From 8162e2e67b7b817d6b01a60df52bb67256fd2ba3 Mon Sep 17 00:00:00 2001
From: Thomas Waldmann <tw@waldmann-edv.de>
Date: Thu, 14 Jan 2021 20:41:57 +0100
Subject: [PATCH] cached_hash is only used in archive, move it there

---
 src/borg/archive.py  | 29 ++++++++++++++++++++++++++++-
 src/borg/chunker.pyx | 26 --------------------------
 2 files changed, 28 insertions(+), 27 deletions(-)

diff --git a/src/borg/archive.py b/src/borg/archive.py
index 90013bd29..b597a1aa6 100644
--- a/src/borg/archive.py
+++ b/src/borg/archive.py
@@ -19,7 +19,7 @@
 logger = create_logger()
 
 from . import xattr
-from .chunker import get_chunker, Chunk, cached_hash
+from .chunker import get_chunker, Chunk
 from .cache import ChunkListEntry
 from .crypto.key import key_factory
 from .compress import Compressor, CompressionSpec
@@ -41,6 +41,7 @@
 from .helpers import os_open, flags_normal, flags_dir
 from .helpers import msgpack
 from .helpers import sig_int
+from .lrucache import LRUCache
 from .patterns import PathPrefixPattern, FnmatchPattern, IECommand
 from .item import Item, ArchiveItem, ItemDiff
 from .platform import acl_get, acl_set, set_flags, get_flags, swidth, hostname
@@ -1088,6 +1089,32 @@ def stat_attrs(self, st, path, fd=None):
         return attrs
 
 
+# remember a few recently used all-zero chunk hashes in this mapping.
+# (hash_func, chunk_length) -> chunk_hash
+# we play safe and have the hash_func in the mapping key, in case we
+# have different hash_funcs within the same borg run.
+zero_chunk_ids = LRUCache(10, dispose=lambda _: None)
+
+
+def cached_hash(chunk, id_hash):
+    allocation = chunk.meta['allocation']
+    if allocation == CH_DATA:
+        data = chunk.data
+        chunk_id = id_hash(data)
+    elif allocation in (CH_HOLE, CH_ALLOC):
+        size = chunk.meta['size']
+        assert size <= len(zeros)
+        data = memoryview(zeros)[:size]
+        try:
+            chunk_id = zero_chunk_ids[(id_hash, size)]
+        except KeyError:
+            chunk_id = id_hash(data)
+            zero_chunk_ids[(id_hash, size)] = chunk_id
+    else:
+        raise ValueError('unexpected allocation type')
+    return chunk_id, data
+
+
 class ChunksProcessor:
     # Processes an iterator of chunks for an Item
 
diff --git a/src/borg/chunker.pyx b/src/borg/chunker.pyx
index 099532308..210ea461c 100644
--- a/src/borg/chunker.pyx
+++ b/src/borg/chunker.pyx
@@ -7,7 +7,6 @@ import os
 from collections import namedtuple
 
 from .constants import CH_DATA, CH_ALLOC, CH_HOLE, MAX_DATA_SIZE, zeros
-from .lrucache import LRUCache
 
 from libc.stdlib cimport free
 
@@ -53,31 +52,6 @@ def Chunk(data, **meta):
     return _Chunk(meta, data)
 
 
-# remember a few recently used all-zero chunk hashes in this mapping.
-# (hash_func, chunk_length) -> chunk_hash
-# we play safe and have the hash_func in the mapping key, in case we
-# have different hash_funcs within the same borg run.
-zero_chunk_ids = LRUCache(10, dispose=lambda _: None)
-
-def cached_hash(chunk, id_hash):
-    allocation = chunk.meta['allocation']
-    if allocation == CH_DATA:
-        data = chunk.data
-        chunk_id = id_hash(data)
-    elif allocation in (CH_HOLE, CH_ALLOC):
-        size = chunk.meta['size']
-        assert size <= len(zeros)
-        data = memoryview(zeros)[:size]
-        try:
-            chunk_id = zero_chunk_ids[(id_hash, size)]
-        except KeyError:
-            chunk_id = id_hash(data)
-            zero_chunk_ids[(id_hash, size)] = chunk_id
-    else:
-        raise ValueError('unexpected allocation type')
-    return chunk_id, data
-
-
 def dread(offset, size, fd=None, fh=-1):
     use_fh = fh >= 0
     if use_fh: