Update app.py

app.py

@@ -1,9 +1,14 @@
-from fastapi import FastAPI
-from pydantic import BaseModel
+# app.py
+"""
+Complaint Prioritization API (Gradio-based)
+Drop this file into a Hugging Face Space (SDK: Gradio) or run locally.
+"""
+
+from typing import List, Dict, Any
+import gradio as gr
 
 # ================================
-# Complaint Prioritization Logic
-# ================================
+# Default priority values mapped 0–1
 DEFAULT_PRIORITY_SCORE = {
     "water": 0.9,
     "electricity": 0.9,
@@ -12,7 +17,8 @@ DEFAULT_PRIORITY_SCORE = {
     "garbage": 0.2
 }
 
-def get_priority_label(score):
+# Helper to convert 0–1 score into labels
+def get_priority_label(score: float) -> str:
     if score >= 0.75:
         return "Critical"
     elif score >= 0.55:
@@ -22,41 +28,85 @@ def get_priority_label(score):
     else:
         return "Low"
 
-def calculate_weighted_score(upvotes, complaints, alpha=0.6, beta=0.4):
-    upvote_score = min(upvotes / 50, 1.0)
-    complaint_score = min(complaints / 20, 1.0)
+# Calculate weighted score (0–1)
+def calculate_weighted_score(upvotes: int, complaints: int, alpha: float = 0.6, beta: float = 0.4) -> float:
+    # Normalize to 0–1 scale (cap at 1)
+    upvote_score = min(upvotes / 50.0, 1.0)       # assume 50 upvotes = max
+    complaint_score = min(complaints / 20.0, 1.0) # assume 20 complaints = max
     weighted = (alpha * complaint_score) + (beta * upvote_score)
     return weighted
 
-def handle_complaint(text, category, complaints, upvotes):
+# Handle a single complaint
+def handle_complaint(text: str, category: str, complaints: int, upvotes: int) -> Dict[str, Any]:
     default_score = DEFAULT_PRIORITY_SCORE.get(category.lower(), 0.3)
     weighted_score = calculate_weighted_score(upvotes, complaints)
     final_score = 0.5 * default_score + 0.5 * weighted_score
+
     return {
         "text": text,
         "category": category,
-        "default_score": default_score,
+        "default_score": round(default_score, 2),
         "weighted_score": round(weighted_score, 2),
         "final_score": round(final_score, 2),
         "final_label": get_priority_label(final_score)
     }
 
-# ================================
-# FastAPI Setup
-# ================================
-app = FastAPI(title="Complaint Prioritization API")
-
-class ComplaintRequest(BaseModel):
-    text: str
-    category: str
-    complaints: int
-    upvotes: int
-
-@app.get("/")
-def root():
-    return {"message": "Complaint Prioritization API is live!"}
-
-@app.post("/prioritize")
-def prioritize_complaint(request: ComplaintRequest):
-    result = handle_complaint(request.text, request.category, request.complaints, request.upvotes)
-    return result
+# API entrypoint for a single complaint (for UI)
+def predict_single(text: str, category: str, complaints: int, upvotes: int):
+    try:
+        complaints = int(complaints)
+        upvotes = int(upvotes)
+    except Exception:
+        return {"error": "complaints and upvotes must be integers"}
+    return handle_complaint(text, category, complaints, upvotes)
+
+# API entrypoint for batch complaints (JSON text input)
+def predict_batch(json_string: str):
+    """
+    Accepts a JSON array string (list of dicts with keys: text, category, complaints, upvotes)
+    Example:
+    [
+      {"text":"Pothole","category":"road","complaints":2,"upvotes":5},
+      {"text":"Leak","category":"water","complaints":15,"upvotes":8}
+    ]
+    """
+    import json
+    try:
+        items = json.loads(json_string)
+        if not isinstance(items, list):
+            return {"error": "Expected a JSON array/list of complaints"}
+    except Exception as e:
+        return {"error": f"Invalid JSON: {str(e)}"}
+
+    results = []
+    for it in items:
+        t = it.get("text", "")
+        cat = it.get("category", "")
+        complaints = int(it.get("complaints", 0))
+        upvotes = int(it.get("upvotes", 0))
+        results.append(handle_complaint(t, cat, complaints, upvotes))
+    return {"results": results}
+
+# Small UI using Gradio; Spaces will host this and expose a programmatic endpoint automatically
+with gr.Blocks() as demo:
+    gr.Markdown("## Complaint Prioritization API\nProvide a single complaint or a JSON list for batch processing.")
+    with gr.Tab("Single complaint"):
+        txt = gr.Textbox(label="Complaint text", value="Huge pothole on main road, damaging cars daily.")
+        cat = gr.Dropdown(label="Category", choices=["water","electricity","gas","road","garbage","other"], value="road")
+        comp = gr.Number(label="Number of complaints", value=2, precision=0)
+        upv = gr.Number(label="Number of upvotes", value=5, precision=0)
+        out = gr.JSON(label="Result")
+        btn = gr.Button("Predict")
+        btn.click(fn=predict_single, inputs=[txt, cat, comp, upv], outputs=out)
+
+    with gr.Tab("Batch (JSON array)"):
+        batch_in = gr.Textbox(label="JSON array of complaints", lines=10, value='[{"text":"Leak near market","category":"water","complaints":15,"upvotes":8}]')
+        batch_out = gr.JSON(label="Batch results")
+        batch_btn = gr.Button("Predict batch")
+        batch_btn.click(fn=predict_batch, inputs=batch_in, outputs=batch_out)
+
+    gr.Markdown("### Programmatic usage\nUse the Spaces' `gradio` API endpoint `/api/predict/` (or use the `predict_batch` JSON route). See README for examples.")
+
+# When running locally in debug mode, this will start a Gradio server.
+if __name__ == "__main__":
+    demo.launch()