From 20869a13b3490097ae80f83100cee1aab357fa45 Mon Sep 17 00:00:00 2001
From: dgtlmoon <dgtlmoon@gmail.com>
Date: Mon, 12 Sep 2022 12:45:29 +0200
Subject: [PATCH] tweaks to saving

---
 changedetectionio/fetch_processor/__init__.py |  1 -
 changedetectionio/fetch_processor/image.py    |  4 ++--
 changedetectionio/model/Watch.py              | 11 ++++++++++-
 changedetectionio/update_worker.py            |  2 +-
 requirements.txt                              |  1 +
 5 files changed, 14 insertions(+), 5 deletions(-)

diff --git a/changedetectionio/fetch_processor/__init__.py b/changedetectionio/fetch_processor/__init__.py
index 74ae2c87..2b9a7773 100644
--- a/changedetectionio/fetch_processor/__init__.py
+++ b/changedetectionio/fetch_processor/__init__.py
@@ -3,7 +3,6 @@ available_fetchers = [('json_html_plaintext', 'JSON/HTML/Text'), ('image', 'Stat
 class fetch_processor():
     contents = b''
     screenshot = None
-    history_artifact_suffix = 'txt'
 
     """
     base class for all fetch processors
diff --git a/changedetectionio/fetch_processor/image.py b/changedetectionio/fetch_processor/image.py
index ffe1589f..ed006601 100644
--- a/changedetectionio/fetch_processor/image.py
+++ b/changedetectionio/fetch_processor/image.py
@@ -80,6 +80,8 @@ class perform_site_check(fetch_processor):
 
         update_obj["last_check_status"] = fetcher.get_last_status_code()
 
+        self.contents = fetcher.raw_content
+
         image = Image.open(io.BytesIO(fetcher.raw_content))
 
         # @todo different choice?
@@ -97,6 +99,4 @@ class perform_site_check(fetch_processor):
         if not watch.get('previous_md5'):
             watch['previous_md5'] = fetched_hash
 
-        #self.contents = fetcher.screenshot
-
         return changed_detected, update_obj
diff --git a/changedetectionio/model/Watch.py b/changedetectionio/model/Watch.py
index d624bff0..bb227920 100644
--- a/changedetectionio/model/Watch.py
+++ b/changedetectionio/model/Watch.py
@@ -146,9 +146,18 @@ class model(dict):
         bump = self.history
         return self.__newest_history_key
 
-    def save_history_artifact(self, contents: bytes, timestamp, suffix='txt'):
+    def save_history_artifact(self, contents: bytes, timestamp):
         import uuid
         import logging
+        import magic
+        import re
+        suffix = 'bin'
+        # detect extension type
+        mtype = magic.from_buffer(contents, mime=True)
+        if mtype:
+            r = re.search(r'image/(\w+)', mtype, re.IGNORECASE)
+            if r:
+                suffix = r.group(1)
 
         output_path = "{}/{}".format(self.__datastore_path, self['uuid'])
 
diff --git a/changedetectionio/update_worker.py b/changedetectionio/update_worker.py
index 072de2f6..696be7d4 100644
--- a/changedetectionio/update_worker.py
+++ b/changedetectionio/update_worker.py
@@ -270,7 +270,7 @@ class update_worker(threading.Thread):
                             # For the FIRST time we check a site, or a change detected, save the snapshot.
                             if changed_detected or not watch['last_checked']:
                                 # A change was detected
-                                watch.save_history_artifact(contents=update_handler.contents, timestamp=str(round(time.time())), suffix=update_handler.history_artifact_suffix)
+                                watch.save_history_artifact(contents=update_handler.contents, timestamp=str(round(time.time())))
 
                             self.datastore.update_watch(uuid=uuid, update_obj=update_obj)
 
diff --git a/requirements.txt b/requirements.txt
index c03ae724..55b0947e 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -46,3 +46,4 @@ werkzeug ~= 2.0.0
 
 imagehash ~= 4.3.0
 pillow
+python-magic