TST: Add test for https://gitlab.com/linkahead/linkahead-advanced-user-tools/-/issues/63

unittests/test_table_importer.py

@@ -379,6 +379,48 @@ class CSVImporterTest(TableImporterTest):
         assert df["int_with_gaps"].dtype == "Int64"
         assert df["float"].dtype == float
 
+    def test_wrong_datatype_type_errors(self):
+        """Test for
+        https://gitlab.com/linkahead/linkahead-advanced-user-tools/-/issues/63:
+        Highlight rows and columns in which type errors occur.
+
+        """
+        tmpfile = NamedTemporaryFile(delete=False, suffix=".csv")
+        with open(tmpfile.name, 'w') as tmp:
+            # Wrong types in row 2, columns 1 and 2, and row 4, column 2.
+            tmp.write(
+                "int,float\n"
+                "1,2.3\n"
+                "4.5,word\n"
+                "0,1.2\n"
+                "-12,12+3j\n"
+            )
+        kwargs = {
+            "datatypes": {
+                "int": int,
+                "float": float
+            },
+            "obligatory_columns": ["int"],
+            "converters": {}
+        }
+        importer = CSVImporter(**kwargs)
+        with pytest.raises(DataInconsistencyError) as die:
+            df = importer.read_file(tmpfile.name)
+        msg = str(die.value)
+        assert "Elements with wrong datatypes encountered:\n" in msg
+        # Errors in rows 1 and 3, no errors in 2 and 4
+        assert "* row 1:\n" in msg
+        assert "* row 2:\n" not in msg
+        assert "* row 3:\n" in msg
+        assert "* row 4:\n" not in msg
+        row_1_msgs, row_3_msgs = msg.split("* row 1:\n")[1].split("* row 3:\n")
+        # exactly 2 errors in row 1, exactly 1 in row 3
+        assert len(row_1_msgs.strip().split('\n')) == 2
+        assert len(row_3_msgs.strip().split('\n')) == 1
+        assert "\t* column \"int\"" in row_1_msgs
+        assert "\t* column \"float\"" in row_1_msgs
+        assert "\t* column \"float\"" in row_3_msgs
+
 
 class TSVImporterTest(TableImporterTest):
     def test_full(self):