From ad49c3a3541fcd01b828a65180bf99b260f5f304 Mon Sep 17 00:00:00 2001
From: Morgan McMillian <morgan@mcmillian.dev>
Date: Fri, 1 Jul 2022 10:57:16 -0700
Subject: [PATCH] add config and filters to logging

Basic logging configuration can now be done in the config.yaml file and
filter has been added to redact the access tokens from the log output.
Closes issue #38
---
 config.yaml-sample | 22 ++++++++++++++++++++++
 pnut-matrix.py     | 29 +++++++++++++++++++++++------
 2 files changed, 45 insertions(+), 6 deletions(-)

diff --git a/config.yaml-sample b/config.yaml-sample
index 922e123..e97295a 100644
--- a/config.yaml-sample
+++ b/config.yaml-sample
@@ -13,3 +13,25 @@ MATRIX_PNUT_TOKEN: '<AUTH_TOKEN>'  # pnut.io auth token for the matrix bot
 PNUTCLIENT_ID: '<CLIENT_ID>'  # pnut.io app client ID
 PNUT_APPTOKEN: '<APP TOKEN>'  # pnut.io app token
 PNUT_APPKEY: '<APPSTREAM KEY>'  # pnut.io app stream key
+
+logging:
+  version: 1
+  formatters:
+    precise:
+      format: '%(asctime)s - %(name)s - %(lineno)d - %(levelname)s - %(message)s'
+    normal:
+      format: '%(name)s - %(levelname)s - %(message)s'
+  handlers:
+    console:
+      class: logging.StreamHandler
+      formatter: normal
+  loggers:
+    werkzeug:
+      level: DEBUG
+    appservice:
+      level: DEBUG
+    urllib3.connectionpool:
+      level: DEBUG
+  root:
+    level: DEBUG
+    handlers: [console]
diff --git a/pnut-matrix.py b/pnut-matrix.py
index 0792431..3212f82 100644
--- a/pnut-matrix.py
+++ b/pnut-matrix.py
@@ -2,6 +2,7 @@ import websocket
 import threading
 import time
 import logging
+import logging.config
 import yaml
 import json
 import pnutpy
@@ -9,6 +10,7 @@ import requests
 import magic
 import argparse
 import os
+import re
 
 from matrix_client.api import MatrixHttpApi
 from matrix_client.api import MatrixError, MatrixRequestError
@@ -22,6 +24,21 @@ logger = logging.getLogger()
 _shutdown = threading.Event()
 _reconnect = threading.Event()
 
+class MLogFilter(logging.Filter):
+
+    ACCESS_TOKEN_RE = re.compile(r"(\?.*access(_|%5[Ff])token=)[^&]*(\s.*)$")
+    ACCESS_TOKEN_RE2 = re.compile(r"(\?.*access(_|%5[Ff])token=)[^&]*(.*)$")
+
+    def filter(self, record):
+        if record.name == "werkzeug" and len(record.args) > 0:
+            redacted_uri = MLogFilter.ACCESS_TOKEN_RE.sub(r"\1<redacted>\3", record.args[0])
+            record.args = (redacted_uri, ) + record.args[1:]
+        elif record.name == "urllib3.connectionpool" and len(record.args) > 3:
+            redacted_uri = MLogFilter.ACCESS_TOKEN_RE2.sub(r"\1<redacted>\3", record.args[4])
+            record.args = record.args[:4] + (redacted_uri,) + record.args[5:]
+
+        return True
+
 def new_message(msg, meta):
     logger.debug("channel: " + msg.channel_id)
     logger.debug("username: " + msg.user.username)
@@ -368,17 +385,17 @@ if __name__ == '__main__':
     # )
     args = a_parser.parse_args()
 
-    if args.debug:
-        # websocket.enableTrace(True)
-        logging.basicConfig(level=logging.DEBUG)
-    else:
-        logging.basicConfig(level=logging.INFO)
-
     configyaml = os.environ.get("CONFIG_FILE")
 
     with open(configyaml, "rb") as config_file:
         config = yaml.load(config_file, Loader=yaml.SafeLoader)
 
+    # websocket.enableTrace(True)
+    logging.config.dictConfig(config['logging'])
+    redact_filter = MLogFilter()
+    logging.getLogger("werkzeug").addFilter(redact_filter)
+    logging.getLogger("urllib3.connectionpool").addFilter(redact_filter)
+
     ws_url = 'wss://stream.pnut.io/v1/app?access_token='
     ws_url += config['PNUT_APPTOKEN'] + '&key=' + config['PNUT_APPKEY']
     ws_url += '&include_raw=1'