Upload 3 files

Dockerfile

@@ -0,0 +1,25 @@
+FROM mcr.microsoft.com/playwright:v1.47.0-noble
+
+RUN apt-get update -q && apt-get install -qy python3 python3-pip
+
+# Switch to the "ubuntu" user
+USER ubuntu
+
+# Set home to the ubuntu's home directory
+ENV HOME=/home/ubuntu \
+	PATH=/home/ubuntu/.local/bin:$PATH
+
+# Set the working directory to the user's home directory
+WORKDIR $HOME/app
+
+# Try and run pip command after setting the user with `USER user` to avoid permission issues with Python
+COPY --chown=ubuntu requirements.txt $HOME/app
+RUN pip install --no-cache-dir --upgrade -r requirements.txt --break-system-packages
+
+# Copy the current directory contents into the container at $HOME/app setting the owner to the user
+COPY --chown=ubuntu app.py $HOME/app
+
+EXPOSE 7860
+
+ENTRYPOINT [ "python3" ]
+CMD [ "app.py" ]

app.py

@@ -0,0 +1,74 @@
+import gradio as gr
+from bs4 import BeautifulSoup
+from markdownify import MarkdownConverter
+from playwright.sync_api import sync_playwright
+
+
+def md(soup, **options):
+    return MarkdownConverter(**options).convert_soup(soup)
+
+
+def main_fn(url: str, check: list[int], request: gr.Request):
+
+    user_agent = request.headers["user-agent"]
+
+    with sync_playwright() as p:
+
+        browser = p.chromium.launch(
+            args=[
+                "--single-process",
+                "--no-zygote",
+                "--no-sandbox",
+                "--disable-gpu",
+                "--disable-dev-shm-usage",
+                "--headless=new",
+            ]
+        )
+
+        context = browser.new_context(user_agent=user_agent)
+        page = context.new_page()
+
+        response = page.goto(url=url)
+        status = response.status
+
+        content = page.content()
+        title = page.title()
+
+        browser.close()
+
+    soup = BeautifulSoup(content, features="html.parser")
+
+    for tag in ["script", "style"]:
+        target = soup.find_all(tag)
+        for t in target:
+            t.clear
+
+    body = soup.find("body")
+    main = soup.find("main")
+
+    body = md(body)
+
+    if main:
+        body = md(main, strip=check)
+
+    return f"{title}\n======\n\n{body}"
+
+
+demo = gr.Interface(
+    main_fn,
+    title="URL to Markdown",
+    description="""<div style="width: fit-content; margin: 0 auto;">Gets HTML given by URL and converts it to Markdown.Does not support dynamically generated HTML such as React.</div>
+    <div style="width: fit-content; margin: 0 auto;">URLで与えたHTMLを取得してMarkdownに変換します。Reactなどの動的に生成されるHTMLには対応していません</div>""",
+    inputs=[
+        gr.Text(label="URL", placeholder="https://*****"),
+        gr.CheckboxGroup(
+            label="Ignore tags（無視するタグ）",
+            choices=["a", "img", "noscript"],
+            value=["a", "img"],
+        ),
+    ],
+    outputs=[gr.TextArea(label="Markdown", show_copy_button=True)],
+    allow_flagging="never",
+)
+
+demo.launch(server_name="0.0.0.0")

requirements.txt

@@ -0,0 +1,5 @@
+beautifulsoup4
+gradio
+markdownify
+playwright==1.47.0
+requests