Added helper function to clean whitespaces and newlines in text

functions/gradio.py

@@ -6,6 +6,7 @@ Functions for handling Gradio UI interactions and processing user inputs.
 
 import logging
 from pathlib import Path
+from functions.helper import clean_text_whitespace
 from functions.linkedin_resume import extract_text_from_linkedin_pdf
 from functions.github import get_github_repositories
 # from functions.job_call import summarize_job_call
@@ -55,7 +56,7 @@ def process_inputs(
     logger = logging.getLogger(f'{__name__}.process_inputs')
     logger.info("LinkedIn PDF: %s", linkedin_pdf_path)
     logger.info("GitHub URL: %s", github_url)
-    logger.info("Job post: %s", job_post_text[:100])
+    logger.info("Job post: %s", clean_text_whitespace(job_post_text[:100]).replace("\n", " "))
     logger.info("User instructions: %s", user_instructions[:100] if user_instructions else "None")
     result = ""
 

functions/helper.py

@@ -0,0 +1,33 @@
+"""
+helper.py
+
+Utility functions for text processing and data cleaning.
+"""
+
+import re
+
+
+def clean_text_whitespace(text: str) -> str:
+    """
+    Clean up text by normalizing whitespace and newlines.
+    
+    Args:
+        text (str): Input text string to clean
+        
+    Returns:
+        str: Cleaned text with normalized whitespace and newlines
+    """
+    if not text or not isinstance(text, str):
+        return text
+
+    # Replace multiple whitespace characters (spaces, tabs) with a single space
+    # This handles spaces, tabs, and other whitespace characters except newlines
+    text = re.sub(r'[^\S\n]+', ' ', text)
+
+    # Replace multiple consecutive newlines with a single newline
+    text = re.sub(r'\n{2,}', '\n', text)
+
+    # Strip leading and trailing whitespace
+    text = text.strip()
+
+    return text