Kano001's picture
Upload 376 files
122d3ff verified
raw
history blame
31.2 kB
# This file is part of h5py, a Python interface to the HDF5 library.
#
# http://www.h5py.org
#
# Copyright 2008-2013 Andrew Collette and contributors
#
# License: Standard 3-clause BSD; see "license.txt" for full license terms
# and contributor agreement.
"""
Implements support for high-level access to HDF5 groups.
"""
from contextlib import contextmanager
import posixpath as pp
import numpy
from .compat import filename_decode, filename_encode
from .. import h5, h5g, h5i, h5o, h5r, h5t, h5l, h5p
from . import base
from .base import HLObject, MutableMappingHDF5, phil, with_phil
from . import dataset
from . import datatype
from .vds import vds_support
class Group(HLObject, MutableMappingHDF5):
""" Represents an HDF5 group.
"""
def __init__(self, bind):
""" Create a new Group object by binding to a low-level GroupID.
"""
with phil:
if not isinstance(bind, h5g.GroupID):
raise ValueError("%s is not a GroupID" % bind)
super().__init__(bind)
_gcpl_crt_order = h5p.create(h5p.GROUP_CREATE)
_gcpl_crt_order.set_link_creation_order(
h5p.CRT_ORDER_TRACKED | h5p.CRT_ORDER_INDEXED)
_gcpl_crt_order.set_attr_creation_order(
h5p.CRT_ORDER_TRACKED | h5p.CRT_ORDER_INDEXED)
def create_group(self, name, track_order=None):
""" Create and return a new subgroup.
Name may be absolute or relative. Fails if the target name already
exists.
track_order
Track dataset/group/attribute creation order under this group
if True. If None use global default h5.get_config().track_order.
"""
if track_order is None:
track_order = h5.get_config().track_order
with phil:
name, lcpl = self._e(name, lcpl=True)
gcpl = Group._gcpl_crt_order if track_order else None
gid = h5g.create(self.id, name, lcpl=lcpl, gcpl=gcpl)
return Group(gid)
def create_dataset(self, name, shape=None, dtype=None, data=None, **kwds):
""" Create a new HDF5 dataset
name
Name of the dataset (absolute or relative). Provide None to make
an anonymous dataset.
shape
Dataset shape. Use "()" for scalar datasets. Required if "data"
isn't provided.
dtype
Numpy dtype or string. If omitted, dtype('f') will be used.
Required if "data" isn't provided; otherwise, overrides data
array's dtype.
data
Provide data to initialize the dataset. If used, you can omit
shape and dtype arguments.
Keyword-only arguments:
chunks
(Tuple or int) Chunk shape, or True to enable auto-chunking. Integers can
be used for 1D shape.
maxshape
(Tuple or int) Make the dataset resizable up to this shape. Use None for
axes you want to be unlimited. Integers can be used for 1D shape.
compression
(String or int) Compression strategy. Legal values are 'gzip',
'szip', 'lzf'. If an integer in range(10), this indicates gzip
compression level. Otherwise, an integer indicates the number of a
dynamically loaded compression filter.
compression_opts
Compression settings. This is an integer for gzip, 2-tuple for
szip, etc. If specifying a dynamically loaded compression filter
number, this must be a tuple of values.
scaleoffset
(Integer) Enable scale/offset filter for (usually) lossy
compression of integer or floating-point data. For integer
data, the value of scaleoffset is the number of bits to
retain (pass 0 to let HDF5 determine the minimum number of
bits necessary for lossless compression). For floating point
data, scaleoffset is the number of digits after the decimal
place to retain; stored values thus have absolute error
less than 0.5*10**(-scaleoffset).
shuffle
(T/F) Enable shuffle filter.
fletcher32
(T/F) Enable fletcher32 error detection. Not permitted in
conjunction with the scale/offset filter.
fillvalue
(Scalar) Use this value for uninitialized parts of the dataset.
track_times
(T/F) Enable dataset creation timestamps.
track_order
(T/F) Track attribute creation order if True. If omitted use
global default h5.get_config().track_order.
external
(Iterable of tuples) Sets the external storage property, thus
designating that the dataset will be stored in one or more
non-HDF5 files external to the HDF5 file. Adds each tuple
of (name, offset, size) to the dataset's list of external files.
