Update _widget_reader.py

src/yt_napari/_tests/test_widget_reader.py

@@ -1,17 +1,21 @@
+import json
+
+# note: the cache is disabled for all the tests in this file due to flakiness
+# in github CI. It may be that loading from a true file, rather than the
+# yt_ugrid_ds_fn fixture would fix that...
+import os
 from functools import partial
+from unittest.mock import patch
 
 import numpy as np
 
 from yt_napari import _widget_reader as _wr
+from yt_napari._data_model import InputModel
 from yt_napari._ds_cache import dataset_cache
 
 # import ReaderWidget, SelectionEntry, TimeSeriesReader
 from yt_napari._special_loaders import _construct_ugrid_timeseries
 
-# note: the cache is disabled for all the tests in this file due to flakiness
-# in github CI. It may be that loading from a true file, rather than the
-# yt_ugrid_ds_fn fixture would fix that...
-
 
 def test_widget_reader_add_selections(make_napari_viewer, yt_ugrid_ds_fn):
     viewer = make_napari_viewer()
@@ -45,6 +49,57 @@ def _rebuild_data(final_shape, data):
     return np.random.random(final_shape) * data.mean()
 
 
+def test_save_widget_reader(make_napari_viewer, yt_ugrid_ds_fn, tmp_path):
+    viewer = make_napari_viewer()
+    r = _wr.ReaderWidget(napari_viewer=viewer)
+    r.ds_container.filename.value = yt_ugrid_ds_fn
+    r.ds_container.store_in_cache.value = False
+    r.add_new_button.click()
+    sel = list(r.active_selections.values())[0]
+    assert isinstance(sel, _wr.SelectionEntry)
+
+    mgui_region = sel.selection_container_raw
+    mgui_region.fields.field_type.value = "enzo"
+    mgui_region.fields.field_name.value = "Density"
+    mgui_region.resolution.value = (400, 400, 400)
+
+    rebuild = partial(_rebuild_data, mgui_region.resolution.value)
+    r._post_load_function = rebuild
+
+    temp_file = tmp_path / "test.json"
+
+    with patch("PyQt5.QtWidgets.QFileDialog.exec_") as mock_exec, patch(
+        "PyQt5.QtWidgets.QFileDialog.selectedFiles"
+    ) as mock_selectedFiles:
+        # Set the return values for the mocked functions
+        mock_exec.return_value = 1
+        mock_selectedFiles.return_value = [temp_file]
+
+        r.save_selection()
+
+    assert os.path.exists(temp_file)
+    with open(temp_file, "r") as json_file:
+        saved_data = json.load(json_file)
+
+    assert (
+        saved_data["datasets"][0]["selections"]["regions"][0]["fields"][0]["field_type"]
+        == "enzo"
+    )
+    assert (
+        saved_data["datasets"][0]["selections"]["regions"][0]["fields"][0]["field_name"]
+        == "Density"
+    )
+    assert saved_data["datasets"][0]["selections"]["regions"][0]["resolution"] == [
+        400,
+        400,
+        400,
+    ]
+
+    # ensure that the saved json is a valid model
+    _ = InputModel.parse_obj(saved_data)
+    r.deleteLater()
+
+
 def test_widget_reader(make_napari_viewer, yt_ugrid_ds_fn):
     viewer = make_napari_viewer()
     r = _wr.ReaderWidget(napari_viewer=viewer)
@@ -138,4 +193,42 @@ def test_timeseries_widget_reader(make_napari_viewer, tmp_path):
     tsr.load_data()
     assert len(viewer.layers) == 2
 
+    temp_file = tmp_path / "test.json"
+
+    # Use patch to replace the actual QFileDialog functions with mock functions
+    with patch("PyQt5.QtWidgets.QFileDialog.exec_") as mock_exec, patch(
+        "PyQt5.QtWidgets.QFileDialog.selectedFiles"
+    ) as mock_selectedFiles:
+        # Set the return values for the mocked functions
+        mock_exec.return_value = 1  # Assuming QDialog::Accepted is 1
+        mock_selectedFiles.return_value = [temp_file]
+
+        # Call the save_selection method
+        tsr.save_selection()
+
+    assert os.path.exists(temp_file)
+    with open(temp_file, "r") as json_file:
+        saved_data = json.load(json_file)
+
+    assert (
+        saved_data["timeseries"][0]["selections"]["regions"][0]["fields"][0][
+            "field_type"
+        ]
+        == "stream"
+    )
+    assert (
+        saved_data["timeseries"][0]["selections"]["regions"][0]["fields"][0][
+            "field_name"
+        ]
+        == "density"
+    )
+    assert saved_data["timeseries"][0]["selections"]["regions"][0]["resolution"] == [
+        10,
+        10,
+        10,
+    ]
+
+    # ensure that the saved json is a valid model
+    _ = InputModel.parse_obj(saved_data)
+
     tsr.deleteLater()

