#!/usr/bin/env python3

#
# bbb-download.py
#
# BBB recording downloader
# Tries to record all files that are part of a BBB recording "as-is" to disk
#
# Copyright (c) 2023 Sylvain Munaut <tnt@246tNt.com>
# SPDX-License-Identifier: MIT
#

import argparse
import json
import pathlib
import re
import requests
import urllib.parse

import bbb


def bbb_slides_dl(fh):

	rv = []
	doc = json.load(fh)

	for fid,fc in doc.items():
		# Validate file id
		if not re.match(r'^[0-9a-fA-F]{40,64}-[0-9]+$', fid):
			raise RuntimeError('Unknown document ID format in presentation_text.json')

		# Iterate pages
		for fp in fc.keys():
			# Validate it's a slide
			if not fp.startswith('slide-'):
				raise RuntimeError('Unknown document page type in presentation_text.json')

			# Add to the list of "to download"
			rv.append((f'presentation/{fid:s}/{fp:s}.png', ['presentation', fid, f'{fp:s}.png']))

	return rv


def bbb_shapes_svg_check(fh):
	print(bbb.parse_shapes_svg(fh))
	return []


BASE_DL = [
	# URL Path                    Local Path               Handler
	( 'captions.json',            'captions.json',          ),
	( 'cursor.xml',               'cursor.xml',             ),
	( 'deskshare.xml',            'deskshare.xml',          ),
	( 'deskshare/deskshare.webm', 'deskshare.webm',         ),
	( 'external_videos.json',     'external_videos.json',   ),
	( 'metadata.xml',             'metadata.xml',           ),
	( 'notes.xml',                'notes.xml',              ),
	( 'panzooms.xml',             'panzooms.xml',           ),
	( 'polls.json',               'polls.json',             ),
	( 'presentation_text.json',   'presentation_text.json', bbb_slides_dl ),
	( 'shapes.svg',               'shapes.svg',             bbb_shapes_svg_check ),
	( 'slides_new.xml',           'slides_new.xml',         ),
	( 'video/webcams.webm',       'webcams.webm',           ),
]


def download_file(url, filename):
	with requests.get(url, stream=True) as r:
		r.raise_for_status()
		with open(filename, 'wb') as f:
			for chunk in r.iter_content(chunk_size=128*1024):
				f.write(chunk)


def download_all_recursive(opts):

	# Create the base directory
	p_base = opts.dest.absolute()
	p_base.mkdir(parents=True)

	# Create the base url
	u_base = opts.url
	u_base = urllib.parse.urljoin(u_base, f'presentation/{opts.meeting:s}/')

	# Loop init
	to_dl = list(BASE_DL)
	i = 0

	# Iterate until we're done
	while i < len(to_dl):
		# Entry to fetch
		e = to_dl[i]

		e_url     = e[0]
		e_path    = e[1]
		e_handler = e[2] if len(e) > 2 else None

		# Source URI
		e_url = urllib.parse.urljoin(u_base, e_url)

		# Destination path
		if not isinstance(e_path, list):
			e_path = [e_path]
		e_path = pathlib.Path(p_base, *e_path)
		e_path.parent.mkdir(parents=True, exist_ok=True)

		# Trigger fetch
		print(e_url, e_path)
		try:
			download_file(e_url, e_path)
		except:
			pass

		# If we have a handler, call it
		if callable(e_handler):
			with open(e_path, 'rb') as fh:
				to_dl.extend(e_handler(fh))

		# Next
		i = i + 1


def parse_opt():
	parser = argparse.ArgumentParser(
		prog = 'bbb-download.py',
		description = 'BBB recording downloader',
	)

	parser.add_argument('-u', '--url', required=True,
		help="BBB instance base URL")
	parser.add_argument('-m', '--meeting', required=True,
		help="Meeting ID")
	parser.add_argument('-d', '--dest', required=True, type=pathlib.Path,
		help="Destination directory")

	return parser.parse_args()


def main():
	opts = parse_opt()
	download_all_recursive(opts)


if __name__ == '__main__':
	main()