Spaces:
Running
Running
Upload 3 files
Browse files- Dockerfile +7 -23
- app.py +35 -50
- requirements.txt +3 -3
Dockerfile
CHANGED
|
@@ -1,24 +1,6 @@
|
|
| 1 |
-
|
|
|
|
| 2 |
|
| 3 |
-
# Install system dependencies needed for Chrome and Selenium
|
| 4 |
-
ENV DEBIAN_FRONTEND=noninteractive
|
| 5 |
-
RUN apt-get update && apt-get install -y \
|
| 6 |
-
wget \
|
| 7 |
-
gnupg \
|
| 8 |
-
unzip \
|
| 9 |
-
curl \
|
| 10 |
-
--no-install-recommends \
|
| 11 |
-
&& rm -rf /var/lib/apt/lists/*
|
| 12 |
-
|
| 13 |
-
# Install Google Chrome Stable safely without hanging on prompts
|
| 14 |
-
RUN wget -q -O - https://dl.google.com/linux/linux_signing_key.pub | gpg --dearmor -o /usr/share/keyrings/google-chrome.gpg \
|
| 15 |
-
&& sh -c 'echo "deb [arch=amd64 signed-by=/usr/share/keyrings/google-chrome.gpg] http://dl.google.com/linux/chrome/deb/ stable main" >> /etc/apt/sources.list.d/google-chrome.list' \
|
| 16 |
-
&& apt-get update \
|
| 17 |
-
&& apt-get install -y google-chrome-stable \
|
| 18 |
-
--no-install-recommends \
|
| 19 |
-
&& rm -rf /var/lib/apt/lists/*
|
| 20 |
-
|
| 21 |
-
# Create user securely to run in HF Spaces instead of root
|
| 22 |
RUN useradd -m -u 1000 user
|
| 23 |
USER user
|
| 24 |
ENV HOME=/home/user \
|
|
@@ -27,15 +9,17 @@ ENV HOME=/home/user \
|
|
| 27 |
|
| 28 |
WORKDIR $HOME/app
|
| 29 |
|
| 30 |
-
# Install
|
| 31 |
COPY --chown=user requirements.txt .
|
| 32 |
RUN pip install --no-cache-dir -r requirements.txt
|
| 33 |
|
|
|
|
|
|
|
|
|
|
| 34 |
# Copy our backend script into the container
|
| 35 |
COPY --chown=user app.py .
|
| 36 |
|
| 37 |
-
# HF Spaces requires the container to expose port 7860
|
| 38 |
EXPOSE 7860
|
| 39 |
|
| 40 |
-
# Start Gradio
|
| 41 |
CMD ["python", "app.py"]
|
|
|
|
| 1 |
+
# Use the official Microsoft Playwright image. It has ALL complex system dependencies pre-installed!
|
| 2 |
+
FROM mcr.microsoft.com/playwright/python:v1.40.0-jammy
|
| 3 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 4 |
RUN useradd -m -u 1000 user
|
| 5 |
USER user
|
| 6 |
ENV HOME=/home/user \
|
|
|
|
| 9 |
|
| 10 |
WORKDIR $HOME/app
|
| 11 |
|
| 12 |
+
# Install our python requirements
|
| 13 |
COPY --chown=user requirements.txt .
|
| 14 |
RUN pip install --no-cache-dir -r requirements.txt
|
| 15 |
|
| 16 |
+
# Download the Chromium portable browser incredibly fast without apt-get!
|
| 17 |
+
RUN playwright install chromium
|
| 18 |
+
|
| 19 |
# Copy our backend script into the container
|
| 20 |
COPY --chown=user app.py .
|
| 21 |
|
|
|
|
| 22 |
EXPOSE 7860
|
| 23 |
|
| 24 |
+
# Start Gradio background daemon
|
| 25 |
CMD ["python", "app.py"]
|
app.py
CHANGED
|
@@ -2,92 +2,77 @@ import gradio as gr
|
|
| 2 |
import threading
|
| 3 |
import time
|
| 4 |
import schedule
|
| 5 |
-
from
|
| 6 |
-
from selenium.webdriver.chrome.options import Options
|
| 7 |
import random
|
| 8 |
import datetime
|
| 9 |
|
| 10 |
URL = "https://streamer1-gvbgwaea2w2qmwgjey7ksg.streamlit.app/"
|
| 11 |
-
logs = ["Service started.
