h5netcdf 1.6.3__py3-none-any.whl → 1.7.0__py3-none-any.whl

h5netcdf/utils.py

@@ -1,5 +1,12 @@
 from collections.abc import Mapping
 
+import h5py
+import numpy as np
+
+
+class CompatibilityError(Exception):
+    """Raised when using features that are not part of the NetCDF4 API."""
+
 
 class Frozen(Mapping):
     """Wrapper around an object implementing the mapping interface to make it
@@ -24,3 +31,201 @@ class Frozen(Mapping):
 
     def __repr__(self):
         return f"{type(self).__name__}({self._mapping!r})"
+
+
+def _create_classic_string_dataset(gid, name, value, shape, chunks):
+    """Write a string dataset to an HDF5 object with control over the strpad.
+
+    Parameters
+    ----------
+    gid : h5py.h5g.GroupID
+        Group ID where to write the dataset.
+    name : str
+        Dataset name.
+    value : str
+        Dataset contents to be written.
+    shape : tuple
+        Dataset shape.
+    chunks : tuple or None
+        Chunk shape.
+    """
+    tid = h5py.h5t.C_S1.copy()
+    tid.set_size(1)
+    tid.set_strpad(h5py.h5t.STR_NULLTERM)
+    kwargs = {}
+    if len(shape) == 0:
+        sid = h5py.h5s.create(h5py.h5s.SCALAR)
+    else:
+        if chunks is not None:
+            # for resizing, we need to provide maxshape
+            # with unlimited as the first dimension
+            maxshape = (h5py.h5s.UNLIMITED,) + shape[1:]
+            # and we also need to create a chunked dataset
+            dcpl = h5py.h5p.create(h5py.h5p.DATASET_CREATE)
+            # try automatic chunking
+            dcpl.set_chunk(chunks)
+            kwargs["dcpl"] = dcpl
+        else:
+            maxshape = shape
+        sid = h5py.h5s.create_simple(shape, maxshape)
+    did = h5py.h5d.create(gid, name.encode(), tid, sid, **kwargs)
+    if value is not None:
+        value = np.array(np.bytes_(value))
+        did.write(h5py.h5s.ALL, h5py.h5s.ALL, value, mtype=did.get_type())
+
+
+def _create_enum_dataset(group, name, shape, enum_type, fillvalue=None):
+    """Create a dataset with a transient enum dtype.
+
+    Parameters
+    ----------
+    group : h5netcdf.Group
+    name : str
+        dataset name
+    shape : tuple
+        dataset shape
+    enum_type : h5netcdf.EnumType
+
+    Keyword arguments
+    -----------------
+    fillvalue : optional scalar fill value
+    """
+    # copy from existing committed type
+    enum_tid = enum_type._h5ds.id.copy()
+    space = h5py.h5s.create_simple(shape)
+
+    dcpl = h5py.h5p.create(h5py.h5p.DATASET_CREATE)
+    if fillvalue is not None:
+        dcpl.set_fill_value(np.array(fillvalue, dtype=enum_type.dtype))
+
+    h5py.h5d.create(group._h5group.id, name.encode("ascii"), enum_tid, space, dcpl=dcpl)
+    enum_tid.close()
+
+
+def _create_enum_dataset_attribute(ds, name, value, enum_type):
+    """Create an enum attribute at the dataset.
+
+    Parameters
+    ----------
+    ds : h5netcdf.Variable
+    name : str
+        dataset name
+    value : array of ints
+        enum values
+    enum_type : h5netcdf.EnumType
+    """
+    tid = enum_type._h5ds.id.copy()
+    space = h5py.h5s.create_simple((1,))
+
+    aid = h5py.h5a.create(ds._h5ds.id, name.encode("ascii"), tid, space)