src/yt_napari/_widget_reader.py

@@ -1,14 +1,23 @@
+import json
 from collections import defaultdict
 from typing import Callable, Optional
 
 import napari
 from magicgui import widgets
 from napari.qt.threading import thread_worker
 from qtpy import QtCore
-from qtpy.QtWidgets import QComboBox, QHBoxLayout, QPushButton, QVBoxLayout, QWidget
+from qtpy.QtWidgets import (
+    QComboBox,
+    QFileDialog,
+    QHBoxLayout,
+    QPushButton,
+    QVBoxLayout,
+    QWidget,
+)
 
 from yt_napari import _data_model, _gui_utilities, _model_ingestor
 from yt_napari._ds_cache import dataset_cache
+from yt_napari._schema_version import schema_name
 from yt_napari.viewer import _check_for_reference_layer
 
 
@@ -68,7 +77,9 @@ def add_spatial_selection_widgets(self):
         self.layout().addLayout(removal_group_layout)
 
     def add_load_group_widgets(self):
-        pass
+        """
+        add the widgets related to the Load button
+        """
 
     def add_a_selection(self):
         selection_type = self.new_selection_type.currentText()
@@ -106,6 +117,26 @@ def add_load_group_widgets(self):
         load_group.addWidget(cc.native)
         self.layout().addLayout(load_group)
 
+        ss = widgets.PushButton(text="Save Selection")
+        ss.clicked.connect(self.save_selection)
+        load_group.addWidget(ss.native)
+
+    def save_selection(self):
+        py_kwargs = {}
+        py_kwargs = self._validate_data_model()
+
+        file_dialog = QFileDialog()
+        file_dialog.setFileMode(QFileDialog.AnyFile)
+        file_dialog.setAcceptMode(QFileDialog.AcceptSave)
+        file_dialog.setNameFilter("JSON Files (*.json);;All Files (*)")
+
+        if file_dialog.exec_():
+            file_path = file_dialog.selectedFiles()[0]
+            if file_path:
+                # Save the JSON data to the selected file
+                with open(file_path, "w") as json_file:
+                    json.dump(py_kwargs, json_file, indent=4)
+
     def clear_cache(self):
         dataset_cache.rm_all()
 
@@ -114,7 +145,29 @@ def load_data(self):
         # instantiate pydantic objects, which are then handed off to the
         # same data ingestion function as the json loader.
 
-        # first, get the pydantic args for each selection type, embed in lists
+        py_kwargs = {}
+        py_kwargs = self._validate_data_model()
+        model = _data_model.InputModel.parse_obj(py_kwargs)
+
+        # process each layer
+        layer_list, _ = _model_ingestor._process_validated_model(model)
+
+        # align all layers after checking for or setting the reference layer
+        ref_layer = _check_for_reference_layer(self.viewer.layers)
+        if ref_layer is None:
+            ref_layer = _model_ingestor._choose_ref_layer(layer_list)
+        layer_list = ref_layer.align_sanitize_layers(layer_list)
+
+        for new_layer in layer_list:
+            im_arr, im_kwargs, _ = new_layer
+            if self._post_load_function is not None:
+                im_arr = self._post_load_function(im_arr)
+
+            # add the new layer
+            self.viewer.add_image(im_arr, **im_kwargs)
+
+    def _validate_data_model(self):
+        # this function save json data
         selections_by_type = defaultdict(list)
         for selection in self.active_selections.values():
             py_kwargs = selection.get_current_pydantic_kwargs()
@@ -129,34 +182,17 @@ def load_data(self):
             py_kwargs,
             ignore_attrs="selections",
         )
-
         # add selections in
         py_kwargs["selections"] = selections_by_type
 
