Fix the scanner

INSTALLING.md

@@ -10,17 +10,16 @@ Kyoo also needs 3 files to work properly. Two of them can simply be copy-pasted
 Those files can be put in any directory of your choice.
 
 Those 3 files are:
- - A `docker-compose.yml` (simply copy docker-compose.prod.yml from [here](https://raw.githubusercontent.com/zoriya/Kyoo/master/docker-compose.prod.yml)).
- - A `nginx.conf.template` copied from [here](https://raw.githubusercontent.com/zoriya/Kyoo/master/nginx.conf.template).
- - A `.env` file that you will need to **fill**. Look at the example [.env.example](https://raw.githubusercontent.com/zoriya/Kyoo/master/.env.example)
 
+- A `docker-compose.yml` (simply copy docker-compose.prod.yml from [here](https://raw.githubusercontent.com/zoriya/Kyoo/master/docker-compose.prod.yml)).
+- A `nginx.conf.template` copied from [here](https://raw.githubusercontent.com/zoriya/Kyoo/master/nginx.conf.template).
+- A `.env` file that you will need to **fill**. Look at the example [.env.example](https://raw.githubusercontent.com/zoriya/Kyoo/master/.env.example)
 
 > If you want an explanation of what are those files, you can read the following:
 > The `docker-compose.yml` file describes the different services of Kyoo, where they should be downloaded and their start order. \
 > The `nignx.conf.template` file describes which service will be called when accessing the URL of Kyoo. \
 > The `.env` file contains all the configuration options that the services in `docker-compose.yml` will read.
 
-
 To retrieve metadata, Kyoo will need to communicate with an external service. For now, that is `the movie database`.
 For this purpose, you will need to get an API Key. For that, go to [themoviedb.org](https://www.themoviedb.org/) and create an account, then
 go [here](https://www.themoviedb.org/settings/api) and copy the `API Key (v3 auth)`, paste it after the `THEMOVIEDB_APIKEY=` on the `.env` file.
@@ -34,9 +33,9 @@ Congratulation, everything is now ready to use Kyoo. You can navigate to `http:/
 
 1. Install docker & docker-compose
 2. Download the
-[`docker-compose.yml`](https://raw.githubusercontent.com/zoriya/Kyoo/master/docker-compose.prod.yml),
-[`nginx.conf.template`](https://raw.githubusercontent.com/zoriya/Kyoo/master/nginx.conf.template) and
-[`.env`](https://raw.githubusercontent.com/zoriya/Kyoo/master/.env.example) files
+   [`docker-compose.yml`](https://raw.githubusercontent.com/zoriya/Kyoo/master/docker-compose.prod.yml),
+   [`nginx.conf.template`](https://raw.githubusercontent.com/zoriya/Kyoo/master/nginx.conf.template) and
+   [`.env`](https://raw.githubusercontent.com/zoriya/Kyoo/master/.env.example) files
 3. Fill the `.env` file with your configuration options (and an API Key from [themoviedb.org](https://www.themoviedb.org/))
 4. Run `docker-compose up -d`
 
@@ -48,6 +47,9 @@ unsure that your `.env` contains all the options specified in the updated `.env.
 After that, you will need to update Kyoo's services. For that, open a terminal in the configuration's directory and run
 the command `docker-compose pull`. You are now ready to restart Kyoo, you can run `docker-compose up -d`.
 
+You can also enable automatic updates via an external tool like [watchtower](https://containrrr.dev/watchtower/).
+TLDR: `docker run -d --name watchtower -e WATCHTOWER_CLEANUP=true -e WATCHTOWER_POLL_INTERVAL=86400 -v /var/run/docker.sock:/var/run/docker.sock containrrr/watchtower`
+
 # Uninstalling
 
 To uninstall Kyoo, you need to open a terminal in the configuration's directory and run `docker-compose down`. This will

scanner/requirements.txt

@@ -2,4 +2,4 @@ guessit
 aiohttp
 jsons
 black-with-tabs
-watchdog
+watchfiles

scanner/scanner/monitor.py

@@ -1,74 +1,19 @@
-import asyncio
-from functools import wraps
-from watchdog.observers import Observer
-from watchdog.events import (
-	FileSystemEventHandler,
-	DirCreatedEvent,
-	FileCreatedEvent,
-	DirMovedEvent,
-	FileMovedEvent,
-	DirDeletedEvent,
-	FileDeletedEvent,
-)
-
-from scanner.utils import log_errors
-
+import logging
+from watchfiles import awatch, Change
+from .utils import ProviderError
 from .scanner import Scanner
 
