trailforks_html_to_gpx.py

#!/usr/bin/env python3
import sys
import json
from html.parser import HTMLParser
import re


gpx_template = """<?xml version='1.0' encoding='UTF-8'?>
<gpx version="1.1"
creator="skookum"
xmlns="http://www.topografix.com/GPX/1/1"
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xsi:schemaLocation="http://www.topografix.com/GPX/1/1
http://www.topografix.com/GPX/1/1/gpx.xsd">
<trk>
<name>%s</name>
<trkseg>
%s</trkseg>
</trk>
</gpx>
"""

trkpt_template = '<trkpt lat="%s" lon="%s"></trkpt>\n'


class TrailforksParser(HTMLParser):
    def __init__(self):
        super().__init__()
        self.points = list()
        self.title = "Trail"

    def handle_starttag(self, tag, attrs):
        if tag == "meta":
            d = dict(attrs)
            if d.get('property', None) == "og:title":
                self.title = d.get('content', self.title)

    def find_sections_with_points(self, data: str):
        for line in data.split("\n"):  # assume 'sections:" is on its own line
            m = re.match(r'\s*sections: \[(.*)\],\s*$', line)
            if m:
                return m.group(1)
        assert False, "Found no sections variable"

    def handle_data(self, data):
        if 'sections:' in data:
            # print("Encountered some data  :", data)
            sections = self.find_sections_with_points(data)
            d = json.loads(sections)
            text_points = d['points']
            self.points += self.extract_point_pairs(text_points)

    def extract_point_pairs(self, text_points: dict) -> list[str, str]:
        return ((elem['lat'], elem['lng']) for elem in text_points)

    def form_gpx(self) -> str:
        trkpoints = "".join((trkpt_template % entry) for entry in self.points)
        gpx_text = gpx_template % (self.title, trkpoints)
        return gpx_text


def main(argv):
    parser = TrailforksParser()
    with open(sys.argv[1]) as f:
        parser.feed(f.read())
    assert len(parser.points), "Collected GPX points are empty"
    return parser.form_gpx()


if __name__ == "__main__":
    print(main(sys.argv[1]))
    sys.exit(0)