Each name must be a str, bytes, or os.PathLike; each offset and
size, an integer. If only a name is given instead of an iterable
of tuples, it is equivalent to [(name, 0, h5py.h5f.UNLIMITED)].
efile_prefix
(String) External dataset file prefix for dataset access property
list. Does not persist in the file.
virtual_prefix
(String) Virtual dataset file prefix for dataset access property
list. Does not persist in the file.
allow_unknown_filter
(T/F) Do not check that the requested filter is available for use.
This should only be used with ``write_direct_chunk``, where the caller
compresses the data before handing it to h5py.
rdcc_nbytes
Total size of the dataset's chunk cache in bytes. The default size
is 1024**2 (1 MiB).
rdcc_w0
The chunk preemption policy for this dataset. This must be
between 0 and 1 inclusive and indicates the weighting according to
which chunks which have been fully read or written are penalized
when determining which chunks to flush from cache. A value of 0
means fully read or written chunks are treated no differently than
other chunks (the preemption is strictly LRU) while a value of 1
means fully read or written chunks are always preempted before
other chunks. If your application only reads or writes data once,
this can be safely set to 1. Otherwise, this should be set lower
depending on how often you re-read or re-write the same data. The
default value is 0.75.
rdcc_nslots
The number of chunk slots in the dataset's chunk cache. Increasing
this value reduces the number of cache collisions, but slightly
increases the memory used. Due to the hashing strategy, this value
should ideally be a prime number. As a rule of thumb, this value
should be at least 10 times the number of chunks that can fit in
rdcc_nbytes bytes. For maximum performance, this value should be set
approximately 100 times that number of chunks. The default value is
521.
"""
if 'track_order' not in kwds:
kwds['track_order'] = h5.get_config().track_order
if 'efile_prefix' in kwds:
kwds['efile_prefix'] = self._e(kwds['efile_prefix'])
if 'virtual_prefix' in kwds:
kwds['virtual_prefix'] = self._e(kwds['virtual_prefix'])
with phil:
group = self
if name:
name = self._e(name)
if b'/' in name.lstrip(b'/'):
parent_path, name = name.rsplit(b'/', 1)
group = self.require_group(parent_path)
dsid = dataset.make_new_dset(group, shape, dtype, data, name, **kwds)
dset = dataset.Dataset(dsid)
return dset
if vds_support:
def create_virtual_dataset(self, name, layout, fillvalue=None):
"""Create a new virtual dataset in this group.
See virtual datasets in the docs for more information.
name
(str) Name of the new dataset
layout
(VirtualLayout) Defines the sources for the virtual dataset
fillvalue
The value to use where there is no data.
"""
with phil:
group = self
if name:
name = self._e(name)
if b'/' in name.lstrip(b'/'):
parent_path, name = name.rsplit(b'/', 1)
group = self.require_group(parent_path)
dsid = layout.make_dataset(
group, name=name, fillvalue=fillvalue,
)
dset = dataset.Dataset(dsid)
return dset
@contextmanager
def build_virtual_dataset(
self, name, shape, dtype, maxshape=None, fillvalue=None
):
"""Assemble a virtual dataset in this group.
This is used as a context manager::
with f.build_virtual_dataset('virt', (10, 1000), np.uint32) as layout:
layout[0] = h5py.VirtualSource('foo.h5', 'data', (1000,))
name
(str) Name of the new dataset
shape
(tuple) Shape of the dataset
dtype
A numpy dtype for data read from the virtual dataset
maxshape
(tuple, optional) Maximum dimensions if the dataset can grow.
Use None for unlimited dimensions.
fillvalue
The value used where no data is available.
