src/streamlit_app.py

import os
import fitz
import tempfile
import streamlit as st
from PIL import Image
from chromadb import PersistentClient
from chromadb.utils.data_loaders import ImageLoader
from chromadb.utils.embedding_functions import OpenCLIPEmbeddingFunction
from skimage import data as skdata
from skimage.io import imsave
import uuid

# Use safe temp directories for Streamlit or restricted environments
TEMP_DIR = tempfile.gettempdir()
IMAGES_DIR = os.path.join(TEMP_DIR, "extracted_images")
DB_PATH = os.path.join(TEMP_DIR, "image_vdb")
os.makedirs(IMAGES_DIR, exist_ok=True)

@st.cache_resource
def get_chroma_collection():
    chroma_client = PersistentClient(path=DB_PATH)
    image_loader = ImageLoader()
    embedding_fn = OpenCLIPEmbeddingFunction()
    collection = chroma_client.get_or_create_collection(
        name="image", embedding_function=embedding_fn, data_loader=image_loader
    )
    return collection

image_collection = get_chroma_collection()

# === Image Extraction ===
def extract_images_from_pdf(pdf_bytes):
    pdf = fitz.open(stream=pdf_bytes, filetype="pdf")
    saved_images = []

    for page_num in range(len(pdf)):
        page = pdf.load_page(page_num)
        images = page.get_images(full=True)

        for img_idx, img in enumerate(images):
            xref = img[0]
            base_image = pdf.extract_image(xref)
            img_bytes = base_image["image"]
            ext = base_image["ext"]
            filename = f"page_{page_num+1}_img_{img_idx+1}.{ext}"
            path = os.path.join(IMAGES_DIR, filename)

            with open(path, "wb") as f:
                f.write(img_bytes)

            saved_images.append(path)

    return saved_images

# === Indexing ===
def index_images(image_paths):
    ids = []
    uris = []
    for path in sorted(image_paths):
        if path.lower().endswith((".png", ".jpeg", ".jpg")):
            ids.append(str(uuid.uuid4()))
            uris.append(path)

    if ids:
        image_collection.add(ids=ids, uris=uris)

# === Querying ===
def query_similar_images(image_file, top_k=5):
    with tempfile.NamedTemporaryFile(delete=False, suffix=".jpg") as tmp:
        tmp.write(image_file.read())
        tmp_path = tmp.name

    try:
        results = image_collection.query(query_uris=[tmp_path], n_results=top_k)
        return results['uris'][0]
    finally:
        os.remove(tmp_path)

# === Demo images ===
def load_skimage_demo_images():
    demo_images = {
        "astronaut": skdata.astronaut(),
        "coffee": skdata.coffee(),
        "camera": skdata.camera(),
        "chelsea": skdata.chelsea(),
        "rocket": skdata.rocket()
    }
    saved_paths = []

    for name, img in demo_images.items():
        path = os.path.join(IMAGES_DIR, f"{name}.png")
        imsave(path, img)
        saved_paths.append(path)

    return saved_paths

# === Streamlit UI ===
st.title("🔍 Image Similarity Search from PDF or Custom Dataset")

source = st.radio(
    "Select Image Source",
    ["Upload PDF", "Upload Images", "Load Demo Dataset"],
    horizontal=True
)

if source == "Upload PDF":
    uploaded_pdf = st.file_uploader("📤 Upload PDF", type=["pdf"])
    if uploaded_pdf:
        with st.spinner("Extracting images..."):
            images = extract_images_from_pdf(uploaded_pdf.read())
            index_images(images)
            st.success(f"{len(images)} images extracted and indexed.")
            st.image(images, width=150)

elif source == "Upload Images":
    uploaded_imgs = st.file_uploader(
        "📤 Upload one or more images", type=["jpg", "jpeg", "png"], accept_multiple_files=True
    )
    if uploaded_imgs:
        saved_paths = []
        for img in uploaded_imgs:
            img_path = os.path.join(IMAGES_DIR, img.name)
            with open(img_path, "wb") as f:
                f.write(img.read())
            saved_paths.append(img_path)

        index_images(saved_paths)
        st.success(f"{len(saved_paths)} images indexed.")
        st.image(saved_paths, width=150)

elif source == "Load Demo Dataset":
    if st.button("🔄 Load Demo Images (skimage)"):
        demo_paths = load_skimage_demo_images()
        index_images(demo_paths)
        st.success("Demo images loaded and indexed.")
        st.image(demo_paths, width=150)

st.divider()

st.subheader("🔎 Search for Similar Images")
query_img = st.file_uploader("Upload a query image", type=["jpg", "jpeg", "png"])
if query_img:
    st.image(query_img, caption="Query Image", width=200)
    with st.spinner("Searching..."):
        matches = query_similar_images(query_img, top_k=5)

    st.subheader("📊 Top Matches:")
    for match in matches:
        st.image(match, width=200, caption=os.path.basename(match))