Upload 64 files

.github/workflows/sync-to-hf.yml

@@ -0,0 +1,25 @@
+name: Sync to Hugging Face Space
+
+on:
+  push:
+    branches:
+      - main
+
+jobs:
+  sync-to-huggingface:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v3
+        with:
+          fetch-depth: 0
+          lfs: true
+
+      - name: Push to Hugging Face Space
+        env:
+          HF_TOKEN: ${{ secrets.HF_TOKEN }}
+        run: |
+          git config --global user.email "github-actions[bot]@users.noreply.github.com"
+          git config --global user.name "github-actions[bot]"
+          git remote add hf https://HF_USERNAME:$HF_TOKEN@huggingface.co/spaces/Anmol4521/jansahayak
+          git push hf main --force

app.py

@@ -33,28 +33,6 @@ VECTORSTORES_INITIALIZED = False
 SKIP_VECTORSTORES = os.environ.get('SKIP_VECTORSTORES', 'false').lower() == 'true'
 
 
-def load_saved_result(session_id: str):
-    """Load a previously saved result from disk if the in-memory session is unavailable."""
-    if not os.path.exists('outputs'):
-        return None
-
-    for filename in os.listdir('outputs'):
-        if not filename.endswith('.json'):
-            continue
-
-        filepath = os.path.join('outputs', filename)
-        try:
-            with open(filepath, 'r', encoding='utf-8') as f:
-                data = json.load(f)
-
-            if data.get('session_id') == session_id or session_id in filename:
-                return data
-        except Exception:
-            continue
-
-    return None
-
-
 def initialize_vectorstores():
     """Load vectorstores lazily on first use to avoid blocking port binding"""
     global SCHEME_VECTORSTORE, EXAM_VECTORSTORE, VECTORSTORES_INITIALIZED
@@ -244,9 +222,6 @@ def analyze():
         # Ensure user_profile key exists in result
         if 'user_profile' not in result and 'profile' in result:
             result['user_profile'] = result['profile']
-
-        result['session_id'] = session_id
-        result['generated_at'] = datetime.now().isoformat()
         
         # Update session
         sessions[session_id]['status'] = 'completed'
@@ -255,7 +230,7 @@ def analyze():
         
         # Save to file
         timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
-        filename = f"outputs/results_{session_id}.json"
+        filename = f"outputs/results_{timestamp}.json"
         os.makedirs('outputs', exist_ok=True)
         
         with open(filename, 'w', encoding='utf-8') as f:
@@ -294,16 +269,8 @@ def analyze():
 def result(session_id):
     """Display results page"""
     if session_id not in sessions:
-        saved_result = load_saved_result(session_id)
-        if saved_result:
-            sessions[session_id] = {
-                'status': 'completed',
-                'result': saved_result,
-                'completed_at': saved_result.get('generated_at', datetime.now().isoformat())
-            }
-        else:
-            return render_template('error.html', 
-                                 error='Session not found'), 404
+        return render_template('error.html', 
+                             error='Session not found'), 404
     
     session_data = sessions[session_id]
     
@@ -341,8 +308,7 @@ def history():
                 data = json.load(f)
                 output_files.append({
                     'filename': filename,
-                    'timestamp': data.get('generated_at', filename.replace('results_', '').replace('.json', '')),
-                    'session_id': data.get('session_id', filename.replace('results_', '').replace('.json', '')),
+                    'timestamp': filename.replace('results_', '').replace('.json', ''),
                     'profile': data.get('user_profile', {}),
                     'errors': data.get('errors', [])
                 })
@@ -367,14 +333,7 @@ def download_pdf(session_id):
     """Generate and download PDF report"""
     try:
         if session_id not in sessions:
-            saved_result = load_saved_result(session_id)
-            if not saved_result:
-                return jsonify({'error': 'Session not found'}), 404
-            sessions[session_id] = {
-                'status': 'completed',
-                'result': saved_result,
-                'completed_at': saved_result.get('generated_at', datetime.now().isoformat())
-            }
+            return jsonify({'error': 'Session not found'}), 404
         
         session_data = sessions[session_id]
         result = session_data.get('result', {})

graph/workflow.py

@@ -114,8 +114,22 @@ def scheme_node(state: AgentState) -> dict:
             print(f"✅ Profile has {len(useful_fields)} useful fields")
         
         result = run_scheme_agent(profile, use_web_search=True, vectorstore=scheme_vectorstore)
+        
+        if "error" in result:
+            print(f"❌ Scheme Agent returned Error: {result['error']}")
+            return {
+                "scheme_recommendations": f"Error generating recommendations: {result['error']}",
+                "errors": [f"Scheme: {result['error']}"]
+            }
+            
         print("✅ Scheme recommendations generated")
-        return {"scheme_recommendations": result.get("recommendations", "")}
+        
+        # Ensure it's a string, not a list on failure
+        recs = result.get("recommendations", "")
+        if isinstance(recs, list):
+            recs = " ".join(recs) if recs else "Failed to generate recommendations"
+            
+        return {"scheme_recommendations": recs}
         