"""
from .vds import VirtualLayout
layout = VirtualLayout(shape, dtype, maxshape, self.file.filename)
yield layout
self.create_virtual_dataset(name, layout, fillvalue)
def require_dataset(self, name, shape, dtype, exact=False, **kwds):
""" Open a dataset, creating it if it doesn't exist.
If keyword "exact" is False (default), an existing dataset must have
the same shape and a conversion-compatible dtype to be returned. If
True, the shape and dtype must match exactly.
If keyword "maxshape" is given, the maxshape and dtype must match
instead.
If any of the keywords "rdcc_nslots", "rdcc_nbytes", or "rdcc_w0" are
given, they will be used to configure the dataset's chunk cache.
Other dataset keywords (see create_dataset) may be provided, but are
only used if a new dataset is to be created.
Raises TypeError if an incompatible object already exists, or if the
shape, maxshape or dtype don't match according to the above rules.
"""
if 'efile_prefix' in kwds:
kwds['efile_prefix'] = self._e(kwds['efile_prefix'])
if 'virtual_prefix' in kwds:
kwds['virtual_prefix'] = self._e(kwds['virtual_prefix'])
with phil:
if name not in self:
return self.create_dataset(name, *(shape, dtype), **kwds)
if isinstance(shape, int):
shape = (shape,)
try:
dsid = dataset.open_dset(self, self._e(name), **kwds)
dset = dataset.Dataset(dsid)
except KeyError:
dset = self[name]
raise TypeError("Incompatible object (%s) already exists" % dset.__class__.__name__)
if shape != dset.shape:
if "maxshape" not in kwds:
raise TypeError("Shapes do not match (existing %s vs new %s)" % (dset.shape, shape))
elif kwds["maxshape"] != dset.maxshape:
raise TypeError("Max shapes do not match (existing %s vs new %s)" % (dset.maxshape, kwds["maxshape"]))
if exact:
if dtype != dset.dtype:
raise TypeError("Datatypes do not exactly match (existing %s vs new %s)" % (dset.dtype, dtype))
elif not numpy.can_cast(dtype, dset.dtype):
raise TypeError("Datatypes cannot be safely cast (existing %s vs new %s)" % (dset.dtype, dtype))
return dset
def create_dataset_like(self, name, other, **kwupdate):
""" Create a dataset similar to `other`.
name
Name of the dataset (absolute or relative). Provide None to make
an anonymous dataset.
other
The dataset which the new dataset should mimic. All properties, such
as shape, dtype, chunking, ... will be taken from it, but no data
or attributes are being copied.
Any dataset keywords (see create_dataset) may be provided, including
shape and dtype, in which case the provided values take precedence over
those from `other`.
"""
for k in ('shape', 'dtype', 'chunks', 'compression',
'compression_opts', 'scaleoffset', 'shuffle', 'fletcher32',
'fillvalue'):
kwupdate.setdefault(k, getattr(other, k))
# TODO: more elegant way to pass these (dcpl to create_dataset?)
dcpl = other.id.get_create_plist()
kwupdate.setdefault('track_times', dcpl.get_obj_track_times())
kwupdate.setdefault('track_order', dcpl.get_attr_creation_order() > 0)
# Special case: the maxshape property always exists, but if we pass it
# to create_dataset, the new dataset will automatically get chunked
# layout. So we copy it only if it is different from shape.
if other.maxshape != other.shape:
kwupdate.setdefault('maxshape', other.maxshape)
return self.create_dataset(name, **kwupdate)
def require_group(self, name):
# TODO: support kwargs like require_dataset
"""Return a group, creating it if it doesn't exist.
TypeError is raised if something with that name already exists that
isn't a group.