|
| 12 |
|
| 13 |
def add_log(message):
|
| 14 |
timestamp = datetime.datetime.now().strftime('%Y-%m-%d %H:%M:%S')
|
| 15 |
logs.append(f"[{timestamp}] {message}")
|
| 16 |
-
# Keep the last 30 logs in memory to show on the UI
|
| 17 |
if len(logs) > 30:
|
| 18 |
logs.pop(0)
|
| 19 |
|
| 20 |
def visit_and_scroll():
|
| 21 |
add_log(f"Started keeping alive {URL}...")
|
| 22 |
|
| 23 |
-
options = Options()
|
| 24 |
-
options.add_argument("--headless=new")
|
| 25 |
-
# These flags are critically important for running Chrome inside a Docker container
|
| 26 |
-
options.add_argument("--no-sandbox")
|
| 27 |
-
options.add_argument("--disable-dev-shm-usage")
|
| 28 |
-
options.add_argument("--disable-gpu")
|
| 29 |
-
options.add_argument("--window-size=1920,1080")
|
| 30 |
-
|
| 31 |
-
driver = None
|
| 32 |
try:
|
| 33 |
-
|
| 34 |
-
|
| 35 |
-
|
| 36 |
-
|
| 37 |
-
|
| 38 |
-
|
| 39 |
-
|
| 40 |
-
|
| 41 |
-
|
| 42 |
-
|
| 43 |
-
|
| 44 |
-
|
| 45 |
-
|
| 46 |
-
|
| 47 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 48 |
|
| 49 |
-
|
| 50 |
-
|
| 51 |
-
scroll_step = random.randint(100, 500)
|
| 52 |
-
current_scroll -= scroll_step
|
| 53 |
-
if current_scroll < 0:
|
| 54 |
-
current_scroll = 0
|
| 55 |
-
driver.execute_script(f"window.scrollTo(0, {current_scroll});")
|
| 56 |
-
time.sleep(random.uniform(0.3, 1.2))
|
| 57 |
|
| 58 |
-
add_log("Done navigating! Closing background browser...")
|
| 59 |
-
|
| 60 |
except Exception as e:
|
| 61 |
add_log(f"Error occurred: {e}")
|
| 62 |
-
finally:
|
| 63 |
-
if driver:
|
| 64 |
-
time.sleep(2)
|
| 65 |
-
driver.quit()
|
| 66 |
|
| 67 |
def run_schedule():
|
| 68 |
-
# Initial run upon startup
|
| 69 |
visit_and_scroll()
|
| 70 |
-
|
| 71 |
-
# Schedule to run every 1 hour
|
| 72 |
schedule.every(1).hours.do(visit_and_scroll)
|
| 73 |
while True:
|
| 74 |
schedule.run_pending()
|
| 75 |
time.sleep(60)
|
| 76 |
|
| 77 |
-
# Start
|
| 78 |
thread = threading.Thread(target=run_schedule, daemon=True)
|
| 79 |
thread.start()
|
| 80 |
|
| 81 |
def get_logs():
|
| 82 |
return "\n".join(logs)
|
| 83 |
|
| 84 |
-
# Create a minimal Gradio UI specifically to satisfy Hugging Face Spaces port mapping
|
| 85 |
with gr.Blocks() as demo:
|
| 86 |
-
gr.Markdown("# 🤖 Keep-Alive
|
| 87 |
-
gr.Markdown(f"Currently maintaining a background connection to **{URL}** every 1 hour.")
|
| 88 |
output = gr.Textbox(label="Live System Logs", lines=15, value=get_logs(), interactive=False)
|
| 89 |
refresh_btn = gr.Button("Refresh Logs")
|
| 90 |
refresh_btn.click(fn=get_logs, outputs=output)
|
| 91 |
|
| 92 |
-
# Expose the port commonly used for HF Spaces deployments
|
| 93 |
demo.launch(server_name="0.0.0.0", server_port=7860)
|
|
|
|
| 2 |
import threading
|
| 3 |
import time
|
| 4 |
import schedule
|
| 5 |
+
from playwright.sync_api import sync_playwright
|
|
|
|
| 6 |
import random
|
| 7 |
import datetime
|
| 8 |
|
| 9 |
URL = "https://streamer1-gvbgwaea2w2qmwgjey7ksg.streamlit.app/"
|
| 10 |
+
logs = ["Service started. Waiting for next schedule..."]
