Spaces:
Sleeping
Sleeping
Sonu Prasad commited on
Commit ·
2861cb3
1
Parent(s): 84f0d92
Use Chromium instead of Chrome for compatibility
Browse files- Dockerfile +18 -32
- bot.py +25 -13
- requirements.txt +0 -2
Dockerfile
CHANGED
|
@@ -4,40 +4,34 @@ FROM python:3.11-slim
|
|
| 4 |
# Set environment variables
|
| 5 |
ENV PYTHONDONTWRITEBYTECODE=1
|
| 6 |
ENV PYTHONUNBUFFERED=1
|
| 7 |
-
ENV
|
|
|
|
| 8 |
|
| 9 |
-
# Install
|
| 10 |
RUN apt-get update && apt-get install -y --no-install-recommends \
|
|
|
|
|
|
|
| 11 |
wget \
|
| 12 |
gnupg \
|
| 13 |
-
|
| 14 |
-
|
| 15 |
-
libgconf-2-4 \
|
| 16 |
-
libxss1 \
|
| 17 |
-
libnss3 \
|
| 18 |
-
libnspr4 \
|
| 19 |
libasound2 \
|
| 20 |
-
libatk1.0-0 \
|
| 21 |
libatk-bridge2.0-0 \
|
|
|
|
| 22 |
libcups2 \
|
|
|
|
| 23 |
libdrm2 \
|
| 24 |
-
|
|
|
|
|
|
|
|
|
|
| 25 |
libxcomposite1 \
|
| 26 |
libxdamage1 \
|
|
|
|
|
|
|
| 27 |
libxrandr2 \
|
| 28 |
-
|
| 29 |
-
|
| 30 |
-
libcairo2 \
|
| 31 |
-
fonts-liberation \
|
| 32 |
-
ca-certificates \
|
| 33 |
-
curl \
|
| 34 |
-
&& rm -rf /var/lib/apt/lists/*
|
| 35 |
-
|
| 36 |
-
# Install Chrome
|
| 37 |
-
RUN wget -q -O /tmp/chrome.deb https://dl.google.com/linux/direct/google-chrome-stable_current_amd64.deb \
|
| 38 |
-
&& apt-get update \
|
| 39 |
-
&& apt-get install -y --no-install-recommends /tmp/chrome.deb \
|
| 40 |
-
&& rm /tmp/chrome.deb \
|
| 41 |
&& rm -rf /var/lib/apt/lists/*
|
| 42 |
|
| 43 |
# Create app directory
|
|
@@ -51,15 +45,7 @@ RUN pip install --no-cache-dir -r requirements.txt
|
|
| 51 |
COPY . .
|
| 52 |
|
| 53 |
# Create directories for uploads with proper permissions
|
| 54 |
-
RUN mkdir -p documents uploads && chmod
|
| 55 |
-
|
| 56 |
-
# Create a non-root user for HuggingFace
|
| 57 |
-
RUN useradd -m -u 1000 user
|
| 58 |
-
USER user
|
| 59 |
-
ENV HOME=/home/user
|
| 60 |
-
ENV PATH=/home/user/.local/bin:$PATH
|
| 61 |
-
|
| 62 |
-
WORKDIR /app
|
| 63 |
|
| 64 |
# Expose port for HuggingFace Spaces
|
| 65 |
EXPOSE 7860
|
|
|
|
| 4 |
# Set environment variables
|
| 5 |
ENV PYTHONDONTWRITEBYTECODE=1
|
| 6 |
ENV PYTHONUNBUFFERED=1
|
| 7 |
+
ENV CHROME_BIN=/usr/bin/chromium
|
| 8 |
+
ENV CHROMEDRIVER_PATH=/usr/bin/chromedriver
|
| 9 |
|
| 10 |
+
# Install Chromium and dependencies
|
| 11 |
RUN apt-get update && apt-get install -y --no-install-recommends \
|
| 12 |
+
chromium \
|
| 13 |
+
chromium-driver \
|
| 14 |
wget \
|
| 15 |
gnupg \
|
| 16 |
+
ca-certificates \
|
| 17 |
+
fonts-liberation \
|
|
|
|
|
|
|
|
|
|
|
|
|
| 18 |
libasound2 \
|
|
|
|
| 19 |
libatk-bridge2.0-0 \
|
| 20 |
+
libatk1.0-0 \
|
| 21 |
libcups2 \
|
| 22 |
+
libdbus-1-3 \
|
| 23 |
libdrm2 \
|
| 24 |
+
libgbm1 \
|
| 25 |
+
libgtk-3-0 \
|
| 26 |
+
libnspr4 \
|
| 27 |
+
libnss3 \
|
| 28 |
libxcomposite1 \
|
| 29 |
libxdamage1 \
|
| 30 |
+
libxfixes3 \
|
| 31 |
+
libxkbcommon0 \
|
| 32 |
libxrandr2 \
|
| 33 |
+
xdg-utils \
|
| 34 |
+
&& apt-get clean \
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 35 |
&& rm -rf /var/lib/apt/lists/*
|
| 36 |
|
| 37 |
# Create app directory
|
|
|
|
| 45 |
COPY . .
