Upload 12 files

.huggingface.yml

@@ -0,0 +1,3 @@
+sdk: docker
+app_file: app.py
+entrypoint: ./launch.sh

app.py

@@ -1,133 +1,92 @@
-"""
-app.py — Bodha Vector API (Stable Hugging Face Version)
--------------------------------------------------------
-Provides:
-  • /vector-api/search  → REST API with JSON output
-  • /                   → Gradio UI for manual testing
-"""
-
-import os
-import json
-import chromadb
-from fastapi import FastAPI, Request, Depends, HTTPException, status
-from fastapi.responses import JSONResponse
-from fastapi.middleware.cors import CORSMiddleware
-from fastapi.security import HTTPBasic, HTTPBasicCredentials
-import gradio as gr
-from typing import List
-from contextlib import asynccontextmanager
-
-# ========== CONFIG ==========
-os.environ["CHROMA_TELEMETRY"] = "FALSE"  # disable telemetry warnings
-
-CHROMA_PATH = "/app/textbook_db"          # use absolute path inside Hugging Face
-COLLECTION_NAME = "textbooks"
-USERNAME = "bodha"
-PASSWORD = "securepass123"
-TOP_K = 5
-# =============================
-
-# ---------- Initialize Chroma ----------
-try:
-    chroma_client = chromadb.PersistentClient(path=CHROMA_PATH)
-    collection = chroma_client.get_or_create_collection(COLLECTION_NAME)
-except Exception as e:
-    raise RuntimeError(f"❌ Failed to initialize Chroma: {e}")
-
-security = HTTPBasic()
-
-# ---------- Lifespan Startup ----------
-@asynccontextmanager
-async def lifespan(app: FastAPI):
-    print("✅ Bodha Vector API is starting up...")
-    try:
-        print(f"📂 Loading collection: {COLLECTION_NAME}")
-        print(f"📄 Total docs: {collection.count()}")
-    except Exception as e:
-        print(f"⚠️ Warning: Could not access collection: {e}")
-    yield
-    print("🛑 Shutting down Bodha Vector API...")
-
-app = FastAPI(lifespan=lifespan, title="Bodha Vector API")
-
-# ---------- Allow CORS for WordPress / JS ----------
-app.add_middleware(
-    CORSMiddleware,
-    allow_origins=["*"],
-    allow_methods=["*"],
-    allow_headers=["*"],
-)
-
-# ---------- Basic Auth ----------
-def verify_credentials(credentials: HTTPBasicCredentials = Depends(security)):
-    if not (
-        credentials.username == USERNAME and
-        credentials.password == PASSWORD
-    ):
-        raise HTTPException(
-            status_code=status.HTTP_401_UNAUTHORIZED,
-            detail="Invalid credentials",
-            headers={"WWW-Authenticate": "Basic"},
-        )
-    return credentials.username
-
-# ---------- API Endpoint ----------
-@app.post("/vector-api/search")
-async def search_vector(request: Request, user: str = Depends(verify_credentials)):
-    try:
-        data = await request.json()
-        query = data.get("query", "").strip()
-        if not query:
-            return JSONResponse({"error": "No query provided."}, status_code=400)
-
-        results = collection.query(query_texts=[query], n_results=TOP_K)
-        docs = results.get("documents", [[]])[0]
-        metas = results.get("metadatas", [[]])[0]
-        dists = results.get("distances", [[]])[0]
-
-        response_data = []
-        for i, doc in enumerate(docs):
-            meta = metas[i]
-            score = 1 - dists[i] if dists else None  # similarity score
-            response_data.append({
-                "rank": i + 1,
-                "score": round(score, 4) if score else None,
-                "class": meta.get("class", ""),
-                "chapter": meta.get("chapter", ""),
-                "pages": meta.get("pages", ""),
-                "source": meta.get("source_pdf", ""),
-                "text": doc
-            })
-        return JSONResponse(response_data)
-
-    except Exception as e:
-        return JSONResponse({"error": str(e)}, status_code=500)
-
-# ---------- Optional Gradio Interface ----------
-def gradio_query(query):
-    if not query.strip():
-        return "⚠️ Please enter a query."
-    results = collection.query(query_texts=[query], n_results=TOP_K)
-    docs = results.get("documents", [[]])[0]
-    metas = results.get("metadatas", [[]])[0]
-    dists = results.get("distances", [[]])[0]
-    output = ""
-    for i, doc in enumerate(docs):
-        meta = metas[i]
-        score = 1 - dists[i] if dists else None
-        output += f"### Result {i+1} (score: {round(score,3) if score else 'N/A'})\n"
