From d126265fe4e0b8016004e898a58faad3a9127339 Mon Sep 17 00:00:00 2001
From: Lee Bousfield <ljbousfield@gmail.com>
Date: Thu, 28 Jul 2016 15:57:03 -0400
Subject: [PATCH] borg mount: cache partially read data chunks

Cherry-pick of bfb00df from master to 1.0-maint
---
 borg/archiver.py |  5 +++++
 borg/fuse.py     | 17 +++++++++++++++--
 2 files changed, 20 insertions(+), 2 deletions(-)

diff --git a/borg/archiver.py b/borg/archiver.py
index d91bdd3e0..76aa33ea1 100644
--- a/borg/archiver.py
+++ b/borg/archiver.py
@@ -1275,6 +1275,11 @@ def build_parser(self, args=None, prog=None):
         option is given the command will run in the background until the filesystem
         is ``umounted``.
 
+        The BORG_MOUNT_DATA_CACHE_ENTRIES environment variable is meant for advanced users
+        to tweak the performance. It sets the number of cached data chunks; additional
+        memory usage can be up to ~8 MiB times this number. The default is the number
+        of CPU cores.
+
         For mount options, see the fuse(8) manual page. Additional mount options
         supported by borg:
 
diff --git a/borg/fuse.py b/borg/fuse.py
index 7b75b8b8e..09c05d6f8 100644
--- a/borg/fuse.py
+++ b/borg/fuse.py
@@ -13,6 +13,7 @@
 from .archive import Archive
 from .helpers import daemonize, bigint_to_int
 from .logger import create_logger
+from .lrucache import LRUCache
 logger = create_logger()
 
 
@@ -62,6 +63,9 @@ def __init__(self, key, repository, manifest, archive, cached_repo):
         self.pending_archives = {}
         self.accounted_chunks = {}
         self.cache = ItemCache()
+        data_cache_capacity = int(os.environ.get('BORG_MOUNT_DATA_CACHE_ENTRIES', os.cpu_count() or 1))
+        logger.debug('mount data cache capacity: %d chunks', data_cache_capacity)
+        self.data_cache = LRUCache(capacity=data_cache_capacity, dispose=lambda _: None)
         if archive:
             self.process_archive(archive)
         else:
@@ -282,8 +286,17 @@ def read(self, fh, offset, size):
                 offset -= s
                 continue
             n = min(size, s - offset)
-            chunk = self.key.decrypt(id, self.repository.get(id))
-            parts.append(chunk[offset:offset + n])
+            if id in self.data_cache:
+                data = self.data_cache[id]
+                if offset + n == len(data):
+                    # evict fully read chunk from cache
+                    del self.data_cache[id]
+            else:
+                data = self.key.decrypt(id, self.repository.get(id))
+                if offset + n < len(data):
+                    # chunk was only partially read, cache it
+                    self.data_cache[id] = data
+            parts.append(data[offset:offset + n])
             offset = 0
             size -= n
             if not size: