Remove the testing shows whoops

Use this instead of the strange torrent class behaviour

README.md

@@ -1,7 +1,8 @@
 # Automagnet
 RSS feed parser, primarily for use on the seven seas.
 
-Extend by adding entries to the `providers` list. Includes sample provider for horriblesubs.info
+Extend by adding entries to the `providers` array in config.json.
+Includes sample provider for horriblesubs.info.
 
 ## Quickstart
 1. Install pip and pipenv
@@ -15,12 +16,29 @@ Extend by adding entries to the `providers` list. Includes sample provider for h
    pipenv install
    ```
 
-3. Add some shows to wants.txt
+3. Add some shows to the `wants` array in config.json
-   ```bash
+   ```json
-   echo 'weeb nonsense' >> wants.txt
+   {
+     "providers": [
+       ...
+     ],
+     "wants": [
+       {
+         "name": "Some weeb nonsense",
+         "target-quality": "1080p"
+       },
+       {
+         "name": "More weeb nonsense",
+         "min-quality": "720p"
+       },
+       {
+         "name": "Even more weeb nonsense"
+       }
+     ]
+   }
    ```
 
-4. Get some magnets, but only once, and do with them what you will.
+4. Get some magnets / torrents, but only once, and do with them what you will. Handy to run in a cronjob every so often
    ```bash
-   pipenv run ./automagnet.py has.txt wants.txt | transmission-cli -a
+   pipenv run ./automagnet.py | transmission-cli -a
    ```

automagnet.py

@@ -4,47 +4,179 @@ import feedparser
 import subprocess
 import click
 import re
+import json
+import hashlib
+import sys
+from operator import itemgetter
+
+def printe(*args, **kwargs):
+	print(*args, file=sys.stderr, **kwargs)
+
+class Resolution:
+	# In decending order of quality, since indexes are used to tell whats "better"
+	names = [
+		('8k', '8K', '4320p'),
+		('5k', '5K', '2880p', 'UHD+'),
+		('4k', '4K', '2160p', 'UHD'),
+		('2k', '2K', '1440p', 'QHD', 'WQHD'),
+		('1080p', 'FHD'),
+		('720p', 'HD', 'HDTV'),
+		('480p', 'SD', 'VGA'),
+		('360p', 'nHD'),
+		('240p', 'QVGA'),
+		('Unkown', 'None', None)
+	]
+	def __init__(self, _name):
+		self.name = _name
+		for family in self.names:
+			if self.name in family:
+				self.family = family
+				break
+		else:
+			raise ValueError("Did not recongnize resolution name. See Resolution.names")
+
+	# Rich comparison support
+	def __lt__(self, other):
+		if not isinstance(other, Resolution):
+			raise TypeError
+		return self.names.index(self.family) > other.names.index(other.family)
+	def __eq__(self, other):
+		if not isinstance(other, Resolution):
+			raise TypeError
+		return self.names.index(self.family) == other.names.index(other.family)
+	def __ne__(self, other):
+		return not self.__eq__(other)
+	def __gt__(self, other):
+		return not self.__lt__(other)
+	def __le__(self, other):
+		return self.__lt__(other) or self.__eq__(other)
+	def __ge__(self, other):
+		return self.__gt__(other) or self.__eq__(other)
+
+
+class Show:
+	def __init__(self, name, episode=None):
+		self.name = name
+		try:
+			self.episode = int(episode) if episode is not None else None
+		except (ValueError, TypeError):
+			raise ValueError("episode paramater must be a number or None")
+		self.key = '{name} - {episode}'.format(name=self.name, episode=self.episode if self.episode else 'None')
+
+	def getConfig(self, config):
+		try:
+			show_config = next(want for want in config['wants'] if want['name'] == self.name)
+		except StopIteration:
+			show_config = None
+		return show_config
+
+	def __repr__(self):
+		return "<Show {name} - {episode}>".format(name=self.name, episode=self.episode if self.episode else 'None')
+
+	def __hash__(self):
+		return hash(self.key)
+
+	def __eq__(self, other):
+		return self.key == other.key
+
+class Torrent:
+	def __init__(self, name, link, id=None, episode=None, quality=None, encoding=None):
+		self.name = name
+		self.link = link
+		self.id = id if id else hashlib.sha256(link).hexdigest()
+		try:
+			self.episode = int(episode) if episode is not None else None
+		except (ValueError, TypeError):
+			raise ValueError("episode paramater must be a number or None")
+		self.quality = Resolution(quality)
+		self.encoding = encoding
+
+	def isShow(self, show):
+		return self.name == show.name and (show.episode is None or self.episode == show.episode)
+
+	def __repr__(self):
+		return "<Torrent {id} for {name} - {episode}>".format(id=self.id, name=self.name, episode=self.episode)
+
+	def __hash__(self):
+		return hash(self.id)
+
+	def __eq__(self, other):
+		return self.id == other.id
 
-# List of provider RSS feeds to parse, and a regex we use to parse them. Regex should contain at least a "name" group
-providers = [
-	('https://horriblesubs.info/rss.php?res=all', '\[HorribleSubs\]\s(?P<name>.*?)(?:\s-\s(?P<episode>[0-9]+))?\s\[(?P<quality>[0-9]+.)\]\.(?P<encoding>.*)')
-]
 
