From 5045ee68b6caf8b51121f6f85e4dfcae5e5a0ba0 Mon Sep 17 00:00:00 2001
From: Anton Khirnov <anton@khirnov.net>
Date: Thu, 13 May 2021 15:53:02 +0200
Subject: db: implement an email message cache

Not in master because it doesn't seem to be the bottleneck, didn't check
whether the difference is measurable.
---
 alot/db/manager.py | 31 +++++++++++++++++++++++++++++++
 alot/db/message.py |  3 +--
 2 files changed, 32 insertions(+), 2 deletions(-)

diff --git a/alot/db/manager.py b/alot/db/manager.py
index efd70e70..9d76ddeb 100644
--- a/alot/db/manager.py
+++ b/alot/db/manager.py
@@ -3,8 +3,10 @@
 # For further details see the COPYING file
 
 import asyncio
+from   collections import OrderedDict
 from   concurrent.futures import ThreadPoolExecutor
 from   contextlib import closing
+import email
 from   functools  import partial, partialmethod
 import logging
 import os
@@ -112,6 +114,31 @@ class _DBWriteList:
     def __str__(self):
         return '%s:%s' % (self.__class__.__name__, self._ops)
 
+class _EmailCache:
+    _cache = None
+
+    def __init__(self, max_size = 128):
+        self._cache    = OrderedDict()
+        self._max_size = max_size
+
+    def get(self, path):
+        with open(path, 'rb') as f:
+            mtime = os.stat(f.fileno()).st_mtime
+
+            mtime_cached, mail = self._cache.pop(path, (None, None))
+            if mtime_cached != mtime:
+                mail = email.message_from_bytes(f.read(), policy = email.policy.SMTP)
+                logging.info('loading from disk')
+            else:
+                logging.info('reading from cache')
+
+            # update the cache
+            if len(self._cache) >= self._max_size:
+                self._cache.popitem(last = False)
+            self._cache[path] = (mtime, mail)
+
+            return mail
+
 class DBManager:
     """
     Keeps track of your index parameters, maintains a write-queue and
@@ -130,6 +157,8 @@ class DBManager:
     _write_task    = None
     _write_queue   = None
 
+    _email_cache   = None
+
     def __init__(self, loop, path=None, ro=False):
         """
         :param path: absolute path to the notmuch index
@@ -148,6 +177,8 @@ class DBManager:
         self._exclude_tags = frozenset(settings.get('exclude_tags'))
         self._property_tags = frozenset(settings.get('property_tags'))
 
+        self._email_cache = _EmailCache()
+
     def _db_ro(self):
         return closing(Database(path = self.path, mode = Database.MODE.READ_ONLY))
 
diff --git a/alot/db/message.py b/alot/db/message.py
index 74139315..f7aea829 100644
--- a/alot/db/message.py
+++ b/alot/db/message.py
@@ -459,8 +459,7 @@ class Message:
     @cached_property
     def _email(self):
         try:
-            with open(self.filename, 'rb') as f:
-                mail = email.message_from_bytes(f.read(), policy = email.policy.SMTP)
+            mail = self._dbman._email_cache.get(self.filename)
         except IOError:
             warning = b"Subject: Caution!\n"\
                       b"Message file is no longer accessible:\n%s" % self.filename
-- 
cgit v1.2.3