     except Exception as e:
         print(f"❌ Scheme Agent Error: {str(e)}")
@@ -151,8 +165,22 @@ def exam_node(state: AgentState) -> dict:
             # Still try with whatever we have
         
         result = run_exam_agent(profile, use_web_search=True, vectorstore=exam_vectorstore)
+        
+        if "error" in result:
+            print(f"❌ Exam Agent returned Error: {result['error']}")
+            return {
+                "exam_recommendations": f"Error generating recommendations: {result['error']}",
+                "errors": [f"Exam: {result['error']}"]
+            }
+            
         print("✅ Exam recommendations generated")
-        return {"exam_recommendations": result.get("recommendations", "")}
+        
+        # Ensure it's a string
+        recs = result.get("recommendations", "")
+        if isinstance(recs, list):
+            recs = " ".join(recs) if recs else "Failed to generate recommendations"
+            
+        return {"exam_recommendations": recs}
         
     except Exception as e:
         print(f"❌ Exam Agent Error: {str(e)}")
@@ -174,18 +202,31 @@ def benefit_node(state: AgentState) -> dict:
         profile = state.get("profile", {})
         scheme_recommendations = state.get("scheme_recommendations", "")
         
-        if not profile or not scheme_recommendations:
+        if not profile or not scheme_recommendations or scheme_recommendations == "[]":
             print("⚠️ Insufficient data for benefit calculation")
             return {"missed_benefits": "Insufficient data"}
         
         result = calculate_missed_benefits(profile, scheme_recommendations)
+        
+        if "error" in result:
+            print(f"❌ Benefit Agent returned Error: {result['error']}")
+            return {
+                "missed_benefits": f"Error calculating benefits: {result['error']}",
+                "errors": [f"Benefit: {result['error']}"]
+            }
+            
         print("✅ Benefit calculation completed")
-        return {"missed_benefits": result.get("calculation", "")}
+        
+        calc = result.get("calculation", "")
+        if not calc:
+            calc = "Benefit calculation returned no result."
+            
+        return {"missed_benefits": calc}
         
     except Exception as e:
-        print(f"❌ Benefit Agent Error: {str(e)}")
+        print(f"❌ Benefit Agent Exception: {str(e)}")
         return {
-            "missed_benefits": "",
+            "missed_benefits": f"Error calculating benefits: {str(e)}",
             "errors": [f"Benefit: {str(e)}"]
         }
 

start_web.sh

@@ -1,21 +1,21 @@
-#!/bin/bash
-
-echo "================================"
-echo "JanSahayak - Starting Web Server"
-echo "================================"
-echo
-
-# Check if virtual environment exists
-if [ ! -d ".venv" ]; then
-    echo "Virtual environment not found!"
-    echo "Please run setup first."
-    exit 1
-fi
-
-# Activate virtual environment
-source .venv/bin/activate
-
-# Start Flask app
-echo "Starting Flask application..."
-echo
-python app.py
+#!/bin/bash
+
+echo "================================"
+echo "JanSahayak - Starting Web Server"
+echo "================================"
+echo
+
+# Check if virtual environment exists
+if [ ! -d ".venv" ]; then
+    echo "Virtual environment not found!"
+    echo "Please run setup first."
+    exit 1
+fi
+
+# Activate virtual environment
+source .venv/bin/activate
+
+# Start Flask app
+echo "Starting Flask application..."
+echo
+python app.py

test_pipeline.py

@@ -0,0 +1,32 @@
+import os
+import sys
+
+sys.path.append(os.path.dirname(os.path.abspath(__file__)))
+
+from graph.workflow import build_workflow
+
+def test():
+    app = build_workflow()
+    
+    initial_state = {
+        "user_input": "I am a 25 year old farmer with 2 acres of land",
+        "user_interests": ["schemes", "exams"],
+        "profile": {},
+        "scheme_vectorstore": None,
+        "exam_vectorstore": None,
+        "errors": []
+    }
+    
+    try:
+        # We need GROQ_API_KEY for this to work
+        os.environ["GROQ_API_KEY"] = "gsk_dummy"  # But this will fail authentication
+        # So instead we mock the LLM calls to see the graph flow.
+        
+        # Actually let's just print the graph flow if possible?
+        # Or better yet, we can see if scheme_recommendations is not being populated correctly.
+        pass
+    except Exception as e:
+        print(f"Error: {e}")
+
+if __name__ == "__main__":
+    test()