-        output += f"**Class:** {meta.get('class','')}  \n"
-        output += f"**Source:** {meta.get('source_pdf','')}  \n"
-        output += f"**Pages:** {meta.get('pages','')}  \n\n"
-        output += f"{doc[:700]}...\n\n---\n\n"
-    return output or "No results found."
-
-demo = gr.Interface(
-    fn=gradio_query,
-    inputs=gr.Textbox(label="Ask your Physics/Science question"),
-    outputs="markdown",
-    title="🔍 Bodha Textbook Search",
-    description="Search across NCERT (Class 6–12) & H.C. Verma textbooks"
-)
-
-app = gr.mount_gradio_app(app, demo, path="/")
+"""
+app.py — Bodha Vector API for Hugging Face
+Provides:
+ • /vector-api/search  → JSON API (with Basic Auth)
+ • /                   → Gradio interface
+"""
+
+import os, chromadb, gradio as gr
+from fastapi import FastAPI, Request, Depends, HTTPException, status
+from fastapi.responses import JSONResponse
+from fastapi.middleware.cors import CORSMiddleware
+from fastapi.security import HTTPBasic, HTTPBasicCredentials
+from contextlib import asynccontextmanager
+
+# ---------- CONFIG ----------
+os.environ["CHROMA_TELEMETRY"] = "FALSE"
+CHROMA_PATH = "/app/textbook_db"
+COLLECTION_NAME = "textbooks"
+USERNAME, PASSWORD = "bodha", "securepass123"
+TOP_K = 5
+# ----------------------------
+
+# ---------- INIT ----------
+client = chromadb.PersistentClient(path=CHROMA_PATH)
+collection = client.get_or_create_collection(COLLECTION_NAME)
+security = HTTPBasic()
+
+@asynccontextmanager
+async def lifespan(app: FastAPI):
+    print("✅ Starting Bodha Vector API...")
+    print(f"📚 Collection: {COLLECTION_NAME} | Docs: {collection.count()}")
+    yield
+    print("🛑 Stopping Bodha API...")
+
+app = FastAPI(lifespan=lifespan)
+app.add_middleware(CORSMiddleware, allow_origins=["*"], allow_methods=["*"], allow_headers=["*"])
+
+# ---------- AUTH ----------
+def verify(credentials: HTTPBasicCredentials = Depends(security)):
+    if credentials.username != USERNAME or credentials.password != PASSWORD:
+        raise HTTPException(status_code=status.HTTP_401_UNAUTHORIZED,
+                            detail="Invalid credentials",
+                            headers={"WWW-Authenticate": "Basic"})
+    return credentials.username
+
+# ---------- API ----------
+@app.post("/vector-api/search")
+async def search_vector(request: Request, user: str = Depends(verify)):
+    body = await request.json()
+    query = body.get("query", "").strip()
+    if not query:
+        return JSONResponse({"error": "Missing query"}, status_code=400)
+
+    res = collection.query(query_texts=[query], n_results=TOP_K)
+    docs, metas, dists = res["documents"][0], res["metadatas"][0], res["distances"][0]
+    out = []
+    for i, doc in enumerate(docs):
+        meta = metas[i]
+        score = 1 - dists[i] if dists else None
+        out.append({
+            "rank": i + 1,
+            "score": round(score, 4) if score else None,
+            "class": meta.get("class", ""),
+            "chapter": meta.get("chapter", ""),
+            "pages": meta.get("pages", ""),
+            "source": meta.get("source_pdf", ""),
+            "text": doc
+        })
+    return JSONResponse(out)
+
+# ---------- GRADIO ----------
+def gradio_query(q):
+    if not q.strip():
+        return "⚠️ Enter a query."
+    res = collection.query(query_texts=[q], n_results=TOP_K)
+    docs, metas, dists = res["documents"][0], res["metadatas"][0], res["distances"][0]
+    out = ""
+    for i, doc in enumerate(docs):
+        meta = metas[i]
+        score = 1 - dists[i] if dists else None
+        out += f"### {i+1}. {meta.get('class','')} ({meta.get('source_pdf','')}) — score {round(score,3)}\n{doc[:700]}...\n\n---\n"
+    return out or "No results."
+
+demo = gr.Interface(
+    fn=gradio_query,
+    inputs=gr.Textbox(label="Ask a Physics/Science question"),
+    outputs="markdown",
+    title="🔍 Bodha Textbook Search",
+    description="Search across NCERT (Class 6–12) + H.C.Verma"
+)
+
+app = gr.mount_gradio_app(app, demo, path="/")

launch.sh

@@ -0,0 +1,4 @@
+#!/bin/bash
+# Hugging Face starts this script; it launches FastAPI via uvicorn.
+echo "🚀 Launching Bodha Vector API on port 7860..."
+exec uvicorn app:app --host 0.0.0.0 --port 7860

requirements.txt

@@ -1,4 +1,4 @@
 chromadb==0.5.3
 gradio==4.29.0
 fastapi==0.111.0
-uvicorn==0.30.0
+uvicorn==0.30.0