feat: improve upload validation, add loading spinner, implement user authentication

.gitignore

@@ -22,6 +22,9 @@ temp.jpg
 # Streamlit cache
 .streamlit/secrets.toml
 
+# Auth credentials (generated at runtime)
+code/auth.yaml
+
 # OS files
 .DS_Store
 Thumbs.db

code/app.py

@@ -18,13 +18,81 @@
 from report_generator import generate_pdf_report
 from src.image_analyzer import analyze_image
 
+import yaml
+from dotenv import load_dotenv
+
+load_dotenv()
+
+try:
+    import streamlit_authenticator as stauth
+except ImportError:
+    stauth = None
+
 
 APP_TITLE = "ClaimVision AI"
 APP_SUBTITLE = "AI-Powered Damage Claim Verification"
 REPO_ROOT = Path(__file__).resolve().parents[1]
 LOGO_PATH = REPO_ROOT / "assets" / "logo.png"
 TEMP_IMAGE_PATH = Path("temp.jpg")
 OUTPUT_CSV_PATH = REPO_ROOT / "dataset" / "output.csv"
+MAX_FILE_SIZE = 25 * 1024 * 1024
+ALLOWED_EXTENSIONS = {".jpg", ".jpeg", ".png"}
+AUTH_CONFIG_PATH = Path(__file__).parent / "auth.yaml"
+
+
+def ensure_auth_config():
+    if not AUTH_CONFIG_PATH.exists():
+        os.makedirs(AUTH_CONFIG_PATH.parent, exist_ok=True)
+        cookie_key = os.getenv("CLAIMVISION_AUTH_KEY", "claimvision_secret_key_change_me_in_production")
+        default = {
+            "credentials": {
+                "usernames": {
+                    "admin": {
+                        "email": "admin@claimvision.ai",
+                        "name": "Admin",
+                        "password": "$2b$12$rGtJKo3o4t5EtOPW9LTbKu34HLnk5F5EJjy/RPv/BWT6eRmGMDgAG",
+                    }
+                }
+            },
+            "cookie": {
+                "name": "claimvision_auth",
+                "key": cookie_key,
+                "expiry_days": 30,
+            },
+        }
+        with open(AUTH_CONFIG_PATH, "w") as f:
+            yaml.dump(default, f, default_flow_style=False)
+    return AUTH_CONFIG_PATH
+
+
+def load_auth_config():
+    ensure_auth_config()
+    with open(AUTH_CONFIG_PATH) as f:
+        return yaml.safe_load(f)
+
+
+def save_auth_config(config):
+    with open(AUTH_CONFIG_PATH, "w") as f:
+        yaml.dump(config, f, default_flow_style=False)
+
+
+def validate_uploaded_file(uploaded_file):
+    errors = []
+    if uploaded_file is None:
+        errors.append("No file was uploaded.")
+        return errors
+
+    ext = Path(uploaded_file.name).suffix.lower()
+    if ext not in ALLOWED_EXTENSIONS:
+        errors.append(f"Unsupported file type '{ext}'. Only JPG, JPEG, and PNG images are allowed.")
+
+    file_size = len(uploaded_file.getbuffer())
+    if file_size == 0:
+        errors.append("The uploaded file is empty.")
+    elif file_size > MAX_FILE_SIZE:
+        errors.append(f"File size exceeds {MAX_FILE_SIZE // (1024 * 1024)} MB.")
+
+    return errors
 
 
 def get_logo_image():
@@ -3242,6 +3310,9 @@ def create_pdf_report(
 
 
 def save_uploaded_image(uploaded_file):
+    errors = validate_uploaded_file(uploaded_file)
+    if errors:
+        raise ValueError("; ".join(errors))
     TEMP_IMAGE_PATH.write_bytes(uploaded_file.getbuffer())
     return str(TEMP_IMAGE_PATH)
 
@@ -3251,43 +3322,102 @@ def main():
     inject_css()
     render_chatbot_widget()
 
