Create scrape.py
Browse files
scrape.py
ADDED
|
@@ -0,0 +1,49 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import streamlit as st
|
| 2 |
+
from scrapegraphai.graphs import SmartScraperGraph
|
| 3 |
+
|
| 4 |
+
st.set_page_config(page_title="Web Scraping AI Agent", layout="centered")
|
| 5 |
+
|
| 6 |
+
st.title("🌐 Web Scraping AI Agent")
|
| 7 |
+
st.caption("This app allows you to scrape a site using LLMs (Ollama / OpenAI).")
|
| 8 |
+
|
| 9 |
+
# Graph config
|
| 10 |
+
graph_config = {
|
| 11 |
+
"llm": {
|
| 12 |
+
"model": "ollama/llama3",
|
| 13 |
+
"temperature": 0,
|
| 14 |
+
"format": "json",
|
| 15 |
+
"base_url": "http://localhost:11434",
|
| 16 |
+
},
|
| 17 |
+
"embeddings": {
|
| 18 |
+
"model": "ollama/nomic-embed-text",
|
| 19 |
+
"base_url": "http://localhost:11434",
|
| 20 |
+
},
|
| 21 |
+
"browser": "chromium",
|
| 22 |
+
"playwright": {
|
| 23 |
+
"headless": True, # don’t open window
|
| 24 |
+
"timeout": 20000 # stop after 20s
|
| 25 |
+
},
|
| 26 |
+
"verbose": True,
|
| 27 |
+
}
|
| 28 |
+
|
| 29 |
+
|
| 30 |
+
# Inputs
|
| 31 |
+
url = st.text_input("🔗 Enter URL you want to scrape:")
|
| 32 |
+
user_prompt = st.text_area("📝 What do you want the AI Agent to scrape from the site?")
|
| 33 |
+
|
| 34 |
+
if st.button("🚀 Scrape"):
|
| 35 |
+
if not url or not user_prompt:
|
| 36 |
+
st.warning("⚠️ Please enter both a URL and a prompt.")
|
| 37 |
+
else:
|
| 38 |
+
try:
|
| 39 |
+
smart_scraper_graph = SmartScraperGraph(
|
| 40 |
+
prompt=user_prompt,
|
| 41 |
+
source=url,
|
| 42 |
+
config=graph_config
|
| 43 |
+
)
|
| 44 |
+
with st.spinner("Scraping in progress... ⏳"):
|
| 45 |
+
result = smart_scraper_graph.run()
|
| 46 |
+
st.success("✅ Scraping complete!")
|
| 47 |
+
st.json(result) # Pretty JSON output
|
| 48 |
+
except Exception as e:
|
| 49 |
+
st.error(f"❌ Error: {str(e)}")
|