Added text cleanup

functions/gradio.py

@@ -60,29 +60,29 @@ def process_inputs(
     logger.info("User instructions: %s", user_instructions[:100] if user_instructions else "None")
     result = ""
 
-    # ==================================================================== #
-    # Extract and structure text from the linkedin profile PDF
-    logger.info("Extracting text from LinkedIn PDF: %s", linkedin_pdf_path)
-    linkedin_resume = extract_text(linkedin_pdf_path)
+    # # ==================================================================== #
+    # # Extract and structure text from the linkedin profile PDF
+    # logger.info("Extracting text from LinkedIn PDF: %s", linkedin_pdf_path)
+    # linkedin_resume = extract_text(linkedin_pdf_path)
 
-    if linkedin_resume:
-        logger.info("LinkedIn PDF text extraction successful")
+    # if linkedin_resume:
+    #     logger.info("LinkedIn PDF text extraction successful")
 
-    else:
-        logger.error("LinkedIn PDF text extraction failed")
+    # else:
+    #     logger.error("LinkedIn PDF text extraction failed")
 
-    # ==================================================================== #
-    # Process GitHub profile
-    logger.info("Processing GitHub profile: %s", github_username.strip())
+    # # ==================================================================== #
+    # # Process GitHub profile
+    # logger.info("Processing GitHub profile: %s", github_username.strip())
 
-    # Retrieve repositories from GitHub
-    github_repositories = get_github_repositories(github_username.strip())
+    # # Retrieve repositories from GitHub
+    # github_repositories = get_github_repositories(github_username.strip())
 
-    if github_repositories:
-        logger.info("GitHub repositories retrieved successfully")
+    # if github_repositories:
+    #     logger.info("GitHub repositories retrieved successfully")
 
-    else:
-        logger.error("GitHub repositories retrieval failed")
+    # else:
+    #     logger.error("GitHub repositories retrieval failed")
 
     # ==================================================================== #
     # Process job post text

functions/job_call.py

@@ -3,6 +3,7 @@
 import os
 import json
 import logging
+import unicodedata
 from pathlib import Path
 from datetime import datetime
 from openai import OpenAI
@@ -12,6 +13,8 @@ from configuration import (
     JOB_CALL_EXTRACTION_PROMPT
 )
 
+from functions.helper import clean_text_whitespace
+
 # pylint: disable=broad-exception-caught
 
 
@@ -27,6 +30,9 @@ def summarize_job_call(job_call: str) -> str:
 
     logger = logging.getLogger(f'{__name__}.summarize_job_call')
 
+    # Clean up the job call text
+    job_call = unicodedata.normalize('NFKC', job_call)
+    job_call = clean_text_whitespace(job_call)
 
     client = OpenAI(
         base_url=INFERENCE_URL,
@@ -49,8 +55,6 @@ def summarize_job_call(job_call: str) -> str:
         'messages': messages,
     }
 
-    print(completion_args)
-
     try:
         response = client.chat.completions.create(**completion_args)
 
@@ -63,6 +67,7 @@ def summarize_job_call(job_call: str) -> str:
 
         try:
             summary = json.loads(summary)
+            print(summary.keys())
 
         except json.JSONDecodeError as e:
             logger.error("Failed to parse job call summary JSON: %s", e)