Merge pull request #4 from psc-code/pr/build

build bits

.copier-answers.yml

@@ -0,0 +1,12 @@
+# Changes here will be overwritten by Copier; NEVER EDIT MANUALLY
+_commit: 2024.03.10
+_src_path: gh:scientific-python/cookie
+backend: skbuild
+email: [email protected]
+full_name: Kai Germaschewski
+license: BSD
+org: psc-code
+project_name: xarray-pschdf5
+project_short_description: XArray reader for PSC HDF5 data
+url: https://github.com/psc-code/xarray-pschdf5
+vcs: true

.github/workflows/ci.yml

@@ -28,10 +28,10 @@ jobs:
       - uses: pre-commit/[email protected]
         with:
           extra_args: --hook-stage manual --all-files
-      - name: Run PyLint
-        run: |
-          echo "::add-matcher::$GITHUB_WORKSPACE/.github/matchers/pylint.json"
-          pipx run nox -s pylint
+      # - name: Run PyLint
+      #   run: |
+      #     echo "::add-matcher::$GITHUB_WORKSPACE/.github/matchers/pylint.json"
+      #     pipx run nox -s pylint
 
   checks:
     name: Check Python ${{ matrix.python-version }} on ${{ matrix.runs-on }}

.pre-commit-config.yaml

@@ -52,14 +52,14 @@ repos:
       - id: clang-format
         types_or: [c++, c, cuda]
 
-  - repo: https://github.com/pre-commit/mirrors-mypy
-    rev: "v1.8.0"
-    hooks:
-      - id: mypy
-        files: src|tests
-        args: []
-        additional_dependencies:
-          - pytest
+  # - repo: https://github.com/pre-commit/mirrors-mypy
+  #   rev: "v1.8.0"
+  #   hooks:
+  #     - id: mypy
+  #       files: src|tests
+  #       args: []
+  #       additional_dependencies:
+  #         - pytest
 
   - repo: https://github.com/codespell-project/codespell
     rev: "v2.2.6"

CMakeLists.txt

@@ -1,8 +1,2 @@
 cmake_minimum_required(VERSION 3.15...3.26)
 project(${SKBUILD_PROJECT_NAME} LANGUAGES CXX)
-
-set(PYBIND11_FINDPYTHON ON)
-find_package(pybind11 CONFIG REQUIRED)
-
-pybind11_add_module(_core MODULE src/main.cpp)
-install(TARGETS _core DESTINATION ${SKBUILD_PROJECT_NAME})

pyproject.toml

@@ -1,5 +1,5 @@
 [build-system]
-requires = ["pybind11", "scikit-build-core"]
+requires = ["scikit-build-core"]
 build-backend = "scikit_build_core.build"
 
 
@@ -33,6 +33,7 @@ dynamic = ["version"]
 dependencies = [
   "pugixml >= 0.5.0",
   "h5py",
+  "xarray",
 ]
 
 [project.optional-dependencies]
@@ -145,6 +146,8 @@ ignore = [
   "PLR09",    # Too many <...>
   "PLR2004",  # Magic value used in comparison
   "ISC001",   # Conflicts with formatter
+  "C408",     # like my dict() calls
+  "RET504",
 ]
 isort.required-imports = ["from __future__ import annotations"]
 # Uncomment if using a _compat.typing backport

src/xarray_pschdf5/__init__.py

@@ -9,5 +9,3 @@
 from ._version import version as __version__
 
 __all__ = ["__version__"]
-
-from .pschdf5_backend import PscHdf5Entrypoint

src/xarray_pschdf5/pschdf5_backend.py

@@ -1,10 +1,14 @@
+from __future__ import annotations
+
+import os
 from collections import OrderedDict
-from pugixml import pugi
+from typing import Any, ClassVar
+
 import h5py
+import numpy as np
 import xarray as xr
+from pugixml import pugi
 from xarray.backends import BackendEntrypoint
-import numpy as np
-import os
 
 
 class PscHdf5Entrypoint(BackendEntrypoint):
@@ -18,7 +22,7 @@ def open_dataset(
     ):
         return pschdf5_open_dataset(filename_or_obj, drop_variables=drop_variables)
 
-    open_dataset_parameters = ["filename_or_obj", "drop_variables"]
+    open_dataset_parameters: ClassVar[Any] = ["filename_or_obj", "drop_variables"]
 
     def guess_can_open(self, filename_or_obj):
         if filename_or_obj.endswith(".xdmf"):
@@ -39,90 +43,101 @@ def pschdf5_open_dataset(filename_or_obj, *, drop_variables=None):
 
     vars = dict()
     assert len(grids) == 1
-    for gridname, grid in grids.items():
+    for _, grid in grids.items():
         for fldname, fld in grid["fields"].items():
-            data_dims = fld["dims"]
+            if fldname in drop_variables:
+                continue
+
+            # data_dims = fld["dims"]
             data_path = fld["path"]
-            h5_filename, h5_path = fld['path'].split(':')
-            h5_file = h5py.File(dirname + '/' + h5_filename)
+            h5_filename, h5_path = fld["path"].split(":")
+            h5_file = h5py.File(dirname + "/" + h5_filename)
             data = h5_file[h5_path][:].T
 