|
| 46 |
|
| 47 |
# Create directories for uploads with proper permissions
|
| 48 |
+
RUN mkdir -p documents uploads && chmod 777 documents uploads
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 49 |
|
| 50 |
# Expose port for HuggingFace Spaces
|
| 51 |
EXPOSE 7860
|
bot.py
CHANGED
|
@@ -1,7 +1,8 @@
|
|
| 1 |
"""
|
| 2 |
-
Practice Fusion Bot - Headless
|
| 3 |
"""
|
| 4 |
|
|
|
|
| 5 |
import time
|
| 6 |
import logging
|
| 7 |
from pathlib import Path
|
|
@@ -11,7 +12,6 @@ from selenium import webdriver
|
|
| 11 |
from selenium.webdriver.chrome.service import Service
|
| 12 |
from selenium.webdriver.chrome.options import Options
|
| 13 |
from selenium.webdriver.common.by import By
|
| 14 |
-
from webdriver_manager.chrome import ChromeDriverManager
|
| 15 |
|
| 16 |
import config
|
| 17 |
from utils import (
|
|
@@ -23,7 +23,7 @@ logger = logging.getLogger(__name__)
|
|
| 23 |
|
| 24 |
|
| 25 |
class PracticeFusionBot:
|
| 26 |
-
"""Headless
|
| 27 |
|
| 28 |
def __init__(self, headless: bool = True):
|
| 29 |
self.headless = headless
|
|
@@ -34,23 +34,35 @@ class PracticeFusionBot:
|
|
| 34 |
self._setup_driver()
|
| 35 |
|
| 36 |
def _setup_driver(self):
|
| 37 |
-
"""Initialize
|
| 38 |
chrome_options = Options()
|
| 39 |
|
| 40 |
-
|
| 41 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 42 |
|
| 43 |
-
#
|
| 44 |
-
|
| 45 |
-
|
| 46 |
-
|
| 47 |
-
chrome_options.add_experimental_option("useAutomationExtension", False)
|
| 48 |
|
| 49 |
try:
|
| 50 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 51 |
self.driver = webdriver.Chrome(service=service, options=chrome_options)
|
| 52 |
self.driver.set_page_load_timeout(config.PAGE_LOAD_TIMEOUT)
|
| 53 |
-
self.logger.info("
|
| 54 |
except Exception as e:
|
| 55 |
self.logger.error(f"Failed to initialize driver: {e}")
|
| 56 |
raise
|
|
|
|
| 1 |
"""
|
| 2 |
+
Practice Fusion Bot - Headless Chromium Automation
|
| 3 |
"""
|
| 4 |
|
| 5 |
+
import os
|
| 6 |
import time
|
| 7 |
import logging
|
| 8 |
from pathlib import Path
|
|
|
|
| 12 |
from selenium.webdriver.chrome.service import Service
|
| 13 |
from selenium.webdriver.chrome.options import Options
|
| 14 |
from selenium.webdriver.common.by import By
|
|
|
|
| 15 |
|
| 16 |
import config
|
| 17 |
from utils import (
|
|
|
|
| 23 |
|
| 24 |
|
| 25 |
class PracticeFusionBot:
|
| 26 |
+
"""Headless Chromium bot for Practice Fusion document uploads."""
|
| 27 |
|
| 28 |
def __init__(self, headless: bool = True):
|
| 29 |
self.headless = headless
|
|
|
|
| 34 |
self._setup_driver()
|
| 35 |
|
| 36 |
def _setup_driver(self):
|
| 37 |
+
"""Initialize Chromium WebDriver with headless options."""
|
| 38 |
chrome_options = Options()
|
| 39 |
|
| 40 |
+
# Headless options
|
| 41 |
+
chrome_options.add_argument("--headless=new")
|
| 42 |
+
chrome_options.add_argument("--no-sandbox")
|
| 43 |
+
chrome_options.add_argument("--disable-dev-shm-usage")
|
| 44 |
+
chrome_options.add_argument("--disable-gpu")
|
| 45 |
+
chrome_options.add_argument("--window-size=1920,1080")
|
| 46 |
+
chrome_options.add_argument("--disable-extensions")
|
| 47 |
+
chrome_options.add_argument("--disable-blink-features=AutomationControlled")
|
| 48 |
+
chrome_options.add_argument("--remote-debugging-port=9222")
|
| 49 |
|
| 50 |
+
# Set binary location for Chromium
|
| 51 |
+
chrome_bin = os.environ.get("CHROME_BIN", "/usr/bin/chromium")
|
| 52 |
+
if os.path.exists(chrome_bin):
|
| 53 |
+
chrome_options.binary_location = chrome_bin
|
|
|
|
| 54 |
|
| 55 |
try:
|
| 56 |
+
# Use system chromedriver
|
| 57 |
+
chromedriver_path = os.environ.get("CHROMEDRIVER_PATH", "/usr/bin/chromedriver")
|
| 58 |
+
if os.path.exists(chromedriver_path):
|
| 59 |
+
service = Service(chromedriver_path)
|
| 60 |
+
else:
|
| 61 |
+
service = Service()
|
| 62 |
+
|
| 63 |
self.driver = webdriver.Chrome(service=service, options=chrome_options)
|
| 64 |
self.driver.set_page_load_timeout(config.PAGE_LOAD_TIMEOUT)
|
| 65 |
+
self.logger.info("Chromium driver initialized successfully")
|
| 66 |
except Exception as e:
|
| 67 |
self.logger.error(f"Failed to initialize driver: {e}")
|
| 68 |
raise
|
requirements.txt
CHANGED
|
@@ -1,6 +1,4 @@
|
|
| 1 |
flask==3.0.0
|
| 2 |
flask-cors==4.0.0
|
| 3 |
selenium==4.16.0
|
| 4 |
-
webdriver-manager==4.0.1
|
| 5 |
python-dotenv==1.0.0
|
| 6 |
-
gunicorn==21.2.0
|
|
|
|
| 1 |
flask==3.0.0
|
| 2 |
flask-cors==4.0.0
|
| 3 |
selenium==4.16.0
|
|
|
|
| 4 |
python-dotenv==1.0.0
|
|
|