def __loadIndex(self):
"""load complete index into memory."""
assert self.mCreateMode == False, "asked to read from database opened for writing"
if self.mMethod == "uncompressed":
self.mDatabaseFile = open(self.mDbname, "r")
elif self.mMethod == "dictzip":
import dictzip
self.mDatabaseFile = dictzip.GzipFile(self.mNameDb)
elif self.mMethod == "lzo":
import lzo
self.mDatabaseFile = Uncompressor(self.mNameDb, lzo.decompress)
elif self.mMethod == "gzip":
self.mDatabaseFile = Uncompressor(self.mNameDb, gzip_demangler)
elif self.mMethod == "zlib":
self.mDatabaseFile = Uncompressor(self.mNameDb, zlib.decompress)
elif eslf.mMethod == "bz2":
self.mDatabaseFile = bz2.BZ2File(self.mNameDb)
elif self.mMethod == "debug":
self.mDatabaseFile = Uncompressor(self.mDbname + ".debug",
lambda x: x)
self.mIndex = {}
for line in open(self.mNameIndex, "r"):
if line.startswith("#"): continue
data = line[:-1].split("\t")
# index with random access points
if len(data) > 4:
(identifier, pos_id, block_size, lsequence) = bytes(
data[0]), int(data[1]), int(data[2]), int(data[-1])
points = map(int, data[3:-1])
self.mIndex[int(identifier)] = (pos_id, block_size, lsequence,
points)
else:
(identifier, pos_id, pos_seq, lsequence) = bytes(data[0]), int(
data[1]), int(data[2]), int(data[-1])
self.mIndex[int(identifier)] = (pos_id, pos_seq, lsequence)
self.mIsLoaded = True
def _loadIndex(self, compress=False):
"""load complete index into memory.
if compress is set to true, the index will not be loaded,
but a compressed index will be created instead.
"""
if self.mMethod == "uncompressed":
self.mDatabaseFile = open(self.mDbname, "r")
elif self.mMethod == "dictzip":
import dictzip
self.mDatabaseFile = dictzip.GzipFile(self.mDbname)
elif self.mMethod == "lzo":
import lzo
self.mDatabaseFile = Uncompressor(self.mDbname, lzo.decompress)
elif self.mMethod == "gzip":
self.mDatabaseFile = Uncompressor(self.mDbname, gzip_demangler)
elif self.mMethod == "zlib":
self.mDatabaseFile = Uncompressor(self.mDbname, zlib.decompress)
elif self.mMethod == "bzip2":
import bz2
self.mDatabaseFile = Uncompressor(self.mDbname, bz2.decompress)
elif self.mMethod == "debug":
self.mDatabaseFile = Uncompressor(self.mDbname + ".debug",
lambda x: x)
filename_index = self.mNameIndex + ".dbm"
if compress:
#if os.path.exists(filename_index):
# raise OSError("file %s already exists" % filename_index)
self.mIndex = anydbm.open(filename_index, "n")
elif os.path.exists(filename_index):
self.mIndex = anydbm.open(filename_index, "r")
self.mIsLoaded = True
return
else:
self.mIndex = {}
for line in open(self.mNameIndex, "r"):
data = line[:-1].split("\t")
if len(data) == 2:
# ignore synonyms of non-existent contigs
identifier = data[1]
if data[0] not in self.mIndex:
continue
self.mSynonyms[identifier] = data[0]
else:
# index with random access points
if len(data) > 4:
(identifier, pos_id, block_size, lsequence) = data[0], int(
data[1]), int(data[2]), int(data[-1])
points = map(int, data[3:-1])
self.mIndex[identifier] = (pos_id, block_size, lsequence,
points)
else:
(identifier, pos_id, pos_seq, lsequence) = data[0], int(
data[1]), int(data[2]), int(data[-1])
self.mIndex[identifier] = struct.pack(
"QQi", pos_id, pos_seq, lsequence)
self._addSynonyms()
self.mIsLoaded = True