Update clean text function

src/utils.py

@@ -8,7 +8,7 @@ from PIL import Image
 def clean_text(text):
     """
     Cleans the given text by removing unwanted tokens, extra spaces,
-    and ensures proper spacing between words and after periods.
+    and ensures proper spacing between words and after punctuation marks.
 
     Args:
         text (str): The input text to be cleaned.
@@ -27,13 +27,19 @@ def clean_text(text):
     # Join the cleaned lines into a single string with a space between each line
     cleaned_text = " ".join(cleaned_lines)
 
-    # Ensure proper spacing between words and after periods using regex
+    # Ensure proper spacing using regex
     cleaned_text = re.sub(
         r"\s+", " ", cleaned_text
     )  # Replace multiple spaces with a single space
     cleaned_text = re.sub(
-        r"(?<=[.])(?=[^\s])", r" ", cleaned_text
-    )  # Add space after a period if not followed by a space
+        r"(?<=[.,!?])(?=[^\s])", r" ", cleaned_text
+    )  # Add space after punctuation if not followed by a space
+    cleaned_text = re.sub(
+        r"(?<=[a-z])(?=[A-Z])", r" ", cleaned_text
+    )  # Add space between joined words where a lowercase letter is followed by an uppercase letter
+    cleaned_text = re.sub(
+        r"(\w)([A-Z][a-z])", r"\1 \2", cleaned_text
+    )  # Add space between camel case words
 
     # Return the cleaned text
     return cleaned_text