 
 @click.command()
-@click.argument('has_path', type=click.Path(writable=True, dir_okay=False))
+@click.option('--config', 'config_path', type=click.Path(readable=True, dir_okay=False), default='config.json', help="Config file in JSON format.", )
-@click.argument('wants_file', type=click.File())
+@click.option('--seen-file', 'seen_path', type=click.Path(writable=True, dir_okay=False), default='seen.txt', help="File to store seen torrents in. Pruning is acceptable if pruned torrent id's will not appear in the RSS feeds.")
-def main(has_path, wants_file):
+def main(config_path='config.json', seen_path='seen.txt'):
-	# Get what we want and what we have from the files provided
+	# Get our config
-	wants = [line.strip() for line in wants_file if not line.strip().startswith('#')]
+	with open(config_path, 'r') as config_file:
-	has = [line.strip() for line in open(has_path, 'r')]
+		config_json = ''.join(config_file.readlines())
+	config = json.loads(config_json)
 
-	for provider, regex in providers:
+	providers = config['providers']
+	wants = [want['name'] for want in config['wants']]
+	seen = [line.strip() for line in open(seen_path, 'r')]
+
+	filtered = {}
+
+	for provider in providers:
 		# Compile and check our regex
-		title_parser = re.compile(regex)
+		title_parser = re.compile(provider['regex'])
 		if title_parser.groupindex.get('name') is None:
-			print("ERR: Regex for {url} does not contain a 'name' group")
+			printe("ERROR: Regex for {provider_name} does not contain a 'name' group. Name a capture group in your regex using '(?P<name>...)'".format(provider_name=provider['name']))
 			continue
+		elif title_parser.groupindex.get('quality') is None or title_parser.groupindex.get('episode') is None:
+			printe("WARNING: Regex for {provider_name} does not contain an 'episode' and/or 'quality' group. Unless you are sure duplicates don't exist in the feeds, add optional groups using '(?P<name>...)?'".format(provider_name=provider['name']))
+
 		# Get and parse rss feed
-		rss = feedparser.parse(provider)
+		rss = feedparser.parse(provider['url'])
 		for entry in rss.entries:
 			# Run regex against rss entry title (usually the file / torrent name)
 			matched_title = title_parser.match(entry.title)
 			if matched_title is None:
 				continue
-			# Check if its something we're interested in
 			parsed_title = matched_title.groupdict()
-			if parsed_title['name'].lower() in wants and entry.id not in has:
+			torrent = Torrent(
+				parsed_title['name'],
+				entry.get('link') or entry.links[0].href,
+				id=entry.get('id'),
+				episode=parsed_title.get('episode'),
+				quality=parsed_title.get('quality'),
+				encoding=parsed_title.get('encoding')
+			)
+			show = Show(torrent.name, episode=torrent.episode)
+			# Check if its something we're interested in and add it to a dict
+			if torrent.name in wants and torrent.id not in seen:
+				printe("Found new torrent: {torrent}".format(torrent=entry.title))
+				if filtered.get(show) is None:
+					filtered[show] = [torrent]
+				else:
+					filtered[show].append(torrent)
+				# Record that we have seen it
+				seen.append(torrent.id)
+
+	# Choose the best item
+	for show, torrents in filtered.items():
+		item_config = show.getConfig(config)
+		# Filter
+		if item_config.get('target-quality'):
+			torrents = [torrent for torrent in torrents if torrent.quality == Resolution(item_config['target-quality'])]
+		elif item_config.get('min-quality'):
+			torrents = [torrent for torrent in torrents if torrent.quality >= Resolution(item_config['min-quality'])]
+
+		if len(torrents) > 0:
+			best_entry = sorted(torrents, key=lambda t: t.quality)[0]
 			# Output the magnet / torrent link
+			printe("Choosing {torrent} as best choice for {show}".format(torrent=torrent.id, show=show.name))
 			# TODO: transmission-cli subprocess call here?
-				print(entry.link)
+			print(best_entry.link)
-				# Record that we have it
+		else:
-				# FIXME: should be derived from the magnet itself as not all providers are nice like this
+			printe("No torrents for {show} pass quality restrictions".format(show=show.name))
-				has.append(entry.id)
 
 	# Save our records to a file for later use
-	with open(has_path, 'w') as has_file:
+	with open(seen_path, 'w') as seen_file:
-		has_file.write('\n'.join(has))
+		seen_file.write('\n'.join(seen))
 
 # Usual python stuff
 if __name__ == '__main__':

config.json

@@ -0,0 +1,11 @@
+{
+  "providers": [
+    {
+      "name": "HorribleSubs",
+      "url": "https://horriblesubs.info/rss.php?res=all",
+      "regex": "\\[HorribleSubs\\]\\s(?P<name>.*?)(?:\\s-\\s(?P<episode>[0-9]+))?\\s\\[(?P<quality>[0-9]+.)\\]\\.(?P<encoding>.*)"
+    }
+  ],
+  "wants": [
+  ]
+}

seen.txt

@@ -0,0 +1 @@
+

wants.txt

@@ -1 +0,0 @@
-# Enter shows you want here in strictly lower case