         # now ready to instantiate the base model
         py_kwargs = {
+            "$schema": schema_name,
             "datasets": [
                 py_kwargs,
-            ]
+            ],
         }
-        model = _data_model.InputModel.parse_obj(py_kwargs)
-
-        # process each layer
-        layer_list, _ = _model_ingestor._process_validated_model(model)
-
-        # align all layers after checking for or setting the reference layer
-        ref_layer = _check_for_reference_layer(self.viewer.layers)
-        if ref_layer is None:
-            ref_layer = _model_ingestor._choose_ref_layer(layer_list)
-        layer_list = ref_layer.align_sanitize_layers(layer_list)
-
-        for new_layer in layer_list:
-            im_arr, im_kwargs, _ = new_layer
-            if self._post_load_function is not None:
-                im_arr = self._post_load_function(im_arr)
-
-            # add the new layer
-            self.viewer.add_image(im_arr, **im_kwargs)
+        return py_kwargs
 
 
 class SelectionEntry(QWidget):
@@ -223,7 +259,50 @@ def add_load_group_widgets(self):
         load_group.addWidget(pb.native)
         self.layout().addLayout(load_group)
 
+        ss = widgets.PushButton(text="Save Selection")
+        ss.clicked.connect(self.save_selection)
+        load_group.addWidget(ss.native)
+
+    def save_selection(self):
+        py_kwargs = {}
+        py_kwargs = self._validate_data_model()
+        # model = _data_model.InputModel.parse_obj(py_kwargs)
+
+        file_dialog = QFileDialog()
+        file_dialog.setFileMode(QFileDialog.AnyFile)
+        file_dialog.setAcceptMode(QFileDialog.AcceptSave)
+        file_dialog.setNameFilter("JSON Files (*.json);;All Files (*)")
+
+        if file_dialog.exec_():
+            file_path = file_dialog.selectedFiles()[0]
+            if file_path:
+                # Save the JSON data to the selected file
+                with open(file_path, "w") as json_file:
+                    json.dump(py_kwargs, json_file, indent=4)
+
     def load_data(self):
+        py_kwargs = {}
+        py_kwargs = self._validate_data_model()
+        model = _data_model.InputModel.parse_obj(py_kwargs)
+
+        if _use_threading:  # pragma: no cover
+            worker = time_series_load(model)
+            worker.returned.connect(self.process_timeseries_layers)
+            worker.start()
+        else:
+            _, layer_list = _model_ingestor._process_validated_model(model)
+            self.process_timeseries_layers(layer_list)
+
+    def process_timeseries_layers(self, layer_list):
+        for new_layer in layer_list:
+            im_arr, im_kwargs, _ = new_layer
+            # probably can remove since the _special_loaders can be used
+            # if self._post_load_function is not None:
+            #     im_arr = self._post_load_function(im_arr)
+            # add the new layer
+            self.viewer.add_image(im_arr, **im_kwargs)
+
+    def _validate_data_model(self):
         # first, get the pydantic args for each selection type, embed in lists
         selections_by_type = defaultdict(list)
         for selection in self.active_selections.values():
@@ -254,32 +333,15 @@ def load_data(self):
 
         # now ready to instantiate the base model
         py_kwargs = {
+            "$schema": schema_name,
             "timeseries": [
                 py_kwargs,
-            ]
+            ],
         }
-
-        model = _data_model.InputModel.parse_obj(py_kwargs)
-
-        if _use_threading:
-            worker = time_series_load(model)
-            worker.returned.connect(self.process_timeseries_layers)
-            worker.start()
-        else:
-            _, layer_list = _model_ingestor._process_validated_model(model)
-            self.process_timeseries_layers(layer_list)
-
-    def process_timeseries_layers(self, layer_list):
-        for new_layer in layer_list:
-            im_arr, im_kwargs, _ = new_layer
-            # probably can remove since the _special_loaders can be used
-            # if self._post_load_function is not None:
-            #     im_arr = self._post_load_function(im_arr)
-            # add the new layer
-            self.viewer.add_image(im_arr, **im_kwargs)
+        return py_kwargs
 
 
 @thread_worker(progress=True)
-def time_series_load(model):
+def time_series_load(model):  # pragma: no cover
     _, layer_list = _model_ingestor._process_validated_model(model)
     return layer_list