Upload 4 files

README.md

@@ -1,14 +1,14 @@
 ---
 title: Fake App Scraper
-emoji: 🐠
-colorFrom: gray
-colorTo: yellow
+emoji: 📈
+colorFrom: green
+colorTo: pink
 sdk: gradio
 sdk_version: 5.33.0
 app_file: app.py
 pinned: false
 license: mit
-short_description: test
+short_description: uses playwright to scrape a fake app hosted on vercel
+tag: "mcp-server-track"
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

app.py

@@ -0,0 +1,59 @@
+import gradio as gr
+from scrape_fake_app import get_homework,get_timetable
+
+def fetch_homework(date:str='today') -> str:
+    """
+    description:
+        fetch the homeworks.
+    Args: 
+        date: any string, default "today"
+    Returns:
+        The string describing the homeworks    
+    """
+    return get_homework()
+
+
+def fetch_timetable(date:str='today') -> str:
+    """
+    description:
+        fetch the timetable
+    Args: 
+        date: any string, default "today"
+    Returns:
+        The string describing the timetable    
+    """
+    return get_timetable()
+
+
+
+with gr.Blocks() as demo:
+    
+        # Add title and markdown
+    with gr.Row():
+        gr.Markdown("""## Fake App Dashboard
+                    Use playwright to scrape homework and timetable from [fake-app-omega.vercel.app](https://fake-app-omega.vercel.app)
+                    """)
+    
+    with gr.Row():
+        date = gr.Textbox(label="date",visible=False)
+
+    with gr.Row():
+        with gr.Column():
+            homeworks_btn = gr.Button("Homeworks")
+            homeworks_output = gr.Textbox(label="Homeworks Result", lines=10)
+        
+        with gr.Column():
+            timetable_btn = gr.Button("Timetable")
+            timetable_output = gr.Textbox(label="Timetable Result", lines=10)
+
+
+    homeworks_btn.click(fn=fetch_homework,
+                        inputs=[date],
+                        outputs=homeworks_output)
+
+
+    timetable_btn.click(fn=fetch_timetable,
+                        inputs=[date],
+                        outputs=timetable_output)
+
+demo.launch(mcp_server=True)

requirements.txt

@@ -0,0 +1,3 @@
+gradio[mcp]
+python-dotenv
+playwright

scrape_fake_app.py

@@ -0,0 +1,85 @@
+from playwright.sync_api import sync_playwright
+import os
+from dotenv import load_dotenv
+
+def load_credentials()-> dict:
+    load_dotenv()
+    URL = os.getenv('FAKE_APP_URL')
+    USERNAME = os.getenv('FAKE_APP_USERNAME')
+    PASSWORD = os.getenv('FAKE_APP_PASSWORD')
+    return (URL,USERNAME,PASSWORD)
+
+def extract_homework_text(page) -> str:
+    card = page.get_by_title("homework")
+    content = card.locator("[data-slot='card-content'] section > div").all()
+
+    output = ["Homework:\n"]
+    for section in content:
+        heading = section.locator("h3").inner_text()
+        output.append(heading)
+        items = section.locator("ul > li").all()
+        for item in items:
+            # Extract full inner text including formatting
+            inner = item.inner_text().strip()
+            output.append(f"  {inner}")
+        output.append("")  # Add a blank line between sections
+        
+    return "\n".join(output).strip()
+
+def extract_timetable_text(page):
+    card = page.get_by_title("timetable")
+    items = card.locator("[data-slot='card-content'] ul > li").all()
+
+    output = ["Timetable:\n"]
+    for item in items:
+        # Check if it's a plain text item like "Lunch break"
+        if item.locator("span").count() == 0:
+            output.append(item.inner_text().strip())
+        else:
+            parts = item.locator("span").all()
+            line = " ".join([part.inner_text().strip() for part in parts])
+            output.append(line)
+
+    return "\n".join(output).strip()
+
+
+# print(URL,USERNAME,PASSWORD)
+def get_homework() -> str:
+    URL,USERNAME,PASSWORD=load_credentials()
+    
+    with sync_playwright() as playwright:
+        browser = playwright.firefox.launch(headless=True)
+        page = browser.new_page()
+        page.goto(URL,wait_until="domcontentloaded")
+        
+        page.get_by_role('textbox',name='username').fill(USERNAME)
+        page.get_by_role('textbox',name='password').fill(PASSWORD)
+        page.get_by_role('button',name='login').click()
+        page.wait_for_url("**/dashboard")
+        # page.wait_for_timeout(1000)         
+        homework = extract_homework_text(page)
+        browser.close()        
+    return homework
+
+def get_timetable() -> str:
+    URL,USERNAME,PASSWORD=load_credentials()
+    
+    with sync_playwright() as playwright:
+        browser = playwright.firefox.launch(headless=True)
+        page = browser.new_page()
+        page.goto(URL,wait_until="domcontentloaded")
+        
+        page.get_by_role('textbox',name='username').fill(USERNAME)
+        page.get_by_role('textbox',name='password').fill(PASSWORD)
+        page.get_by_role('button',name='login').click()
+        page.wait_for_url("**/dashboard")
+        # page.wait_for_timeout(1000)        
+        timetable = extract_timetable_text(page)
+        browser.close()
+        
+    return timetable
+
+if __name__=="__main__":
+    print(get_homework())
+    print()
+    print(get_timetable())