From 0b22bf5796ed1ab2ccc2d57e8a3dd5cfc7f21788 Mon Sep 17 00:00:00 2001
From: Eli Ribble <eli@theribbles.org>
Date: Wed, 31 Jul 2024 13:05:25 -0700
Subject: [PATCH] Add logic to talk to paperless

Took me a while to figure out how to set the title parameter correctly.
Even still, like this it shows processing the file as the original file
name and then switches the title when processing is complete.
---
 scan_uploader/__init__.py | 62 ++++++++++++++++++++++++++++++++++-----
 1 file changed, 55 insertions(+), 7 deletions(-)

diff --git a/scan_uploader/__init__.py b/scan_uploader/__init__.py
index 2d722d6..19b896c 100644
--- a/scan_uploader/__init__.py
+++ b/scan_uploader/__init__.py
@@ -1,31 +1,79 @@
 import argparse
-import inotify.adapters
+import datetime
 import logging
 from pathlib import Path
+import os
+
+import inotify.adapters
+import requests
+import requests.auth
 
 LOGGER = logging.getLogger(__name__)
 
+class TokenAuth(requests.auth.AuthBase):
+	def __init__(self, token: str):
+		self.token = token
+
+	def __call__(self, request: requests.Request):
+		request.headers["Authorization"] = f"Token {self.token}"
+		return request
+
 def main():
 	parser = argparse.ArgumentParser()
 	parser.add_argument("--verbose", "-v", action="store_true", help="Enable verbose logging.")
+	parser.add_argument("--paperless-url", help="The URL to your paperless-ngx service")
+	parser.add_argument("--paperless-api-key", help="The API key to your paperless-ngx service")
+	parser.add_argument("--backlog", action="store_true", help="If present, process all files currently present as if they were just scanned")
+	parser.add_argument("watch_path", type=Path, help="The root directory to watch for new scans")
 	args = parser.parse_args()
 
 	logging.basicConfig(level=logging.DEBUG if args.verbose else logging.INFO)
+
+	if args.backlog:
+		for i, filename in enumerate(os.listdir(args.watch_path)):
+			path = args.watch_path / filename
+			_handle_write_complete(path, args.paperless_url, args.paperless_api_key, i)
+		return
 	try:
-		run(Path("/mnt/shares/scans"))
+		run(args.watch_path, args.paperless_url, args.paperless_api_key)
 	except KeyboardInterrupt:
 		LOGGER.info("Quitting")
 
-def run(path: Path) -> None:
+def run(watch_path: Path, paperless_url: str, paperless_api_key: str) -> None:
 	i = inotify.adapters.Inotify()
-	i.add_watch(str(path))
+	i.add_watch(str(watch_path))
+	count = 1
 	for event in i.event_gen(yield_nones=False):
 		(_, type_names, path, filename) = event
 		if "IN_CLOSE_WRITE" in type_names:
-			_handle_write_complete(Path(filename))
+			_handle_write_complete(watch_path / filename, paperless_url, paperless_api_key, count)
+			count += 1
 
-def _handle_write_complete(path: Path) -> None:
+def _handle_write_complete(path: Path, paperless_url: str, paperless_api_key: str, count: int) -> None:
 	if path.suffix == ".test":
 		LOGGER.info("Ignoring scanner test write '%s'", path)
 		return
-	LOGGER.info("Pretend I uploaded %s", path)
+	with open(path, "rb") as document:
+		auth = TokenAuth(paperless_api_key)
+		data = {
+			"title": f"{datetime.date.today().isoformat()}-scan-{count}",
+		}
+		files = {
+			#"archive_serial_number"
+			#"correspondent"
+			#"created": datetime.datetime.now(datetime.timezone.utc).isoformat(),
+			#"document_type"
+			"document": document,
+			#"storage_path"
+			#"tags"
+		}
+		url = paperless_url + "/api/documents/post_document/"
+		LOGGER.info("Uploading %s", data["title"])
+		#LOGGER.debug("Created time: %s", data["created"])
+		response = requests.post(url, data=data, files=files, auth=auth)
+		if not response.ok:
+			LOGGER.error("Failed to upload: %s %s", response.status_code, response.text)
+			breakpoint()
+			return
+		LOGGER.info("Upload response: %s", response.status_code)
+	os.remove(path)