, data):
        self.text += data

tree = ET.parse(file_from_site)
root = tree.getroot()

bad_types = ["Waypoint|Trailhead", "Waypoint|Question to Answer", "Waypoint|Reference Point",
            "Waypoint|Final location", "Waypoint|Virtual Stage", "Waypoint|Parking Area"]

out_gpx = ET.Element('gpx')

for wpt in root.findall('{http://www.topografix.com/GPX/1/0}wpt'):
    cache_type = wpt.find('{http://www.topografix.com/GPX/1/0}type').text
    if cache_type in bad_types: continue
    
    lat = wpt.get('lat')
    lon = wpt.get('lon')
    desc = wpt.find('{http://www.topografix.com/GPX/1/0}desc')

    if desc is not None:
        name = desc.text
    else:
        name = ""

    url = wpt.find('{http://www.topografix.com/GPX/1/0}url')
    if url is None:
        url = ""
    else:
        url = url.text

    full_description = wpt.find('{http://www.groundspeak.com/cache/1/0}cache').\
            find('{http://www.groundspeak.com/cache/1/0}long_description').text
    f = HTMLFilter()
    f.feed(full_description)
    full_description = f.text

    comments = []
    comments_cache = wpt.find("{http://www.groundspeak.com/cache/1/0}cache")
    if comments_cache is not None:
        comments_tags = comments_cache.find("{http://www.groundspeak.com/cache/1/0}logs")
        if  comments_tags is not None:
            for i in comments_tags:
                html_conv2 = HTMLFilter()
                comment_html = i.find("{http://www.groundspeak.com/cache/1/0}text")
                if comment_html.text is None: continue
                html_conv2.feed(comment_html.text)
                comment_clean = html_conv2.text
                comments.append(comment_clean)

    # now add to osmand GPX
    geocache = ET.SubElement(out_gpx, 'wpt')
    geocache.set('lat', lat)
    geocache.set('lon', lon)

    name_tag = ET.SubElement(geocache, 'name')
    name_tag.text = name

    # adding comments to description
    for comment in comments:
        full_description = full_description + "\n------------------\n" + comment

    desc_tag = ET.SubElement(geocache, 'desc')
    desc_tag.text = url + "\n\n" + full_description


tree_out = ET.ElementTree(out_gpx)
tree_out.write(result_for_osmand, encoding="utf-8")