+    aid.write(value, mtype=aid.get_type())
+
+
+def _make_enum_tid(enum_dict, basetype):
+    """Make enum tid
+
+    Parameters
+    ----------
+    enum_dict : dict
+        dictionary with Enum field/value pairs
+    basetype : np.dtype
+        basetype of the enum
+    """
+    items = sorted(enum_dict.items(), key=lambda kv: kv[1])  # sort by value
+    base_tid = h5py.h5t.py_create(np.dtype(basetype))
+    tid = h5py.h5t.enum_create(base_tid)
+    for name, val in items:
+        tid.enum_insert(name.encode("utf-8"), int(val))
+    return tid
+
+
+def _commit_enum_type(group, name, enum_dict, basetype):
+    """Commit an enum type to the given group.
+
+    Parameters
+    ----------
+    group : h5netcdf.Group
+    name : str
+        dataset name
+    enum_dict : dict
+        dictionary with Enum field/value pairs
+    basetype : np.dtype
+        basetype of the enum
+    """
+    tid = _make_enum_tid(enum_dict, basetype)
+    tid.commit(group._h5group.id, name.encode("ascii"))
+    tid.close()
+
+
+def _create_string_attribute(gid, name, value):
+    """Create a string attribute to an HDF5 object with control over the strpad.
+
+    Parameters
+    ----------
+    gid : h5py.h5g.GroupID
+      Group ID where to write the attribute.
+    name : str
+        Attribute name.
+    value : str
+        Attributes contents to be written.
+    """
+    # handle charset and encoding
+    charset = h5py.h5t.CSET_ASCII
+    if isinstance(value, str):
+        if not value.isascii():
+            value = value.encode("utf-8")
+            charset = h5py.h5t.CSET_UTF8
+        else:
+            value = value.encode("ascii")
+            charset = h5py.h5t.CSET_ASCII
+
+    # create string type
+    tid = h5py.h5t.C_S1.copy()
+    tid.set_size(len(value))
+    tid.set_strpad(h5py.h5t.STR_NULLTERM)
+    tid.set_cset(charset)
+
+    sid = h5py.h5s.create(h5py.h5s.SCALAR)
+
+    if h5py.h5a.exists(gid, name.encode()):
+        h5py.h5a.delete(gid, name.encode())
+
+    aid = h5py.h5a.create(gid, name.encode(), tid, sid)
+    value = np.array(value)
+    aid.write(value, mtype=aid.get_type())
+
+
+def h5dump(fn: str, dataset=None, strict=False):
+    """Call h5dump on an h5netcdf file."""
+    import re
+    import subprocess
+
+    arglist = ["h5dump", "-A"]
+    if dataset is not None:
+        arglist.append(f"-d {dataset}")
+    arglist.append(fn)
+
+    out = subprocess.run(arglist, check=False, capture_output=True).stdout.decode()
+
+    # Strip non-deterministic components
+    out = re.sub(r"DATASET [0-9]+ ", "DATASET XXXX ", out)
+
+    # Strip the _NCProperties header, which includes software versions which won't match.
+    pattern = (
+        r'ATTRIBUTE "_NCProperties"'  # match the attribute start
+        r"\s*{"  # opening brace
+        r"(?:[^{}]*{[^{}]*}[^{}]*)*"  # match multiple inner braces
+        r"}"  # closing brace
+    )
+    out = re.sub(
+        pattern,
+        'ATTRIBUTE "_NCProperties" { ... }',
+        out,
+        flags=re.DOTALL,
+    )
+
+    if not strict:
+        out = re.sub(r"STRPAD H5T_STR_NULL(?:TERM|PAD);", "STRPAD { ... };", out)
+        out = re.sub(r"CSET H5T_CSET_(?:UTF8|ASCII);", "CSET { ... };", out)
+
+    return out

