fix(xlsform): do not filter duplicates from choices sheet

hotosm · Sep 24, 2024 · dfd60a7 · dfd60a7
1 parent ac8c71e
commit dfd60a7
Show file tree

Hide file tree

Showing 2 changed files with 8 additions and 18 deletions.
diff --git a/osm_fieldwork/update_xlsform.py b/osm_fieldwork/update_xlsform.py
@@ -34,26 +34,24 @@ def merge_dataframes(mandatory_df: pd.DataFrame, user_question_df: pd.DataFrame,
     #     mandatory_df, user_question_df, digitisation_df, fields=["label", "hint", "required_message"]
     # )
 
-    # Find common fields between user_question_df and mandatory_df or digitisation_df
-    duplicate_fields = set(user_question_df[NAME_COLUMN]).intersection(
-        set(mandatory_df[NAME_COLUMN]).union(set(digitisation_df[NAME_COLUMN]))
-    )
-
-    # Is choices sheet, return ordered merged choices
+    # If processing the choices sheet, retain all duplicates
     if "list_name" in user_question_df.columns:
-        user_question_df_filtered = user_question_df[~user_question_df[NAME_COLUMN].isin(duplicate_fields)]
-
         return pd.concat(
             [
                 mandatory_df,
-                user_question_df_filtered,
+                user_question_df,
                 digitisation_df,
             ],
             ignore_index=True,
         )
 
     # Else we are processing the survey sheet, continue
 
+    # Find common fields between user_question_df and mandatory_df or digitisation_df
+    duplicate_fields = set(user_question_df[NAME_COLUMN]).intersection(
+        set(mandatory_df[NAME_COLUMN]).union(set(digitisation_df[NAME_COLUMN]))
+    )
+
     # NOTE filter out 'end group' from duplicate check as they have empty NAME_COLUMN
     end_group_rows = user_question_df[user_question_df["type"].isin(["end group", "end_group"])]
     user_question_df_filtered = user_question_df[

diff --git a/tests/test_update_xlsform.py b/tests/test_update_xlsform.py
@@ -41,7 +41,7 @@ async def test_merge_mandatory_fields():
         merged_xlsform.write(updated_form.getvalue())
 
     check_survey_sheet(workbook)
-    check_choices_sheet(workbook)
+    # NOTE the choices sheet can have duplicates in the 'name' field without issue
     check_entities_sheet(workbook)
     check_form_title(workbook)
 
@@ -113,14 +113,6 @@ def check_survey_sheet(workbook: Workbook) -> None:
     check_for_duplicates(survey_sheet, name_col_index)
 
 
-def check_choices_sheet(workbook: Workbook) -> None:
-    """Check the 'choices' sheet and ensure no duplicates in 'name' column."""
-    choices_sheet = get_sheet(workbook, "choices")
-    name_col_index = get_column_index(choices_sheet, "name")
-
-    check_for_duplicates(choices_sheet, name_col_index)
-
-
 def check_entities_sheet(workbook: Workbook) -> None:
     """Check the 'entities' sheet values."""
     entities_sheet = get_sheet(workbook, "entities")