"""
with phil:
if name not in self:
return self.create_group(name)
grp = self[name]
if not isinstance(grp, Group):
raise TypeError("Incompatible object (%s) already exists" % grp.__class__.__name__)
return grp
@with_phil
def __getitem__(self, name):
""" Open an object in the file """
if isinstance(name, h5r.Reference):
oid = h5r.dereference(name, self.id)
if oid is None:
raise ValueError("Invalid HDF5 object reference")
elif isinstance(name, (bytes, str)):
oid = h5o.open(self.id, self._e(name), lapl=self._lapl)
else:
raise TypeError("Accessing a group is done with bytes or str, "
"not {}".format(type(name)))
otype = h5i.get_type(oid)
if otype == h5i.GROUP:
return Group(oid)
elif otype == h5i.DATASET:
return dataset.Dataset(oid, readonly=(self.file.mode == 'r'))
elif otype == h5i.DATATYPE:
return datatype.Datatype(oid)
else:
raise TypeError("Unknown object type")
def get(self, name, default=None, getclass=False, getlink=False):
""" Retrieve an item or other information.
"name" given only:
Return the item, or "default" if it doesn't exist
"getclass" is True:
Return the class of object (Group, Dataset, etc.), or "default"
if nothing with that name exists
"getlink" is True:
Return HardLink, SoftLink or ExternalLink instances. Return
"default" if nothing with that name exists.
"getlink" and "getclass" are True:
Return HardLink, SoftLink and ExternalLink classes. Return
"default" if nothing with that name exists.
Example:
>>> cls = group.get('foo', getclass=True)
>>> if cls == SoftLink:
"""
# pylint: disable=arguments-differ
with phil:
if not (getclass or getlink):
try:
return self[name]
except KeyError:
return default
if name not in self:
return default
elif getclass and not getlink:
typecode = h5o.get_info(self.id, self._e(name), lapl=self._lapl).type
try:
return {h5o.TYPE_GROUP: Group,
h5o.TYPE_DATASET: dataset.Dataset,
h5o.TYPE_NAMED_DATATYPE: datatype.Datatype}[typecode]
except KeyError:
raise TypeError("Unknown object type")
elif getlink:
typecode = self.id.links.get_info(self._e(name), lapl=self._lapl).type
if typecode == h5l.TYPE_SOFT:
if getclass:
return SoftLink
linkbytes = self.id.links.get_val(self._e(name), lapl=self._lapl)
return SoftLink(self._d(linkbytes))
elif typecode == h5l.TYPE_EXTERNAL:
if getclass:
return ExternalLink
filebytes, linkbytes = self.id.links.get_val(self._e(name), lapl=self._lapl)
return ExternalLink(
filename_decode(filebytes), self._d(linkbytes)
)
elif typecode == h5l.TYPE_HARD:
return HardLink if getclass else HardLink()
else:
raise TypeError("Unknown link type")
def __setitem__(self, name, obj):
""" Add an object to the group. The name must not already be in use.
The action taken depends on the type of object assigned:
Named HDF5 object (Dataset, Group, Datatype)
A hard link is created at "name" which points to the
given object.
SoftLink or ExternalLink
Create the corresponding link.
Numpy ndarray
The array is converted to a dataset object, with default
settings (contiguous storage, etc.).
Numpy dtype
Commit a copy of the datatype as a named datatype in the file.
Anything else
Attempt to convert it to an ndarray and store it. Scalar
values are stored as scalar datasets. Raise ValueError if we
can't understand the resulting array dtype.
"""
with phil:
name, lcpl = self._e(name, lcpl=True)
if isinstance(obj, HLObject):
h5o.link(obj.id, self.id, name, lcpl=lcpl, lapl=self._lapl)
elif isinstance(obj, SoftLink):
self.id.links.create_soft(name, self._e(obj.path), lcpl=lcpl, lapl=self._lapl)
elif isinstance(obj, ExternalLink):
fn = filename_encode(obj.filename)
self.id.links.create_external(name, fn, self._e(obj.path),
lcpl=lcpl, lapl=self._lapl)
elif isinstance(obj, numpy.dtype):
htype = h5t.py_create(obj, logical=True)
htype.commit(self.id, name, lcpl=lcpl)
else:
ds = self.create_dataset(None, data=obj)
h5o.link(ds.id, self.id, name, lcpl=lcpl)
@with_phil
def __delitem__(self, name):
""" Delete (unlink) an item from this group. """
self.id.unlink(self._e(name))
@with_phil
def __len__(self):
""" Number of members attached to this group """
return self.id.get_num_objs()
@with_phil
def __iter__(self):
""" Iterate over member names """
for x in self.id.__iter__():
yield self._d(x)
@with_phil
def __reversed__(self):
""" Iterate over member names in reverse order. """
for x in self.id.__reversed__():
yield self._d(x)
@with_phil
def __contains__(self, name):
""" Test if a member name exists """
if hasattr(h5g, "_path_valid"):
if not self.id:
return False
return h5g._path_valid(self.id, self._e(name), self._lapl)
return self._e(name) in self.id
def copy(self, source, dest, name=None,
shallow=False, expand_soft=False, expand_external=False,
expand_refs=False, without_attrs=False):
"""Copy an object or group.
