app.py

# app.py
import os
from pathlib import Path
import subprocess
from collections import OrderedDict
import streamlit as st

# ------------------------------------------------------------------
# Detect if we're running inside a Hugging Face Space (best-effort)
# ------------------------------------------------------------------
IS_HF = any(k in os.environ for k in ("SPACE_ID", "HF_SPACE_ID", "HF_HOME", "HF_REPO_ID"))

# Opt-in flags (set in HF: Settings → Variables & secrets)
SHOW_MAINTENANCE = os.environ.get("IWEEMS_MAINTENANCE", "0") == "1"
USE_LFS_BOOTSTRAP = IS_HF or (os.environ.get("IWEEMS_USE_LFS_BOOTSTRAP", "0") == "1")

# ------------------------------------------------------------------
# Optional maintenance panel (HF-only, and only when you enable it)
# ------------------------------------------------------------------
if IS_HF and SHOW_MAINTENANCE:
    with st.expander("🔧 Maintenance / Clear caches"):
        # Hugging Face hub cache
        if st.button("Clear Hugging Face hub cache"):
            try:
                from huggingface_hub import scan_cache_dir
                report = scan_cache_dir().delete()  # removes ~/.cache/huggingface/*
                freed = getattr(report, "reclaimed_size", 0) / 1e6
                st.success(f"Cleared hub cache (~{freed:.1f} MB).")
            except Exception as e:
                st.error(f"Hub cache clear failed: {e}")

        # Streamlit caches
        if st.button("Clear Streamlit caches"):
            st.cache_data.clear()
            st.cache_resource.clear()
            st.success("Streamlit caches cleared. Please rerun.")

# ------------------------------------------------------------------
# LFS bootstrap (ensures real files, not pointer stubs) — HF only
# ------------------------------------------------------------------
def _is_lfs_pointer(p: Path) -> bool:
    try:
        if not p.exists() or p.stat().st_size > 300:
            return False
        return p.read_bytes().startswith(b"version https://git-lfs.github.com")
    except Exception:
        return False

def ensure_lfs_materialized(paths, repo_id: str = "vojkas/iweems"):
    # Fast path: try git-lfs pull; ignore failures if not available
    try:
        subprocess.run(["git", "lfs", "install", "--local"], check=False,
                       stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL)
        subprocess.run(["git", "lfs", "pull"], check=False,
                       stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL)
    except Exception:
        pass

    # Fallback: if any file still looks like a pointer, fetch actual bytes via hub
    try:
        from huggingface_hub import hf_hub_download
        for rel in paths:
            p = Path(rel)
            if not p.exists() or _is_lfs_pointer(p):
                real = hf_hub_download(repo_id=repo_id, filename=str(p))
                p.parent.mkdir(parents=True, exist_ok=True)
                p.write_bytes(Path(real).read_bytes())
    except Exception:
        # If hub isn't available, leave as-is; open() will surface a clear error.
        pass

# Only do this on HF (or if you explicitly opt-in locally)
if USE_LFS_BOOTSTRAP:
    ensure_lfs_materialized([
        "data/vectors_dict_comp.pkl",
        "data/coordinates3d_tsne_dict.pkl",
        "data/filtered_vocab_df.json",
        "data/cos_sim_matrices_dict_f16.pkl",  # used by crosscorpora_comparisons
    ])

# ------------------------------------------------------------------
# Your original app wiring (unchanged)
# ------------------------------------------------------------------
import overview
import explorer
import crosscorpora_comparisons

st.set_page_config(
    page_title="iWEEMS",
    page_icon="📊",
    layout="wide",
)

PAGES = OrderedDict({
    "📊 About": overview,
    "🔍 Visualization App": explorer,
    "🔍 Crosscorpora Comparisons": crosscorpora_comparisons,
})
DEFAULT_PAGE = "🔍 Visualization App"

if "page" not in st.session_state:
    st.session_state.page = DEFAULT_PAGE

st.sidebar.title("📌 Navigation")
for idx, (page_name, module) in enumerate(PAGES.items()):
    if st.sidebar.button(page_name, key=f"nav_btn_{idx}"):
        st.session_state.page = page_name

selected_page = PAGES.get(st.session_state.page, PAGES[DEFAULT_PAGE])
selected_page.app()