{h5netcdf-1.6.3.dist-info → h5netcdf-1.7.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: h5netcdf
-Version: 1.6.3
+Version: 1.7.0
 Summary: netCDF4 via h5py
 Author-email: Stephan Hoyer <shoyer@gmail.com>, Kai Mühlbauer <kmuehlbauer@wradlib.org>
 Maintainer-email: h5netcdf developers <devteam@h5netcdf.org>
@@ -124,10 +124,10 @@ Usage
 -----
 
 h5netcdf has two APIs, a new API and a legacy API. Both interfaces currently
-reproduce most of the features of the netCDF interface, with the notable
-exception of support for operations that rename or delete existing objects.
-We simply haven't gotten around to implementing this yet. Patches
-would be very welcome.
+reproduce most of the features of the netCDF interface, including the ability
+to write NETCDF4 and NETCDF4_CLASSIC formatted files. Support for operations
+that rename or delete existing objects is still missing, and patches would be
+very welcome.
 
 New API
 ~~~~~~~

h5netcdf-1.7.0.dist-info/RECORD

@@ -0,0 +1,16 @@
+h5netcdf/__init__.py,sha256=Y0EBCcmlJctwl1kCmj7yLijTVy9AioBTr2091vInAtw,456
+h5netcdf/_version.py,sha256=oGRWiKvEGHesjf5wCNHGVlYfAA3dInDJeL5EiMaru6A,704
+h5netcdf/attrs.py,sha256=yPwGhaEhr7TtihH--Tx24j7gKgRDQIXs-mtx6ngIZRo,4494
+h5netcdf/core.py,sha256=Fg_68fF9vcyozlk-2vO1Qa0U_oRbvWfE976ax--CvxM,69698
+h5netcdf/dimensions.py,sha256=pxtt3ID55bTMp6djIpAka3RgXzioIUs9qBBgMqd9HhU,8336
+h5netcdf/legacyapi.py,sha256=MIZlht5Ad4hDFF1Slz2vXmKkgbv7Fhhf2YwNIe16Lfk,7682
+h5netcdf/utils.py,sha256=btxKI-VZP-Wn0Rk_wmnhnUls-mrxO42w0s_uX_su4FI,6528
+h5netcdf/tests/conftest.py,sha256=qS7XTZxos0NIRFtMCJwVEyx0paZw8Le1loPK1MtoQ_0,2350
+h5netcdf/tests/pytest.ini,sha256=ruJxrLdCIA4bCPVuPQjxsLSlvVxuIsIakK6iQOmz-ak,107
+h5netcdf/tests/test_h5netcdf.py,sha256=s5m3N3huMsxmIAN_W78-YnawB-lKWQPmQgSutO6bwzA,116652
+h5netcdf-1.7.0.dist-info/licenses/AUTHORS.txt,sha256=LTKzUh9o4Wc_oT3aFC48cyDCCP6tdm6VEV_6RrNy4uo,272
+h5netcdf-1.7.0.dist-info/licenses/LICENSE,sha256=Xer1Jg8iL_n9Da0xt0S99blk6tsg9tee_JdgH1rWTjs,1505
+h5netcdf-1.7.0.dist-info/METADATA,sha256=lNPv7WhB1VSjMJYsmC0qJX_wRvRP_deMsIXidu0Ch3A,13396
+h5netcdf-1.7.0.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+h5netcdf-1.7.0.dist-info/top_level.txt,sha256=Fb_KIpOE6MBqjSvxV1Ay7oYce1mdmQ1pO9JQJPDeGqg,9
+h5netcdf-1.7.0.dist-info/RECORD,,

h5netcdf-1.6.3.dist-info/RECORD

@@ -1,16 +0,0 @@
-h5netcdf/__init__.py,sha256=Y0EBCcmlJctwl1kCmj7yLijTVy9AioBTr2091vInAtw,456
-h5netcdf/_version.py,sha256=RrM0JkMSEquWJIGj8RCqoST__VpnXFzYf3k-uS1FCko,511
-h5netcdf/attrs.py,sha256=4IvV4ULLWkz4igFsvu9S2LB745wgUKrIdIuSeO5kpX8,3581
-h5netcdf/core.py,sha256=xO62cSVvXXd-EQ7xuLjZC_gc99ou0naNmyorW7bKZxs,64150
-h5netcdf/dimensions.py,sha256=ln6n3J-BE2PlJDPlr9YiqPYHhXRUHVH3j_7_1O0VGS0,7802
-h5netcdf/legacyapi.py,sha256=MIZlht5Ad4hDFF1Slz2vXmKkgbv7Fhhf2YwNIe16Lfk,7682
-h5netcdf/utils.py,sha256=6E-HAIE0ONMyL4SxI3oUyQvrDgDWifR5EPde91V9rT0,674
-h5netcdf/tests/conftest.py,sha256=4fLa2qoB8br2UpokaOn1-mjHgqVUgVV0G3QLIUzfbZo,2133
-h5netcdf/tests/pytest.ini,sha256=ruJxrLdCIA4bCPVuPQjxsLSlvVxuIsIakK6iQOmz-ak,107
-h5netcdf/tests/test_h5netcdf.py,sha256=4VslXYKJNx6vOhhyc4_-mCc7M4W_l8fftZPoMk7x-50,110275
-h5netcdf-1.6.3.dist-info/licenses/AUTHORS.txt,sha256=LTKzUh9o4Wc_oT3aFC48cyDCCP6tdm6VEV_6RrNy4uo,272
-h5netcdf-1.6.3.dist-info/licenses/LICENSE,sha256=Xer1Jg8iL_n9Da0xt0S99blk6tsg9tee_JdgH1rWTjs,1505
-h5netcdf-1.6.3.dist-info/METADATA,sha256=SctpwYrbXsezPM2YQwNsV2dkXn3mEB5PSR7zwdvWRtI,13387
-h5netcdf-1.6.3.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-h5netcdf-1.6.3.dist-info/top_level.txt,sha256=Fb_KIpOE6MBqjSvxV1Ay7oYce1mdmQ1pO9JQJPDeGqg,9
-h5netcdf-1.6.3.dist-info/RECORD,,