# Authors: David Alexander, Jim Bullard
__all__ = [ "BasH5Reader" ,
"BaxH5Reader" ,
"BasH5Collection" ]
try:
import h5py
except ImportError:
from pbcore.util import h5py_dummy
h5py = h5py_dummy()
import numpy as np
import os.path as op
from bisect import bisect_left, bisect_right
from operator import getitem
from itertools import groupby
from collections import OrderedDict
from pbcore.io.FofnIO import readFofn
from pbcore.chemistry import (decodeTriple,
tripleFromMetadataXML,
ChemistryLookupError)
from pbcore.model import ExtraBaseRegionsMixin, HQ_REGION
from ._utils import arrayFromDataset, CommonEqualityMixin
# ZMW hole Types
SEQUENCING_ZMW = 0
from __future__ import absolute_import
from __future__ import print_function
try:
import h5py
except ImportError:
from pbcore.util import h5py_dummy
h5py = h5py_dummy()
import numpy as np
from cStringIO import StringIO
def arrayFromDataset(ds, offsetBegin, offsetEnd):
"""
Extract a one-dimensional array from an HDF5 dataset.
"""
shape = (offsetEnd - offsetBegin,)
a = np.ndarray(shape=shape, dtype=ds.dtype)
mspace = h5py.h5s.create_simple(shape)
fspace = ds.id.get_space()
fspace.select_hyperslab((offsetBegin,), shape, (1,))
ds.id.read(mspace, fspace, a)
return a
def splitFileContents(f, delimiter, BLOCKSIZE=8192):
"""
Same semantics as f.read().split(delimiter), but with memory usage
determined by largest chunk rather than entire file size
"""