From 9979672b46eb36b5ab66505e243df23d3bb53c58 Mon Sep 17 00:00:00 2001
From: Joan <lgray.foxl@gmail.com>
Date: Mon, 6 Mar 2023 22:34:43 +0100
Subject: [PATCH] Added captcha solving method

---
 bot/Dockerfile       |  5 +++--
 bot/bot.py           | 31 ++++++++++++++++++++++++++++---
 bot/requirements.txt |  5 +++--
 3 files changed, 34 insertions(+), 7 deletions(-)

diff --git a/bot/Dockerfile b/bot/Dockerfile
index f55301f..7a6e618 100644
--- a/bot/Dockerfile
+++ b/bot/Dockerfile
@@ -1,7 +1,5 @@
 FROM python:3.7
 
-ENV TELEGRAM_API_KEY=5707028834:AAFlX07ObRGWmm15KtHgwqcPZ4OHy-MkMks
-
 # Adding trusting keys to apt for repositories
 RUN wget -q -O - https://dl-ssl.google.com/linux/linux_signing_key.pub | apt-key add -
 
@@ -12,12 +10,15 @@ RUN sh -c 'echo "deb [arch=amd64] http://dl.google.com/linux/chrome/deb/ stable
 RUN apt-get -y update
 
 # Magic happens
+#RUN wget -O /tmp/google-chrome-stable.deb http://mirror.cs.uchicago.edu/google-chrome/pool/main/g/google-chrome-stable/google-chrome-stable_108.0.5359.71-1_amd64.deb
+#RUN apt install -y /tmp/google-chrome-stable.deb
 RUN apt-get install -y google-chrome-stable
 
 # Installing Unzip
 RUN apt-get install -yqq unzip
 
 # Download the Chrome Driver
+#RUN wget -O /tmp/chromedriver.zip http://chromedriver.storage.googleapis.com/108.0.5359.71/chromedriver_linux64.zip
 RUN wget -O /tmp/chromedriver.zip http://chromedriver.storage.googleapis.com/`curl -sS chromedriver.storage.googleapis.com/LATEST_RELEASE`/chromedriver_linux64.zip
 
 # Unzip the Chrome Driver into /usr/local/bin directory
diff --git a/bot/bot.py b/bot/bot.py
index 94b61c6..c683b66 100644
--- a/bot/bot.py
+++ b/bot/bot.py
@@ -5,8 +5,12 @@ import dbhelper
 import constants
 import helpers
 import parser
+import time
 
+from amazoncaptcha import AmazonCaptcha
 from bs4 import BeautifulSoup
+from selenium.webdriver.common.keys import Keys
+from selenium.webdriver.common.by import By
 from selenium.webdriver.chrome.options import Options
 from selenium import webdriver
 from telegram import Update, ForceReply
@@ -56,15 +60,36 @@ def find_amazon_link(update: Update, context: CallbackContext) -> None:
         driver.get(msg)
         logging.info("Scraping information and closing browser")
         soup = BeautifulSoup(driver.page_source, "lxml")
-        driver.close()
 
         logging.info("Getting title...")
         title, price, image = parser.get_title(soup), parser.get_price(soup), parser.get_image(soup)
         if title == "":
-            logging.info(f"Title not found, not a valid product")
-            return
+            logging.info(f"Title not found, not a valid product or captcha")
+            captcha = AmazonCaptcha.fromdriver(driver)
+            solution = captcha.solve()
+            if solution == 'Not solved':
+                logging.info("Couldn't solve the captcha, if there was any")
+            else:
+                logging.info(f"Captcha solution is {solution}, redirecting")
+                logging.info("Waiting for 5 seconds, humans are not that fast :)")
+                time.sleep(5)
+                fill_captcha_element = driver.find_element(By.ID, 'captchacharacters')
+                fill_captcha_element.send_keys(solution)
+                fill_captcha_element.send_keys(Keys.RETURN)
+                logging.info("Re-loading Amazon webpage")
+                driver.get(msg)
+                logging.info("Scraping information")
+                soup = BeautifulSoup(driver.page_source, "lxml")
+                logging.info("Getting title...")
+                title, price, image = parser.get_title(soup), parser.get_price(soup), parser.get_image(soup)
+                if title == "":
+                    logging.info(f"Title not found, not a valid product or failed captcha")
+                    return
         logging.info(f"Title found: {title}")
 
+        logging.info("Closing browser")
+        driver.close()
+
         context.bot.deleteMessage(chat_id=chat['id'], message_id=message_id)
         referurl = helpers.new_refer_url(productCode, extraparams)
         product_id = dbhelper.check_product(referurl, price)
diff --git a/bot/requirements.txt b/bot/requirements.txt
index f8395ca..4d49785 100644
--- a/bot/requirements.txt
+++ b/bot/requirements.txt
@@ -2,5 +2,6 @@ python-telegram-bot==13.13
 requests==2.28.1
 beautifulsoup4==4.11.1
 lxml==4.9.1
-selenium==4.4.0
-Pillow==9.2.0
\ No newline at end of file
+selenium==4.2.0
+Pillow==9.4.0
+amazoncaptcha==0.5.9
\ No newline at end of file