Add data without session corruption (#111)

* Add data without session corruption

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Add a test on session when adding data

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Move function to glue_utils

* Improving the readability of nested_compare

Co-authored-by: martinRenou <[email protected]>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: martinRenou <[email protected]>

Author: 3 people

gluepyter/glue_session.py

@@ -3,7 +3,6 @@
 import warnings
 from pathlib import Path
 from typing import TYPE_CHECKING, Dict, Optional, Tuple
-
 from glue.core.link_helpers import LinkSame
 from glue.core.state import GlueSerializer
 import glue_jupyter as gj
@@ -356,29 +355,40 @@ def add_data(self, file_path: str) -> None:
         assert os.path.exists(relative_path)
 
         data = self.app.load_data(str(relative_path))
-        self._data[data.label] = data
 
-        # We generate the data representation and merge it into our ycontent
-        serializer = GlueSerializer(data)
-        serialized_data = serializer.dumpo()
+        # Serialize the data on its own (without other context).
+        data_serializer = GlueSerializer(data)
+        data_serializer.dumpo()
+
+        # Serialize the previous data without the new one to create a context.
+        previous_data_serializer = GlueSerializer(self._data)
+        previous_data_serializer.dumpo()
+
+        # We generate the data representation in the previous data context.
+        serialized_data = [
+            (previous_data_serializer.id(obj), previous_data_serializer.do(obj))
+            for oid, obj in list(data_serializer._objs.items())
+        ]
+        serialized_data = dict(serialized_data)
 
         contents = self._document.contents
 
-        # Inject the main data repr
-        contents[data.label] = serialized_data["__main__"]
         # Inject all components
         for key, value in serialized_data.items():
             if key != "__main__":
                 contents[key] = value
+
         # Inject the label in the data collection
-        contents["DataCollection"]["data"].append(data.label)
-        contents["DataCollection"]["cids"].extend(
-            cid for cid, comp in serialized_data["__main__"]["components"]
+        data_collection_name: str = contents.get("__main__", {}).get("data", "")
+        contents[data_collection_name]["data"].append(data.label)
+        contents[data_collection_name]["cids"].extend(
+            cid for cid, comp in serialized_data[data.label]["components"]
         )
-        contents["DataCollection"]["components"].extend(
-            comp for cid, comp in serialized_data["__main__"]["components"]
+        contents[data_collection_name]["components"].extend(
+            comp for cid, comp in serialized_data[data.label]["components"]
         )
 
+        self._data[data.label] = data
         self._document.set(json.dumps(contents))
 
     def _load_data(self) -> None:

gluepyter/glue_utils.py

@@ -10,6 +10,17 @@
 load_plugins()
 
 
+class ErrorWidget:
+    """Wrapper of a HTML widget for showing error message"""
+
+    def __init__(self, e: Exception, path: str) -> None:
+        value = f"{type(e).__name__} at line {e.__traceback__.tb_lineno} of {path}: {e}"
+        self._widget = HTML(value=value)
+
+    def show(self):
+        display(self._widget)
+
+
 def get_function_info(function_or_helper):
     item_info: Dict[str, Dict] = {}
     attributes = ["description", "labels1", "labels2", "display"]
@@ -67,12 +78,28 @@ def get_advanced_links():
     return advanced_links
 
 
-class ErrorWidget:
-    """Wrapper of a HTML widget for showing error message"""
+def nested_compare(value1, value2):
+    # Compare lists
+    if isinstance(value1, list) and isinstance(value2, list):
+        if not len(value1) == len(value2):
+            return False
 
-    def __init__(self, e: Exception, path: str) -> None:
-        value = f"{type(e).__name__} at line {e.__traceback__.tb_lineno} of {path}: {e}"
-        self._widget = HTML(value=value)
+        for v1, v2 in zip(value1, value2):
+            if not nested_compare(v1, v2):
+                return False
 
-    def show(self):
-        display(self._widget)
+        return True
+
+    # Compare dict
+    if isinstance(value1, dict) and isinstance(value2, dict):
+        for k1, v1 in value1.items():
+            if k1 not in value2.keys():
+                return False
+
+            if not nested_compare(v1, value2[k1]):
+                return False
+
+        return True
+
+    # Compare immutable
+    return value1 == value2

gluepyter/tests/test_glue_session.py

@@ -1,6 +1,9 @@
 import y_py as Y
-from gluepyter.glue_session import SharedGlueSession
+from copy import deepcopy
+from pathlib import Path
 from ipywidgets import Output
+from gluepyter.glue_session import SharedGlueSession
+from gluepyter.glue_utils import nested_compare
 
 
 def test_init(session_path):
@@ -61,6 +64,33 @@ def test__read_view_state(yglue_session):
     assert len(state) > 0
 
 
+def test_add_data(yglue_session):
+    yglue_session._load_data()
+    file_path = Path(__file__).parents[2] / "examples" / "w6_psc.vot"
+
+    contents = deepcopy(yglue_session._document.contents)
+    yglue_session.add_data(file_path)
+    updated_contents = yglue_session._document.contents
+
+    assert "w6_psc" in updated_contents.keys()
+
+    # Assert there is no change in previous structure
+    for key, value in contents.items():
+        if key == "DataCollection":
+            continue
+        assert key in updated_contents.keys()
+        assert nested_compare(value, updated_contents[key])
+
+    # Compare the DataCollection
+    for key, value in contents["DataCollection"].items():
+        if key == "data" or key == "cids" or key == "components":
+            assert not nested_compare(value, updated_contents["DataCollection"][key])
+        else:
+            assert nested_compare(value, updated_contents["DataCollection"][key])
+
+    assert "w6_psc" in updated_contents["DataCollection"]["data"]
+
+
 def test_add_identity_link(yglue_session, identity_link):
     yglue_session._load_data()
     change = {"LinkTest": {"action": "add", "newValue": identity_link}}