The source can be a path, Group, Dataset, or Datatype object. The
destination can be either a path or a Group object. The source and
destinations need not be in the same file.
If the source is a Group object, all objects contained in that group
will be copied recursively.
When the destination is a Group object, by default the target will
be created in that group with its current name (basename of obj.name).
You can override that by setting "name" to a string.
There are various options which all default to "False":
- shallow: copy only immediate members of a group.
- expand_soft: expand soft links into new objects.
- expand_external: expand external links into new objects.
- expand_refs: copy objects that are pointed to by references.
- without_attrs: copy object without copying attributes.
Example:
>>> f = File('myfile.hdf5', 'w')
>>> f.create_group("MyGroup")
>>> list(f.keys())
['MyGroup']
>>> f.copy('MyGroup', 'MyCopy')
>>> list(f.keys())
['MyGroup', 'MyCopy']
"""
with phil:
if isinstance(source, HLObject):
source_path = '.'
else:
# Interpret source as a path relative to this group
source_path = source
source = self
if isinstance(dest, Group):
if name is not None:
dest_path = name
elif source_path == '.':
dest_path = pp.basename(h5i.get_name(source.id))
else:
# copy source into dest group: dest_name/source_name
dest_path = pp.basename(h5i.get_name(source[source_path].id))
elif isinstance(dest, HLObject):
raise TypeError("Destination must be path or Group object")
else:
# Interpret destination as a path relative to this group
dest_path = dest
dest = self
flags = 0
if shallow:
flags |= h5o.COPY_SHALLOW_HIERARCHY_FLAG
if expand_soft:
flags |= h5o.COPY_EXPAND_SOFT_LINK_FLAG
if expand_external:
flags |= h5o.COPY_EXPAND_EXT_LINK_FLAG
if expand_refs:
flags |= h5o.COPY_EXPAND_REFERENCE_FLAG
if without_attrs:
flags |= h5o.COPY_WITHOUT_ATTR_FLAG
if flags:
copypl = h5p.create(h5p.OBJECT_COPY)
copypl.set_copy_object(flags)
else:
copypl = None
h5o.copy(source.id, self._e(source_path), dest.id, self._e(dest_path),
copypl, base.dlcpl)
def move(self, source, dest):
""" Move a link to a new location in the file.
If "source" is a hard link, this effectively renames the object. If
"source" is a soft or external link, the link itself is moved, with its
value unmodified.
"""
with phil:
if source == dest:
return
self.id.links.move(self._e(source), self.id, self._e(dest),
lapl=self._lapl, lcpl=self._lcpl)
def visit(self, func):
""" Recursively visit all names in this group and subgroups.
Note: visit ignores soft and external links. To visit those, use
visit_links.
You supply a callable (function, method or callable object); it
will be called exactly once for each link in this group and every
group below it. Your callable must conform to the signature:
func(<member name>) => <None or return value>
Returning None continues iteration, returning anything else stops
and immediately returns that value from the visit method. No
particular order of iteration within groups is guaranteed.
