reverse preprocess_text_lowercase_but_retain_punctuation

Author: sundy1994

src/team_comm_tools/feature_builder.py

@@ -571,7 +571,7 @@ def preprocess_chat_data(self) -> None:
         self.chat_data[self.message_col + "_original"] = self.chat_data[self.message_col]
 
         # create new column that retains punctuation
-        self.chat_data["message_lower_with_punc"] = self.chat_data[self.message_col].astype(str).apply(lambda x: x.lower())
+        self.chat_data["message_lower_with_punc"] = self.chat_data[self.message_col].astype(str).apply(preprocess_text_lowercase_but_retain_punctuation)
 
         # Preprocessing the text in `message_col` and then overwriting the column `message_col`.
         # TODO: We should probably use classes to abstract preprocessing module as well?

src/team_comm_tools/utils/preprocess.py

@@ -119,18 +119,18 @@ def remove_unhashable_cols(df: pd.DataFrame, column_names: dict) -> pd.DataFrame
         df = df.drop(columns=removable_cols)
     return df
 
-# def preprocess_text_lowercase_but_retain_punctuation(text): #NOTE: This function is deprecated 
-#     """Convert the input text to lowercase while retaining punctuation.
-
-#     This function takes a string and converts all characters to lowercase,
-#     keeping any punctuation marks intact.
-
-#     :param text: The input text to process.
-#     :type text: str
-#     :return: The processed text with all characters in lowercase.
-#     :rtype: str
-#     """
-#     return(text.lower())
+def preprocess_text_lowercase_but_retain_punctuation(text):
+    """Convert the input text to lowercase while retaining punctuation.
+
+    This function takes a string and converts all characters to lowercase,
+    keeping any punctuation marks intact.
+
+    :param text: The input text to process.
+    :type text: str
+    :return: The processed text with all characters in lowercase.
+    :rtype: str
+    """
+    return(text.lower())
 
 def preprocess_text(text: str) -> str:
     """Preprocess text by removing non-alphanumeric characters and converting to lowercase.