Upload 2 files

app.py

@@ -0,0 +1,47 @@
+import gradio as gr
+import requests
+from bs4 import BeautifulSoup
+from markdownify import MarkdownConverter
+
+
+def md(soup, **options):
+    return MarkdownConverter(**options).convert_soup(soup)
+
+
+def main_fn(url: str, check: list[int]):
+
+    response = requests.get(url)
+    soup = BeautifulSoup(response.text)
+
+    for tag in ["script", "style"]:
+        target = soup.find_all(tag)
+        for t in target:
+            t.clear
+
+    body = soup.find("body")
+    main = soup.find("main")
+
+    if main:
+        return md(main, strip=check)
+
+    return md(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()

requirements.txt

@@ -0,0 +1,4 @@
+beautifulsoup4
+gradio
+markdownify
+requests