From 8cc1691857d2f65c1fe0351f098aa1dba39e22c9 Mon Sep 17 00:00:00 2001
From: Andrew Godwin <andrew@aeracode.org>
Date: Sun, 1 Oct 2023 10:17:00 -0600
Subject: [PATCH] Delete remote posts after a set horizon time

---
 activities/models/post.py | 25 +++++++++++++++++++++++--
 takahe/settings.py        |  5 +++++
 2 files changed, 28 insertions(+), 2 deletions(-)

diff --git a/activities/models/post.py b/activities/models/post.py
index 2cb3be7..342e4b5 100644
--- a/activities/models/post.py
+++ b/activities/models/post.py
@@ -8,6 +8,7 @@ from urllib.parse import urlparse
 
 import httpx
 import urlman
+from django.conf import settings
 from django.contrib.postgres.indexes import GinIndex
 from django.contrib.postgres.search import SearchVector
 from django.db import models, transaction
@@ -47,14 +48,15 @@ from users.models.system_actor import SystemActor
 
 class PostStates(StateGraph):
     new = State(try_interval=300)
-    fanned_out = State(externally_progressed=True)
+    fanned_out = State(try_interval=86400 * 14)
     deleted = State(try_interval=300)
-    deleted_fanned_out = State(delete_after=24 * 60 * 60)
+    deleted_fanned_out = State(delete_after=86400)
 
     edited = State(try_interval=300)
     edited_fanned_out = State(externally_progressed=True)
 
     new.transitions_to(fanned_out)
+    fanned_out.transitions_to(deleted_fanned_out)
     fanned_out.transitions_to(deleted)
     fanned_out.transitions_to(edited)
 
@@ -87,6 +89,25 @@ class PostStates(StateGraph):
         instance.ensure_hashtags()
         return cls.fanned_out
 
+    @classmethod
+    def handle_fanned_out(cls, instance: "Post"):
+        """
+        For remote posts, sees if we can delete them every so often.
+        """
+        # To be a candidate for deletion, a post must be remote and old enough
+        if instance.local:
+            return
+        if instance.created > timezone.now() - datetime.timedelta(
+            days=settings.SETUP.REMOTE_PRUNE_HORIZON
+        ):
+            return
+        # It must have no local interactions
+        if instance.interactions.filter(identity__local=True).exists():
+            return
+        # OK, delete it!
+        instance.delete()
+        return cls.deleted_fanned_out
+
     @classmethod
     def handle_deleted(cls, instance: "Post"):
         """
diff --git a/takahe/settings.py b/takahe/settings.py
index bd682cd..a5dd4dd 100644
--- a/takahe/settings.py
+++ b/takahe/settings.py
@@ -143,6 +143,11 @@ class Settings(BaseSettings):
     #: Default cache backend
     CACHES_DEFAULT: CacheBackendUrl | None = None
 
+    # How long to wait, in days, until remote posts/profiles are pruned from
+    # our database if nobody local has interacted with them. Must be in rough
+    # multiples of two weeks.
+    REMOTE_PRUNE_HORIZON: int = 60
+
     # Stator tuning
     STATOR_CONCURRENCY: int = 50
     STATOR_CONCURRENCY_PER_MODEL: int = 15