update retention policy to a maximum of 2 days for completed archives and adjust related configurations

app/config.py

@@ -4,6 +4,8 @@ import os
 from dataclasses import dataclass
 from pathlib import Path
 
+MAX_ARCHIVE_RETENTION_DAYS = 2
+
 
 def _as_bool(name: str, default: bool) -> bool:
     value = os.getenv(name)
@@ -59,6 +61,14 @@ class Settings:
         data_dir = Path(os.getenv("SITEHARBOR_DATA_DIR", project_root / "data")).expanduser()
         if not data_dir.is_absolute():
             data_dir = project_root / data_dir
+        max_retention_days = min(
+            _as_int("SITEHARBOR_MAX_RETENTION_DAYS", MAX_ARCHIVE_RETENTION_DAYS),
+            MAX_ARCHIVE_RETENTION_DAYS,
+        )
+        default_retention_days = min(
+            _as_int("SITEHARBOR_DEFAULT_RETENTION_DAYS", MAX_ARCHIVE_RETENTION_DAYS),
+            max_retention_days,
+        )
 
         return cls(
             project_root=project_root,
@@ -77,8 +87,8 @@ class Settings:
             allow_private_networks=_as_bool("SITEHARBOR_ALLOW_PRIVATE_NETWORKS", False),
             allow_nonstandard_ports=_as_bool("SITEHARBOR_ALLOW_NONSTANDARD_PORTS", False),
             respect_robots=_as_bool("SITEHARBOR_RESPECT_ROBOTS", True),
-            default_retention_days=_as_int("SITEHARBOR_DEFAULT_RETENTION_DAYS", 7),
-            max_retention_days=_as_int("SITEHARBOR_MAX_RETENTION_DAYS", 30),
+            default_retention_days=default_retention_days,
+            max_retention_days=max_retention_days,
             max_pages_cap=_as_int("SITEHARBOR_MAX_PAGES", 1_000),
             max_depth_cap=_as_int("SITEHARBOR_MAX_DEPTH", 12),
             max_bytes_cap=_as_int("SITEHARBOR_MAX_BYTES", 1024 * 1024 * 1024),