Initil commit

Dockerfile

@@ -0,0 +1,58 @@
+FROM python:3.9-slim
+
+# 1) Install dependencies for Dev Mode + Playwright
+RUN apt-get update && \
+    apt-get install -y \
+      bash \
+      curl \
+      wget \
+      procps \
+      git \
+      git-lfs \
+      libnss3 \
+      libatk1.0-0 \
+      libatk-bridge2.0-0 \
+      libx11-6 \
+      libx11-xcb1 \
+      libxcomposite1 \
+      libxcursor1 \
+      libxdamage1 \
+      libxext6 \
+      libxfixes3 \
+      libxi6 \
+      libxrandr2 \
+      libxrender1 \
+      libxss1 \
+      libxtst6 \
+      libappindicator1 \
+      libsecret-1-0 \
+      fonts-ipafont-gothic && \
+    rm -rf /var/lib/apt/lists/*
+
+# 2) Copy code into /app
+WORKDIR /app
+COPY . /app
+
+# 3) Install Python dependencies
+RUN pip install --upgrade pip
+RUN pip install gradio 
+RUN pip install -U crawl4ai
+
+# 4) Install Playwright browser(s)
+RUN pip install playwright
+RUN playwright install --with-deps chrome
+
+# 5) Make /app owned by user 1000 (Dev Mode requirement)
+RUN chown -R 1000 /app
+
+# 6) Ensure the HOME variable points to /app (so Git writes to /app/.gitconfig)
+ENV HOME=/app
+
+# 7) Switch to user 1000
+USER 1000
+
+# 8) Expose port for Gradio
+EXPOSE 7860
+
+# 9) Start your Gradio app
+CMD ["python", "app.py"]

app.py

@@ -0,0 +1,29 @@
+import gradio as gr
+import asyncio
+from crawl4ai import AsyncWebCrawler
+from crawl4ai import CrawlerRunConfig, CacheMode
+from crawl4ai.markdown_generation_strategy import DefaultMarkdownGenerator
+
+def crawl_url_to_markdown(url):
+    async def _crawl(u):
+        config = CrawlerRunConfig(
+            cache_mode=CacheMode.BYPASS,
+            markdown_generator=DefaultMarkdownGenerator()
+        )
+        async with AsyncWebCrawler() as crawler:
+            result = await crawler.arun(u, config=config)
+            return result.markdown_v2.raw_markdown
+
+    return asyncio.run(_crawl(url))
+
+demo = gr.Interface(
+    fn=crawl_url_to_markdown,
+    inputs="text",
+    outputs="text",
+    title="Crawl4AI to Markdown",
+    description="Enter a URL to crawl, returns page as Markdown"
+)
+
+if __name__ == "__main__":
+    # Launch on 0.0.0.0:7860 for Hugging Face Spaces
+    demo.launch(server_name="0.0.0.0", server_port=7860)