Update app.py

app.py

@@ -4,6 +4,7 @@ import PyPDF2
 import docx
 import requests
 import json
+import re  # Added for regex pattern replacement
 
 # Function to extract text from PDF
 def extract_text_from_pdf(file):
@@ -29,6 +30,24 @@ def process_uploaded_file(file):
     else:
         raise ValueError("Unsupported file format. Please upload a PDF or Word document.")
 
+# Function to clean JSON string with escaped backslashes
+def clean_json_string(json_str):
+    # Replace escaped backslashes with a temporary marker
+    temp_str = json_str.replace('\\_', '__UNDERSCORE__')
+    
+    # Attempt to fix any other common escape sequence issues
+    temp_str = temp_str.replace('\\n', '\n')
+    temp_str = temp_str.replace('\\t', '\t')
+    temp_str = temp_str.replace('\\r', '\r')
+    
+    # Remove any remaining unmatched backslashes
+    temp_str = temp_str.replace('\\', '')
+    
+    # Restore underscores
+    cleaned_str = temp_str.replace('__UNDERSCORE__', '_')
+    
+    return cleaned_str
+
 # Function to call Together API
 def analyze_with_mistral(resume_text, job_description):
     TOGETHER_API_KEY = os.getenv("HUGGINGFACE_API_KEY")
@@ -50,7 +69,10 @@ Follow these guidelines when analyzing resumes:
 6. Provide specific, actionable recommendations
 7. Be thorough but concise in your analysis
 
-Your analysis must be comprehensive yet practical for the candidate to implement."""
+Your analysis must be comprehensive yet practical for the candidate to implement.
+
+IMPORTANT: Do not use escaped backslashes in your JSON response.
+"""
         },
         {
             "role": "user", 
@@ -63,7 +85,7 @@ RESUME:
 JOB DESCRIPTION:
 {job_description}
 
-Provide a detailed analysis in EXACTLY the following JSON format:
+Provide a detailed analysis in EXACTLY the following JSON format (do not escape underscores with backslashes):
 
 {{
     "ATS_Compatibility": {{
@@ -110,7 +132,7 @@ Provide a detailed analysis in EXACTLY the following JSON format:
     }}
 }}
 
-Ensure all recommendations are specific, actionable, and tailored to both the resume content and job description.
+Do not escape underscores with backslashes in your response. Ensure all recommendations are specific, actionable, and tailored to both the resume content and job description.
 """
         }
     ]
@@ -135,16 +157,37 @@ Ensure all recommendations are specific, actionable, and tailored to both the re
         result = response.json()
         content = result.get("choices", [{}])[0].get("message", {}).get("content", "{}")
         
-        # The key fix: Safely parse the JSON content
+        # Clean the JSON string before parsing
+        cleaned_content = clean_json_string(content)
+        
         try:
-            parsed = json.loads(content)
+            parsed = json.loads(cleaned_content)
             # Check if the JSON structure is valid for our use case
-            # Note: We've updated this validation to match our new JSON structure
             if "ATS_Compatibility" not in parsed or "Overall_Assessment" not in parsed:
-                return {"error": "API returned unexpected JSON structure"}
+                return {
+                    "error": "API returned unexpected JSON structure",
+                    "raw_content": cleaned_content[:500] + "..." if len(cleaned_content) > 500 else cleaned_content
+                }
             return parsed
         except json.JSONDecodeError as e:
-            return {"error": f"Failed to parse API response: {str(e)}", "raw_content": content[:500] + "..." if len(content) > 500 else content}
+            # Try a more aggressive approach to fix the JSON
+            try:
+                # Sometimes the model might include trailing characters
+                # Try to find the closing bracket of the main JSON object
+                match = re.search(r'(\{.*\})', cleaned_content, re.DOTALL)
+                if match:
+                    extracted_json = match.group(1)
+                    parsed = json.loads(extracted_json)
+                    if "ATS_Compatibility" in parsed and "Overall_Assessment" in parsed:
+                        return parsed
+            except:
+                pass
+                
+            # If all attempts fail, return the error
+            return {
+                "error": f"Failed to parse API response: {str(e)}",
+                "raw_content": cleaned_content[:500] + "..." if len(cleaned_content) > 500 else cleaned_content
+            }
         
     except requests.exceptions.RequestException as e:
         return {"error": f"API request failed: {str(e)}"}
@@ -163,7 +206,7 @@ def analyze_resume(file, job_description):
         result = analyze_with_mistral(resume_text, job_description)
         
         # Return as formatted JSON string for better display
-        return json.dumps(result, indent=2)
+        return json.dumps(result, indent=2, ensure_ascii=False)
         
     except Exception as e:
         return json.dumps({"error": f"Error analyzing resume: {str(e)}"}, indent=2)