Use cattrs to serialize conversations in archive

deletefb/deletefb.py

@@ -4,7 +4,7 @@ from .tools.config import settings
 from .tools.likes import unlike_pages
 from .tools.login import login
 from .tools.wall import delete_posts
-from .tools.conversations import delete_conversations
+from .tools.conversations import traverse_conversations
 from .tools.comments import delete_comments
 
 import argparse
@@ -119,7 +119,7 @@ def run_delete():
         delete_comments(driver, args.profile_url)
 
     elif args.mode == "conversations":
-        delete_conversations(driver, year=args.year)
+        traverse_conversations(driver, year=args.year)
 
     else:
         print("Please enter a valid mode")

deletefb/tools/archive.py

@@ -1,13 +1,21 @@
 from .config import settings
 from contextlib import contextmanager
 from pathlib import Path
+from datetime import datetime
 
 import attr
+import cattr
 import json
 
+TIME_FORMAT = "%Y-%m-%d %H:%M:%S"
+
 # Used to avoid duplicates in the log
 from pybloom_live import BloomFilter
 
+cattr.register_unstructure_hook(
+    datetime, lambda dt: datetime.strftime(dt, format=TIME_FORMAT)
+)
+
 def make_filter():
     return BloomFilter(
         capacity=settings["MAX_POSTS"],
@@ -30,7 +38,7 @@ class Archive:
         print("Archiving {0}".format(content))
 
         if content.name not in self._bloom_filter:
-            self.archive_file.write(json.dumps(attr.asdict(content)) + "\n")
+            self.archive_file.write(json.dumps(cattr.unstructure(content)) + "\n")
             self._bloom_filter.add(content.name)
         return
 

deletefb/tools/conversations.py

@@ -68,17 +68,9 @@ def get_conversations(driver):
 
     return conversations
 
-def get_convo_images(driver):
+def parse_conversation(driver):
     """
-    Gets all links to images in a messenger conversation
+    Extracts all messages in a conversation
-    Removes duplicates
-    """
-    for img in set(lxh.fromstring(driver.page_source).xpath("//img")):
-        yield img.get("src")
-
-def get_convo_messages(driver):
-    """
-    Gets all messages in a conversation
     """
 
     for msg in lxh.fromstring(driver.page_source).xpath("//div[@class='msg']/div"):
@@ -91,8 +83,10 @@ def get_convo_messages(driver):
                 date=data_store.get("timestamp")
               )
 
-def archive_conversation(driver, convo):
+def get_messages(driver, convo):
-    print(convo)
+    """
+    Get all of the messages for a given conversation
+    """
     driver.get(convo.url)
 
     wait = WebDriverWait(driver, 20)
@@ -119,12 +113,16 @@ def archive_conversation(driver, convo):
         except SELENIUM_EXCEPTIONS:
             continue
 
-    #for img in get_convo_images(driver):
+    return list(parse_conversation(driver))
-        #print(img)
 
-    convo.messages = list(get_convo_messages(driver))
+def delete_conversation(driver, convo):
+    """
+    Deletes a conversation
+    """
 
-def delete_conversations(driver, year=None):
+    return
+
+def traverse_conversations(driver, year=None):
     """
     Remove all conversations within a specified range
     """
@@ -133,15 +131,18 @@ def delete_conversations(driver, year=None):
 
     convos = get_conversations(driver)
 
-    for convo in convos:
+    with archiver("conversations") as archive_convo:
-        # If the year is set and there is a date
+        for convo in convos:
-        # Then we want to only look at convos from this year
+            # If the year is set and there is a date
+            # Then we want to only look at convos from this year
+
+            if year and convo.date:
+                if convo.date.year == int(year):
+                    convo.messages = get_messages(driver, convo)
+                    archive_convo.archive(convo)
 
-        if year and convo.date:
+            # Otherwise we're looking at all convos
-            if convo.date.year == int(year):
+            elif not year:
-                archive_conversation(driver, convo)
+                convo.messages = get_messages(driver, convo)
-                print(convo.messages)
+                archive_convo.archive(convo)
 
-        # Otherwise we're looking at all convos
-        elif not year:
-            archive_conversation(driver, convo)

deletefb/types.py

@@ -2,6 +2,8 @@ import attr
 import uuid
 import pendulum
 
+from datetime import datetime
+
 def convert_date(text):
     """
     Tries to parse a date into a DateTime instance
@@ -35,7 +37,7 @@ class Comment:
 class Conversation:
     url = attr.ib()
     name = attr.ib()
-    date = attr.ib(converter=convert_date)
+    date : datetime = attr.ib(converter=convert_date)
     messages = attr.ib(default=[])
 
 @attr.s
@@ -44,7 +46,7 @@ class Message:
     content = attr.ib()
 
     # Remove the last 3 digits from FB's dates. They are not standard.
-    date = attr.ib(converter=lambda t: pendulum.from_timestamp(int(str(t)[0:-3])))
+    date : datetime = attr.ib(converter=lambda t: pendulum.from_timestamp(int(str(t)[0:-3])))
 
 @attr.s
 class Page:

setup.py

@@ -25,6 +25,7 @@ setuptools.setup(
         "requests",
         "pybloom-live",
         "attrs",
+        "cattrs",
         "lxml",
         "pendulum"
     ],