-    if "analysis_result" not in st.session_state:
-        st.session_state["analysis_result"] = None
-    if "analysis_completed" not in st.session_state:
-        st.session_state["analysis_completed"] = False
-    if "pdf_report_path" not in st.session_state:
-        st.session_state["pdf_report_path"] = None
-    if "report_error" not in st.session_state:
-        st.session_state["report_error"] = None
-    if "report_success" not in st.session_state:
-        st.session_state["report_success"] = False
-    if "confidence_score" not in st.session_state:
-        st.session_state["confidence_score"] = None
-    if "risk_score" not in st.session_state:
-        st.session_state["risk_score"] = None
-    if "risk_level" not in st.session_state:
-        st.session_state["risk_level"] = None
-    if "risk_reasons" not in st.session_state:
-        st.session_state["risk_reasons"] = []
-    if "explanation" not in st.session_state:
-        st.session_state["explanation"] = None
-    if "cost_estimate" not in st.session_state:
-        st.session_state["cost_estimate"] = None
-    if "claim_history" not in st.session_state:
-        st.session_state["claim_history"] = None
-    if "uploaded_images" not in st.session_state:
-        st.session_state["uploaded_images"] = []
-    if "latest_claim" not in st.session_state:
-        st.session_state["latest_claim"] = None
-    if "analysis_stage" not in st.session_state:
-        st.session_state["analysis_stage"] = ""
-    if "analysis_progress" not in st.session_state:
-        st.session_state["analysis_progress"] = 0
-    if "analysis_status" not in st.session_state:
-        st.session_state["analysis_status"] = "Idle"
+    # --- AUTHENTICATION ---
+    if stauth is None:
+        st.error(
+            "**streamlit-authenticator** is required. "
+            "Install it with: `pip install streamlit-authenticator`"
+        )
+        return
+
+    if "auth_mode" not in st.session_state:
+        st.session_state["auth_mode"] = "login"
+
+    config = load_auth_config()
+    authenticator = stauth.Authenticate(
+        config["credentials"],
+        config["cookie"]["name"],
+        config["cookie"]["key"],
+        config["cookie"]["expiry_days"],
+    )
+
+    if st.session_state["auth_mode"] == "register":
+        try:
+            reg_result = authenticator.register_user(
+                location="main", captcha=False, key="register_user"
+            )
+            if reg_result:
+                status, username, name = reg_result
+                if status == "New user registered":
+                    st.success(
+                        f"Account created for **{username}**! You can now log in."
+                    )
+                    save_auth_config(config)
+                    st.session_state["auth_mode"] = "login"
+                    st.rerun()
+                elif status:
+                    st.info(status)
+        except Exception as e:
+            st.error(str(e))
+
+        if st.button("← Back to login", key="back_to_login_btn"):
+            st.session_state["auth_mode"] = "login"
+            st.rerun()
+        return
+
+    auth_result = authenticator.login(location="main", key="login")
+    if auth_result:
+        name, authentication_status, username = auth_result
+        st.session_state["auth_name"] = name
+        st.session_state["auth_username"] = username
+        st.session_state["authentication_status"] = authentication_status
+    else:
+        authentication_status = st.session_state.get("authentication_status")
+
+    if authentication_status is None:
+        st.markdown("---")
+        col1, col2 = st.columns([3, 1])
+        with col1:
+            st.markdown(f"### Welcome to {APP_TITLE}")
+            st.markdown(f"*{APP_SUBTITLE}*")
+        with col2:
+            if st.button("Register new account", key="goto_register_btn"):
+                st.session_state["auth_mode"] = "register"
+                st.rerun()
+        return
+    elif not authentication_status:
+        return
+
+    # --- SESSION STATE INIT ---
+    init_state = [
+        ("analysis_result", None),
+        ("analysis_completed", False),
+        ("pdf_report_path", None),
+        ("report_error", None),
+        ("report_success", False),
+        ("confidence_score", None),
+        ("risk_score", None),
+        ("risk_level", None),
+        ("risk_reasons", []),
+        ("explanation", None),
+        ("cost_estimate", None),
+        ("claim_history", None),
+        ("uploaded_images", []),
+        ("latest_claim", None),
+        ("analysis_stage", ""),
+        ("analysis_progress", 0),
+        ("analysis_status", "Idle"),
+        ("analyzing", False),
+    ]
+    for key, default in init_state:
+        if key not in st.session_state:
+            st.session_state[key] = default
     init_workflow_state()
 
     active_section = render_sidebar()
+
+    st.sidebar.markdown(f"👤 **{st.session_state['auth_name']}**")
+    authenticator.logout("Logout", "sidebar", key="logout_btn")
     if active_section in ("Overview", "Review Workspace", "AI Analysis", "Report"):
         render_mobile_brand_bar()
         render_hero()
@@ -3347,9 +3477,10 @@ def main():
             st.markdown("### Review Actions")
             action_cols = st.columns([0.35, 0.65])
             with action_cols[0]:
+                is_analyzing = st.session_state["analyzing"]
                 analyze_clicked = st.button(
-                    "Run AI Analysis",
-                    disabled=not uploaded_files,
+                    "Analyzing image..." if is_analyzing else "Run AI Analysis",
+                    disabled=not uploaded_files or is_analyzing,
                     use_container_width=True,
                 )
             with action_cols[1]:
@@ -3370,23 +3501,29 @@ def main():
                 )
 
