Bug fix

* Add `columns_check_with_data`

pydbsmgr/VERSION

@@ -1 +1 @@
-0.9.2
+0.9.3

pydbsmgr/utils/tools/tools.py

@@ -90,8 +90,8 @@ class ColumnsCheck:
     def __init__(self, df: DataFrame):
         self.df = df
 
-    def get_frame(self) -> DataFrame:
-        self.df = self._process_columns()
+    def get_frame(self, **kwargs) -> DataFrame:
+        self.df = self._process_columns(**kwargs)
         return self.df
 
     def _process_columns(self, surrounding: bool = True) -> DataFrame:
@@ -102,7 +102,8 @@ def _process_columns(self, surrounding: bool = True) -> DataFrame:
         df.columns = df.columns.str.replace(r"[^a-zA-Z0-9ñáéíóú_]", "_", regex=True)
 
         df.columns = df.columns.str.replace("_+", "_", regex=True)
-        df.columns = df.columns.str.strip().strip("_")
+        df.columns = df.columns.str.strip()
+        df.columns = df.columns.str.strip("_")
         if surrounding:
             df.columns = [f"[{col}]" for col in df.columns]
 

test/conftest.py

@@ -5,7 +5,7 @@
 
 from pydbsmgr.lightest import LightCleaner
 from pydbsmgr.main import *
-from pydbsmgr.utils.tools import ColumnsDtypes, get_extraction_date
+from pydbsmgr.utils.tools import ColumnsCheck, ColumnsDtypes, get_extraction_date
 
 
 @pytest.fixture()
@@ -80,3 +80,24 @@ def lightest_with_data() -> Callable:
         df["another_date"].astype(str).to_list(),
         df["third_date"].astype(str).to_list(),
     )
+
+
+@pytest.fixture()
+def columns_check_with_data() -> Callable:
+    """
+    Fixture that returns an instance of the class ColumnsCheck with data for testing
+    """
+    df = pd.DataFrame(
+        {
+            "index": ["0", "1", "2", "3"],
+            "Raw Data!": ["10", "06", "18", "25"],
+            "First Data #": ["09", "01", "01", "08"],
+            "First Data 2": ["9", "1", "1", "8"],
+            "Another Data?": ["10", "6", "18", "25"],
+        }
+    )
+
+    handler = ColumnsCheck(df)
+    df = handler.get_frame(surrounding=False)
+
+    return df.columns.to_list()

test/test_functions.py

@@ -61,6 +61,15 @@ def test_lightest(lightest_with_data):
     assert third_date == comparison
 
 
+def test_columnscheck(columns_check_with_data):
+    cols = columns_check_with_data
+    assert cols[0] == "index"
+    assert cols[1] == "raw_data"
+    assert cols[2] == "first_data"
+    assert cols[3] == "first_data_2"
+    assert cols[4] == "another_data"
+
+
 def test_get_extraction_date(_get_extraction_date):
     assert _get_extraction_date("filename_2023-11-17") == "2023-11-17"
     assert _get_extraction_date(["filename_2023-11-17"]) == ["2023-11-17"]