update

src/streamlit_app.py

@@ -2,22 +2,54 @@ import streamlit as st
 from datasets import load_dataset
 
 st.set_page_config(page_title="Paper Explorer", layout="wide")
-st.title("📚 Paper Explorer")
 
-# Streaming=True ensures we don't download the 1GB at once!
-dataset = load_dataset("pwc-archive/papers-with-abstracts", split="train", streaming=True)
+# 1. Initialize Session State for the page number
+if 'page_number' not in st.session_state:
+    st.session_state.page_number = 0
 
-# Simple Pagination in Sidebar
+st.title("📚 Papers with code Explorer")
+
+# 2. Setup constants
+PAGE_SIZE = 10
+dataset_name = "pwc-archive/papers-with-abstracts"
+
+# 3. Sidebar Navigation
 st.sidebar.header("Navigation")
-page_size = st.sidebar.slider("Papers per page", 5, 25, 10)
-offset = st.sidebar.number_input("Start at row", min_value=0, step=page_size)
 
-# Fetch rows
-papers = list(dataset.skip(offset).take(page_size))
+def next_page():
+    st.session_state.page_number += 1
+
+def prev_page():
+    if st.session_state.page_number > 0:
+        st.session_state.page_number -= 1
+
+col1, col2, col3 = st.sidebar.columns([1, 2, 1])
+
+with col1:
+    st.button("⬅️", on_click=prev_page, disabled=(st.session_state.page_number == 0))
+with col2:
+    st.write(f"**Page {st.session_state.page_number + 1}**")
+with col3:
+    st.button("➡️", on_click=next_page)
+
+# 4. Fetch Data based on page
+offset = st.session_state.page_number * PAGE_SIZE
+
+with st.spinner(f"Loading papers {offset} to {offset + PAGE_SIZE}..."):
+    dataset = load_dataset(dataset_name, split="train", streaming=True)
+    papers = list(dataset.skip(offset).take(PAGE_SIZE))
 
-# Display
+# 5. Display the Results
 for p in papers:
-    with st.expander(f"Title: {p['title']}", expanded=True):
-        st.write(f"**Abstract:** {p['abstract'][:500]}...")
-        if p['arxiv_id']:
+    with st.container():
+        st.subheader(p['title'])
+        st.caption(f"ArXiv ID: {p.get('arxiv_id', 'N/A')}")
+        
+        # Display abstract (limiting length for readability)
+        abstract = p.get('abstract', 'No abstract available.')
+        st.write(abstract[:500] + ("..." if len(abstract) > 500 else ""))
+        
+        if p.get('arxiv_id'):
             st.link_button("View on ArXiv", f"https://arxiv.org/abs/{p['arxiv_id']}")
+        
+        st.divider()