             if analyze_clicked and uploaded_files:
-                saved_image_path = save_uploaded_image(uploaded_files[0])
-                start_workflow()
-                st.session_state["analysis_completed"] = False
-                st.session_state["pdf_report_path"] = None
-                st.session_state["report_error"] = None
-                st.session_state["report_success"] = False
-
-                status_slot = st.empty()
-                loader_slot = st.empty()
-                progress_slot = st.empty()
-                status_slot.markdown(verification_badge("Processing"), unsafe_allow_html=True)
-                with loader_slot.container():
-                    render_workflow_card()
-                progress_bar = run_verification_progress(progress_slot, loader_slot)
-
-                with st.spinner("Gemini is reviewing the uploaded evidence..."):
-                    raw_result = analyze_image(saved_image_path, claim_object)
+                file_errors = validate_uploaded_file(uploaded_files[0])
+                if file_errors:
+                    for err in file_errors:
+                        st.error(err)
+                else:
+                    st.session_state["analyzing"] = True
+                    saved_image_path = save_uploaded_image(uploaded_files[0])
+                    start_workflow()
+                    st.session_state["analysis_completed"] = False
+                    st.session_state["pdf_report_path"] = None
+                    st.session_state["report_error"] = None
+                    st.session_state["report_success"] = False
+
+                    status_slot = st.empty()
+                    loader_slot = st.empty()
+                    progress_slot = st.empty()
+                    status_slot.markdown(verification_badge("Processing"), unsafe_allow_html=True)
+                    with loader_slot.container():
+                        render_workflow_card()
+                    progress_bar = run_verification_progress(progress_slot, loader_slot)
+
+                    with st.spinner("Analyzing image..."):
+                        raw_result = analyze_image(saved_image_path, claim_object)
                     claim_history = load_user_history(claim_object)
                     st.session_state["claim_history"] = claim_history
                     completed_result = ensure_analysis_result_contract(
@@ -3401,6 +3538,7 @@ def main():
                         build_output_record(claim_object, user_claim, saved_image_path, completed_result)
                     )
                     st.session_state["analysis_completed"] = True
+                    st.session_state["analyzing"] = False
 
                 if st.session_state["analysis_completed"]:
                     try:
@@ -3451,6 +3589,7 @@ def main():
                         st.session_state["report_success"] = False
                         progress_bar.progress(100, text="Analysis Complete")
                         complete_workflow("AI analysis completed, but the PDF report could not be generated.")
+                        st.session_state["analyzing"] = False
                 st.rerun()
 
             if st.session_state["analysis_result"]:

code/requirements.txt

@@ -12,3 +12,5 @@ shap
 opencv-python
 ruff
 black
+streamlit-authenticator
+pyyaml

code/src/image_analyzer.py

@@ -14,6 +14,29 @@
     "gemini-2.5-flash"
 )
 
+ALLOWED_EXTENSIONS = {".jpg", ".jpeg", ".png"}
+MAX_FILE_SIZE = 25 * 1024 * 1024
+
+
+def validate_image_file(image_path):
+    ext = os.path.splitext(str(image_path))[1].lower()
+    if ext not in ALLOWED_EXTENSIONS:
+        return False, f"Unsupported file type '{ext}'. Only JPG, JPEG, and PNG images are allowed."
+
+    file_size = os.path.getsize(image_path)
+    if file_size == 0:
+        return False, "The uploaded file is empty."
+    if file_size > MAX_FILE_SIZE:
+        return False, f"File size exceeds {MAX_FILE_SIZE // (1024 * 1024)} MB."
+
+    try:
+        with Image.open(image_path) as img:
+            img.verify()
+    except Exception:
+        return False, "The image appears to be corrupted or unreadable."
+
+    return True, ""
+
 
 def _estimate_repair_cost(claim_object, severity):
     severity = str(severity or "unknown").lower()
@@ -137,6 +160,18 @@ def _apply_result_contract(result, claim_object):
 
 def analyze_image(image_path, claim_object):
 
+    is_valid, error = validate_image_file(image_path)
+    if not is_valid:
+        return _apply_result_contract({
+            "object_type": claim_object,
+            "issue_type": "unknown",
+            "object_part": "unknown",
+            "damage_visible": False,
+            "severity": "unknown",
+            "valid_image": False,
+            "quality_flags": [f"validation_error: {error}"]
+        }, claim_object)
+
     image = Image.open(image_path)
 
     prompt = f"""