|
@@ -1,3 +1,4 @@
|
|
|
+from itertools import ifilter
|
|
|
import logging
|
|
|
import msgpack
|
|
|
import os
|
|
@@ -14,7 +15,6 @@ class Cache(object):
|
|
|
self.path = os.path.join(os.path.expanduser('~'), '.dedupestore', 'cache',
|
|
|
'%s.cache' % self.store.uuid)
|
|
|
self.tid = -1
|
|
|
- self.file_chunks = {}
|
|
|
self.open()
|
|
|
if self.tid != self.store.tid:
|
|
|
self.init(crypto)
|
|
@@ -24,21 +24,17 @@ class Cache(object):
|
|
|
return
|
|
|
cache = msgpack.unpackb(open(self.path, 'rb').read())
|
|
|
assert cache['version'] == 1
|
|
|
- if cache['store'] != self.store.uuid:
|
|
|
- raise Exception('Cache UUID mismatch')
|
|
|
- self.chunkmap = cache['chunkmap']
|
|
|
+ self.chunk_counts = cache['chunk_counts']
|
|
|
# Discard old file_chunks entries
|
|
|
- for hash, entry in cache['file_chunks'].iteritems():
|
|
|
- count = entry[0]
|
|
|
- if count < 8:
|
|
|
- self.file_chunks[hash] = [count + 1] + list(entry[1:])
|
|
|
+ self.file_chunks = cache['file_chunks']
|
|
|
self.tid = cache['tid']
|
|
|
|
|
|
def init(self, crypto):
|
|
|
"""Initializes cache by fetching and reading all archive indicies
|
|
|
"""
|
|
|
logging.info('Initializing cache...')
|
|
|
- self.chunkmap = {}
|
|
|
+ self.chunk_counts = {}
|
|
|
+ self.file_chunks = {}
|
|
|
self.tid = self.store.tid
|
|
|
if self.store.tid == 0:
|
|
|
return
|
|
@@ -47,19 +43,19 @@ class Cache(object):
|
|
|
cindex = msgpack.unpackb(data)
|
|
|
for id, size in cindex['chunks']:
|
|
|
try:
|
|
|
- count, size = self.chunkmap[id]
|
|
|
- self.chunkmap[id] = count + 1, size
|
|
|
+ count, size = self.chunk_counts[id]
|
|
|
+ self.chunk_counts[id] = count + 1, size
|
|
|
except KeyError:
|
|
|
- self.chunkmap[id] = 1, size
|
|
|
+ self.chunk_counts[id] = 1, size
|
|
|
self.save()
|
|
|
|
|
|
def save(self):
|
|
|
assert self.store.state == self.store.OPEN
|
|
|
cache = {'version': 1,
|
|
|
- 'store': self.store.uuid,
|
|
|
- 'chunkmap': self.chunkmap,
|
|
|
'tid': self.store.tid,
|
|
|
- 'file_chunks': self.file_chunks,
|
|
|
+ 'chunk_counts': self.chunk_counts,
|
|
|
+ 'file_chunks': dict(ifilter(lambda i: i[1][0] < 8,
|
|
|
+ self.file_chunks.iteritems())),
|
|
|
}
|
|
|
data = msgpack.packb(cache)
|
|
|
cachedir = os.path.dirname(self.path)
|
|
@@ -74,31 +70,31 @@ class Cache(object):
|
|
|
data, hash = crypto.encrypt_read(data)
|
|
|
csize = len(data)
|
|
|
self.store.put(NS_CHUNK, id, data)
|
|
|
- self.chunkmap[id] = (1, csize)
|
|
|
+ self.chunk_counts[id] = (1, csize)
|
|
|
return csize
|
|
|
|
|
|
def seen_chunk(self, id):
|
|
|
- count, size = self.chunkmap.get(id, (0, 0))
|
|
|
- return count
|
|
|
+ return self.chunk_counts.get(id, (0, 0))[0]
|
|
|
|
|
|
def chunk_incref(self, id):
|
|
|
- count, size = self.chunkmap[id]
|
|
|
- self.chunkmap[id] = (count + 1, size)
|
|
|
+ count, size = self.chunk_counts[id]
|
|
|
+ self.chunk_counts[id] = (count + 1, size)
|
|
|
return size
|
|
|
|
|
|
def chunk_decref(self, id):
|
|
|
- count, size = self.chunkmap[id]
|
|
|
+ count, size = self.chunk_counts[id]
|
|
|
if count == 1:
|
|
|
- del self.chunkmap[id]
|
|
|
+ del self.chunk_counts[id]
|
|
|
self.store.delete(NS_CHUNK, id)
|
|
|
else:
|
|
|
- self.chunkmap[id] = (count - 1, size)
|
|
|
+ self.chunk_counts[id] = (count - 1, size)
|
|
|
|
|
|
def file_known_and_unchanged(self, path_hash, st):
|
|
|
entry = self.file_chunks.get(path_hash)
|
|
|
- if (entry and entry[1] == st.st_ino
|
|
|
- and entry[2] == st.st_size and entry[3] == st.st_mtime):
|
|
|
- entry[0] = 0 # reset entry age
|
|
|
+ if (entry and entry[3] == st.st_mtime
|
|
|
+ and entry[2] == st.st_size and entry[1] == st.st_ino):
|
|
|
+ # reset entry age
|
|
|
+ self.file_chunks[path_hash] = (0,) + entry[1:]
|
|
|
return entry[4], entry[2]
|
|
|
else:
|
|
|
return None, 0
|