Example:
>>> # List the entire contents of the file
>>> f = File("foo.hdf5")
>>> list_of_names = []
>>> f.visit(list_of_names.append)
"""
with phil:
def proxy(name):
""" Call the function with the text name, not bytes """
return func(self._d(name))
return h5o.visit(self.id, proxy)
def visititems(self, func):
""" Recursively visit names and objects in this group.
Note: visititems ignores soft and external links. To visit those, use
visititems_links.
You supply a callable (function, method or callable object); it
will be called exactly once for each link in this group and every
group below it. Your callable must conform to the signature:
func(<member name>, <object>) => <None or return value>
Returning None continues iteration, returning anything else stops
and immediately returns that value from the visit method. No
particular order of iteration within groups is guaranteed.
Example:
# Get a list of all datasets in the file
>>> mylist = []
>>> def func(name, obj):
... if isinstance(obj, Dataset):
... mylist.append(name)
...
>>> f = File('foo.hdf5')
>>> f.visititems(func)
"""
with phil:
def proxy(name):
""" Use the text name of the object, not bytes """
name = self._d(name)
return func(name, self[name])
return h5o.visit(self.id, proxy)
def visit_links(self, func):
""" Recursively visit all names in this group and subgroups.
Each link will be visited exactly once, regardless of its target.
You supply a callable (function, method or callable object); it
will be called exactly once for each link in this group and every
group below it. Your callable must conform to the signature:
func(<member name>) => <None or return value>
Returning None continues iteration, returning anything else stops
and immediately returns that value from the visit method. No
particular order of iteration within groups is guaranteed.
Example:
>>> # List the entire contents of the file
>>> f = File("foo.hdf5")
>>> list_of_names = []
>>> f.visit_links(list_of_names.append)
"""
with phil:
def proxy(name):
""" Call the function with the text name, not bytes """
return func(self._d(name))
return self.id.links.visit(proxy)
def visititems_links(self, func):
""" Recursively visit links in this group.
Each link will be visited exactly once, regardless of its target.
You supply a callable (function, method or callable object); it
will be called exactly once for each link in this group and every
group below it. Your callable must conform to the signature:
func(<member name>, <link>) => <None or return value>
Returning None continues iteration, returning anything else stops
and immediately returns that value from the visit method. No
particular order of iteration within groups is guaranteed.
Example:
# Get a list of all softlinks in the file
>>> mylist = []
>>> def func(name, link):
... if isinstance(link, SoftLink):
... mylist.append(name)
...
>>> f = File('foo.hdf5')
>>> f.visititems_links(func)
"""
with phil:
def proxy(name):
""" Use the text name of the object, not bytes """
name = self._d(name)
return func(name, self.get(name, getlink=True))
return self.id.links.visit(proxy)
@with_phil
def __repr__(self):
if not self:
r = u"<Closed HDF5 group>"
else:
namestr = (
'"%s"' % self.name
) if self.name is not None else u"(anonymous)"
r = '<HDF5 group %s (%d members)>' % (namestr, len(self))
return r
class HardLink:
"""
Represents a hard link in an HDF5 file. Provided only so that
Group.get works in a sensible way. Has no other function.
"""
pass
class SoftLink:
"""
Represents a symbolic ("soft") link in an HDF5 file. The path
may be absolute or relative. No checking is performed to ensure
that the target actually exists.
"""
@property
def path(self):
""" Soft link value. Not guaranteed to be a valid path. """
return self._path
def __init__(self, path):
self._path = str(path)
def __repr__(self):
return '<SoftLink to "%s">' % self.path
class ExternalLink:
"""
Represents an HDF5 external link. Paths may be absolute or relative.
No checking is performed to ensure either the target or file exists.
"""
@property
def path(self):
""" Soft link path, i.e. the part inside the HDF5 file. """
return self._path
@property
def filename(self):
""" Path to the external HDF5 file in the filesystem. """
return self._filename
def __init__(self, filename, path):
self._filename = filename_decode(filename_encode(filename))
self._path = path
def __repr__(self):
return '<ExternalLink to "%s" in file "%s"' % (self.path,
self.filename)