-task_list = []
-event = asyncio.Event()
-
 
 async def monitor(path: str, scanner: Scanner):
-	global task_list
-
-	observer = Observer()
-	handler = EventHandler(scanner)
-	observer.schedule(handler, path, recursive=True)
-	observer.start()
-
-	while True:
-		if any(task_list):
-			tl = task_list
-			task_list = []
-			await asyncio.gather(*tl)
-		await event.wait()
-		event.clear()
-	# Should call .join() if the while stops one day.
-	# observer.join()
-
-
-def async_event(f):
-	# Log errors of f and catch them to prevent the gather to throw.
-	f = log_errors(f)
-
-	@wraps(f)
-	def internal(*args, **kwargs):
-		task_list.append(f(*args, **kwargs))
-		event.set()
-
-	return internal
-
-
-class EventHandler(FileSystemEventHandler):
-	def __init__(self, scanner: Scanner):
-		self._scanner = scanner
-
-	@async_event
-	async def on_created(self, event: DirCreatedEvent | FileCreatedEvent):
-		if event.is_directory:
-			return
-		await self._scanner.identify(event.src_path)
-
-	# TODO: Implement the following two methods
-	def on_moved(self, event: DirMovedEvent | FileMovedEvent):
-		if event.is_directory:
-			# TODO: Check if this event is also called for files in the directory or not.
-			return
-		print(event.src_path, event.dest_path)
-
-	def on_deleted(self, event: DirDeletedEvent | FileDeletedEvent):
-		if event.is_directory:
-			# TODO: Check if this event is also called for files in the directory or not.
-			return
-		print(event.src_path)
+	async for changes in awatch(path):
+		for (event, file) in changes:
+			try:
+				if event == Change.added:
+					await scanner.identify(file)
+				else:
+					print(f"Change {event} occured for file {file}")
+			except ProviderError as e:
+				logging.error(str(e))
+			except Exception as e:
+				logging.exception("Unhandled error", exc_info=e)
+		print("end", flush=True)

scanner/scanner/scanner.py

@@ -5,6 +5,7 @@
 from aiohttp import ClientSession
 from pathlib import Path
 from guessit import guessit
+from typing import List
 from providers.provider import Provider
 from providers.types.episode import Episode, PartialShow
 from providers.types.season import Season, SeasonTranslation
@@ -24,34 +25,32 @@ def __init__(
 
 	async def scan(self, path: str):
 		logging.info("Starting the scan. It can take some times...")
-		videos = filter(lambda p: p.is_file(), Path(path).rglob("*"))
+		self.registered = await self.get_registered_paths()
+		videos = (str(p) for p in Path(path).rglob("*") if p.is_file())
 		# We batch videos by 20 because too mutch at once kinda DDOS everything.
 		for group in batch(videos, 20):
 			logging.info("Batch finished. Starting a new one")
 			await asyncio.gather(*map(self.identify, group))
 
-	async def is_registered(self, path: Path) -> bool:
+	async def get_registered_paths(self) -> List[Path]:
 		# TODO: Once movies are separated from the api, a new endpoint should be created to check for paths.
 		async with self._client.get(
-			f"{self._url}/episodes/count",
-			params={"path": f"eq:{path}"},
+			f"{self._url}/episodes",
+			params={"limit": 0},
 			headers={"X-API-Key": self._api_key},
 		) as r:
 			r.raise_for_status()
-			ret = await r.text()
-			if ret != "0":
-				return True
-		return False
+			ret = await r.json()
+			return list(x["path"] for x in ret["items"])
 
 	@log_errors
-	async def identify(self, path: Path):
+	async def identify(self, path: str):
+		if path in self.registered:
+			return
+
 		raw = guessit(path, "--episode-prefer-number")
 
-		if (
-			not "mimetype" in raw
-			or not raw["mimetype"].startswith("video")
-			or await self.is_registered(path)
-		):
+		if not "mimetype" in raw or not raw["mimetype"].startswith("video"):
 			return
 
 		logging.info("Identied %s: %s", path, raw)