merge functionality for submitr

dcicutils/misc_utils.py

@@ -4,6 +4,7 @@
 
 from collections import namedtuple
 import appdirs
+from copy import deepcopy
 import contextlib
 import datetime
 import functools
@@ -2199,28 +2200,58 @@ def merge_key_value_dict_lists(x, y):
     return [key_value_dict(k, v) for k, v in merged.items()]
 
 
-def merge_objects(target: Union[dict, List[Any]], source: Union[dict, List[Any]], full: bool = False) -> dict:
+def merge_objects(target: Union[dict, List[Any]], source: Union[dict, List[Any]],
+                  full: bool = False,  # deprecated
+                  expand_lists: Optional[bool] = None,
+                  primitive_lists: bool = False,
+                  copy: bool = False, _recursing: bool = False) -> Union[dict, List[Any]]:
     """
-    Merges the given source dictionary or list into the target dictionary or list.
-    This MAY well change the given target (dictionary or list) IN PLACE.
-    The the full argument is True then any target lists longer than the
-    source be will be filled out with the last element(s) of the source.
+    Merges the given source dictionary or list into the target dictionary or list and returns the
+    result. This MAY well change the given target (dictionary or list) IN PLACE ... UNLESS the copy
+    argument is True, then the given target will not change as a local copy is made (and returned).
+
+    If the expand_lists argument is True then any target lists longer than the
+    source be will be filled out with the last element(s) of the source; the full
+    argument (is deprecated and) is a synomym for this. The default is False.
+
+    If the primitive_lists argument is True then lists of primitives (i.e. lists in which
+    NONE of its elements are dictionaries, lists, or tuples) will themselves be treated
+    like primitives, meaning the whole of a source list will replace the corresponding
+    target; otherwise they  will be merged normally, meaning each element of a source list
+    will be merged, recursively, into the corresponding target list. The default is False.
     """
+    def is_primitive_list(value: Any) -> bool:  # noqa
+        if not isinstance(value, list):
+            return False
+        for item in value:
+            if isinstance(item, (dict, list, tuple)):
+                return False
+        return True
+
     if target is None:
         return source
+    if expand_lists not in (True, False):
+        expand_lists = full is True
+    if (copy is True) and (_recursing is not True):
+        target = deepcopy(target)
     if isinstance(target, dict) and isinstance(source, dict) and source:
         for key, value in source.items():
-            target[key] = merge_objects(target[key], value, full) if key in target else value
+            if ((primitive_lists is True) and
+                (key in target) and is_primitive_list(target[key]) and is_primitive_list(value)):  # noqa
+                target[key] = value
+            else:
+                target[key] = merge_objects(target[key], value,
+                                            expand_lists=expand_lists, _recursing=True) if key in target else value
     elif isinstance(target, list) and isinstance(source, list) and source:
         for i in range(max(len(source), len(target))):
             if i < len(target):
                 if i < len(source):
-                    target[i] = merge_objects(target[i], source[i], full)
-                elif full:
-                    target[i] = merge_objects(target[i], source[len(source) - 1], full)
+                    target[i] = merge_objects(target[i], source[i], expand_lists=expand_lists, _recursing=True)
+                elif expand_lists is True:
+                    target[i] = merge_objects(target[i], source[len(source) - 1], expand_lists=expand_lists)
             else:
                 target.append(source[i])
-    elif source:
+    elif source not in (None, {}, []):
         target = source
     return target
 

dcicutils/structured_data.py

@@ -74,7 +74,7 @@ def __init__(self, file: Optional[str] = None, portal: Optional[Union[VirtualApp
         self._nrows = 0
         self._autoadd_properties = autoadd if isinstance(autoadd, dict) and autoadd else None
         self._norefs = True if norefs is True else False
-        self._merge = True if merge is True else False
+        self._merge = True if merge is True else False  # New merge functionality (2024-05-25)
         self._debug_sleep = None
         if debug_sleep:
             try:
@@ -347,7 +347,7 @@ def _load_json_file(self, file: str) -> None:
                 (self._portal.get_schema(schema_name_inferred_from_file_name) is not None)):  # noqa
                 # If the JSON file name looks like a schema name then assume it
                 # contains an object or an array of object of that schema type.
-                if self._merge:
+                if self._merge:  # New merge functionality (2024-05-25)
                     data = self._merge_with_existing_portal_object(data, schema_name_inferred_from_file_name)
                 self._add(Schema.type_name(file), data)
             elif isinstance(data, dict):
@@ -356,7 +356,7 @@ def _load_json_file(self, file: str) -> None:
                 # which (each property) contains a list of object of that schema type.
                 for schema_name in data:
                     item = data[schema_name]
-                    if self._merge:
+                    if self._merge:  # New merge functionality (2024-05-25)
                         item = self._merge_with_existing_portal_object(item, schema_name)
                     self._add(schema_name, item)
 
@@ -380,8 +380,7 @@ def _load_reader(self, reader: RowReader, type_name: str) -> None:
                 structured_row_template.set_value(structured_row, column_name, value, reader.file, reader.row_number)
                 if self._autoadd_properties:
                     self._add_properties(structured_row, self._autoadd_properties, schema)
-            # New merge functionality (2024-05-25).
-            if self._merge:
+            if self._merge:  # New merge functionality (2024-05-25)
                 structured_row = self._merge_with_existing_portal_object(structured_row, schema_name)
             if (prune_error := self._prune_structured_row(structured_row)) is not None:
                 self._note_error({"src": create_dict(type=schema_name, row=reader.row_number),
@@ -437,7 +436,7 @@ def _merge_with_existing_portal_object(self, portal_object: dict, portal_type: s
         """
         for identifying_path in self._portal.get_identifying_paths(portal_object, portal_type):
             if existing_portal_object := self._portal.get_metadata(identifying_path, raw=True, raise_exception=False):
-                return merge_objects(existing_portal_object, portal_object)
+                return merge_objects(existing_portal_object, portal_object, primitive_lists=True)
         return portal_object
 
     def _is_ref_lookup_specified_type(ref_lookup_flags: int) -> bool:

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "dcicutils"
-version = "8.9.0.1b2"  # TODO: To become 8.10.0
+version = "8.9.0.1b3"  # TODO: To become 8.10.0
 description = "Utility package for interacting with the 4DN Data Portal and other 4DN resources"
 authors = ["4DN-DCIC Team <[email protected]>"]
 license = "MIT"