|
| 11 |
|
| 12 |
def add_log(message):
|
| 13 |
timestamp = datetime.datetime.now().strftime('%Y-%m-%d %H:%M:%S')
|
| 14 |
logs.append(f"[{timestamp}] {message}")
|
|
|
|
| 15 |
if len(logs) > 30:
|
| 16 |
logs.pop(0)
|
| 17 |
|
| 18 |
def visit_and_scroll():
|
| 19 |
add_log(f"Started keeping alive {URL}...")
|
| 20 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 21 |
try:
|
| 22 |
+
with sync_playwright() as p:
|
| 23 |
+
# Playwright handles headless browsers inside Docker flawlessly
|
| 24 |
+
browser = p.chromium.launch(headless=True, args=['--no-sandbox', '--disable-dev-shm-usage'])
|
| 25 |
+
page = browser.new_page()
|
| 26 |
+
|
| 27 |
+
# Open the Streamlit app
|
| 28 |
+
page.goto(URL, wait_until='networkidle')
|
| 29 |
+
time.sleep(8)
|
| 30 |
+
|
| 31 |
+
total_height = page.evaluate("document.body.scrollHeight")
|
| 32 |
+
add_log("Scrolling down naturally...")
|
| 33 |
+
current_scroll = 0
|
| 34 |
+
while current_scroll < total_height:
|
| 35 |
+
scroll_step = random.randint(100, 500)
|
| 36 |
+
current_scroll += scroll_step
|
| 37 |
+
page.evaluate(f"window.scrollTo(0, {current_scroll});")
|
| 38 |
+
time.sleep(random.uniform(0.3, 1.2))
|
| 39 |
+
new_height = page.evaluate("document.body.scrollHeight")
|
| 40 |
+
if new_height > total_height:
|
| 41 |
+
total_height = new_height
|
| 42 |
+
|
| 43 |
+
add_log("Scrolling up naturally...")
|
| 44 |
+
while current_scroll > 0:
|
| 45 |
+
scroll_step = random.randint(100, 500)
|
| 46 |
+
current_scroll -= scroll_step
|
| 47 |
+
if current_scroll < 0:
|
| 48 |
+
current_scroll = 0
|
| 49 |
+
page.evaluate(f"window.scrollTo(0, {current_scroll});")
|
| 50 |
+
time.sleep(random.uniform(0.3, 1.2))
|
| 51 |
|
| 52 |
+
add_log("Done navigating! Closing Playwright browser...")
|
| 53 |
+
browser.close()
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 54 |
|
|
|
|
|
|
|
| 55 |
except Exception as e:
|
| 56 |
add_log(f"Error occurred: {e}")
|
|
|
|
|
|
|
|
|
|
|
|
|
| 57 |
|
| 58 |
def run_schedule():
|
|
|
|
| 59 |
visit_and_scroll()
|
|
|
|
|
|
|
| 60 |
schedule.every(1).hours.do(visit_and_scroll)
|
| 61 |
while True:
|
| 62 |
schedule.run_pending()
|
| 63 |
time.sleep(60)
|
| 64 |
|
| 65 |
+
# Start background thread
|
| 66 |
thread = threading.Thread(target=run_schedule, daemon=True)
|
| 67 |
thread.start()
|
| 68 |
|
| 69 |
def get_logs():
|
| 70 |
return "\n".join(logs)
|
| 71 |
|
|
|
|
| 72 |
with gr.Blocks() as demo:
|
| 73 |
+
gr.Markdown("# 🤖 Playwright Keep-Alive Status")
|
|
|
|
| 74 |
output = gr.Textbox(label="Live System Logs", lines=15, value=get_logs(), interactive=False)
|
| 75 |
refresh_btn = gr.Button("Refresh Logs")
|
| 76 |
refresh_btn.click(fn=get_logs, outputs=output)
|
| 77 |
|
|
|
|
| 78 |
demo.launch(server_name="0.0.0.0", server_port=7860)
|
requirements.txt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
-
|
| 2 |
-
schedule
|
| 3 |
-
gradio
|
|
|
|
| 1 |
+
playwright
|
| 2 |
+
schedule
|
| 3 |
+
gradio
|