             data_attrs = dict(path=data_path)
             vars[fldname] = xr.DataArray(
-                data=data,
-                dims=["x", "y", "z"],
-                attrs=data_attrs)
-
-    coords = {"xyz"[d]: make_crd(grid['topology']['dims'][d], grid['geometry']
-                                 ['origin'][d], grid['geometry']['spacing'][d]) for d in range(3)}
+                data=data, dims=["x", "y", "z"], attrs=data_attrs
+            )
+
+    coords = {
+        "xyz"[d]: make_crd(
+            grid["topology"]["dims"][d],
+            grid["geometry"]["origin"][d],
+            grid["geometry"]["spacing"][d],
+        )
+        for d in range(3)
+    }
 
     attrs = dict(run=meta["run"], time=meta["time"])
 
     ds = xr.Dataset(vars, coords=coords, attrs=attrs)
-#    ds.set_close(my_close_method)
+    #    ds.set_close(my_close_method)
 
     return ds
 
 
 def make_crd(dim, origin, spacing):
-    return origin + np.arange(.5, dim) * spacing
+    return origin + np.arange(0.5, dim) * spacing
 
 
 def read_xdmf(filename):
     doc = pugi.XMLDocument()
     result = doc.load_file(filename)
     if not result:
-        print('parse error: status=%r description=%r' %
-              (result.status, result.description()))
-        assert False
-
-    grid_collection = doc.child('Xdmf').child('Domain').child('Grid')
-    assert grid_collection.attribute('GridType').value() == 'Collection'
-    grid_time = grid_collection.child('Time')
-    assert grid_time.attribute('Type').value() == 'Single'
-    time = grid_time.attribute('Value').value()
+        raise f"parse error: status={result.status} description={result.description()}"
+
+    grid_collection = doc.child("Xdmf").child("Domain").child("Grid")
+    assert grid_collection.attribute("GridType").value() == "Collection"
+    grid_time = grid_collection.child("Time")
+    assert grid_time.attribute("Type").value() == "Single"
+    time = grid_time.attribute("Value").value()
     rv = {}
-    rv['time'] = time
-    rv['grids'] = {}
+    rv["time"] = time
+    rv["grids"] = {}
     for node in grid_collection.children():
-        if node.name() == 'Grid':
+        if node.name() == "Grid":
             grid = {}
-            grid_name = node.attribute('Name').value()
-            topology = node.child('Topology')
+            grid_name = node.attribute("Name").value()
+            topology = node.child("Topology")
             # assert topology.attribute('TopologyType').value() == '3DCoRectMesh'
-            dims = topology.attribute('Dimensions').value()
+            dims = topology.attribute("Dimensions").value()
             dims = np.asarray([int(d) - 1 for d in dims.split(" ")])[::-1]
-            grid['topology'] = {'type': topology.attribute(
-                'TopologyType').value(), 'dims': dims}
+            grid["topology"] = {
+                "type": topology.attribute("TopologyType").value(),
+                "dims": dims,
+            }
+
+            geometry = node.child("Geometry")
+            assert geometry.attribute("GeometryType").value() == "Origin_DxDyDz"
 
-            geometry = node.child('Geometry')
-            assert geometry.attribute(
-                'GeometryType').value() == 'Origin_DxDyDz'
+            grid["geometry"] = dict()
             for child in geometry.children():
-                if child.attribute('Name').value() == 'Origin':
-                    origin = np.asarray(
-                        [float(x) for x in child.text().as_string().split(" ")])[::-1]
-                if child.attribute('Name').value() == 'Spacing':
-                    spacing = np.asarray(
-                        [float(x) for x in child.text().as_string().split(" ")])[::-1]
+                if child.attribute("Name").value() == "Origin":
+                    grid["geometry"]["origin"] = np.asarray(
+                        [float(x) for x in child.text().as_string().split(" ")]
+                    )[::-1]
 
-            grid['geometry'] = {'origin': origin, 'spacing': spacing}
+                if child.attribute("Name").value() == "Spacing":
+                    grid["geometry"]["spacing"] = np.asarray(
+                        [float(x) for x in child.text().as_string().split(" ")]
+                    )[::-1]
 
             flds = OrderedDict()
             for child in node.children():
-                if child.name() != 'Attribute':
+                if child.name() != "Attribute":
                     continue
 
-                fld = child.attribute('Name').value()
-                item = child.child('DataItem')
-                fld_dims = np.asarray([int(d) for d in item.attribute(
-                    'Dimensions').value().split(" ")])[::-1]
+                fld = child.attribute("Name").value()
+                item = child.child("DataItem")
+                fld_dims = np.asarray(
+                    [int(d) for d in item.attribute("Dimensions").value().split(" ")]
+                )[::-1]
                 assert np.all(fld_dims == dims)
-                assert item.attribute('Format').value() == 'HDF'
+                assert item.attribute("Format").value() == "HDF"
                 path = item.text().as_string()
-                flds[fld] = {'path': path, 'dims': dims}
+                flds[fld] = {"path": path, "dims": dims}
 
-            grid['fields'] = flds
-            rv['grids'][grid_name] = grid
+            grid["fields"] = flds
+            rv["grids"][grid_name] = grid
 
     return rv