now we use undetected-chromedriver

main.py

@@ -11,7 +11,7 @@
 from lib_resume_builder_AIHawk import Resume, FacadeManager, ResumeGenerator, StyleManager
 from typing import Optional
 from constants import PLAIN_TEXT_RESUME_YAML, SECRETS_YAML, WORK_PREFERENCES_YAML
-from src.utils.chrome_utils import chrome_browser_options
+from src.utils.chrome_utils import init_browser
 
 from src.job_application_profile import JobApplicationProfile
 from src.logging import logger
@@ -26,6 +26,7 @@
 from ai_hawk.bot_facade import AIHawkBotFacade
 from ai_hawk.job_manager import AIHawkJobManager
 from ai_hawk.llm.llm_manager import GPTAnswerer
+from ai_hawk.llm.llm_manager import GPTParser
 
 
 class ConfigError(Exception):
@@ -155,14 +156,6 @@ def file_paths_to_dict(resume_file: Path | None, plain_text_resume_file: Path) -
 
         return result
 
-def init_browser() -> webdriver.Chrome:
-    try:
-        options = chrome_browser_options()
-        service = ChromeService(ChromeDriverManager().install())
-        return webdriver.Chrome(service=service, options=options)
-    except Exception as e:
-        raise RuntimeError(f"Failed to initialize browser: {str(e)}")
-
 def create_and_run_bot(parameters, llm_api_key):
     try:
         style_manager = StyleManager()
@@ -182,9 +175,10 @@ def create_and_run_bot(parameters, llm_api_key):
         login_component = get_authenticator(driver=browser, platform='linkedin')
         apply_component = AIHawkJobManager(browser)
         gpt_answerer_component = GPTAnswerer(parameters, llm_api_key)
+undetected-chromedriver        gpt_parser_component = GPTParser(parameters, llm_api_key)
         bot = AIHawkBotFacade(login_component, apply_component)
         bot.set_job_application_profile_and_resume(job_application_profile_object, resume_object)
-        bot.set_gpt_answerer_and_resume_generator(gpt_answerer_component, resume_generator_manager)
+        bot.set_gpt_answerer_and_resume_generator(gpt_parser_component, gpt_answerer_component, resume_generator_manager)
         bot.set_parameters(parameters)
         bot.start_login()
         if (parameters['collectMode'] == True):

requirements.txt

@@ -28,3 +28,4 @@ webdriver-manager==4.0.2
 pytest
 pytest-mock
 pytest-cov
+undetected_chromedriver

src/ai_hawk/bot_facade.py

@@ -46,12 +46,13 @@ def set_job_application_profile_and_resume(self, job_application_profile, resume
         logger.debug("Job application profile and resume set successfully")
 
 
-    def set_gpt_answerer_and_resume_generator(self, gpt_answerer_component, resume_generator_manager):
+    def set_gpt_answerer_and_resume_generator(self, gpt_parser_component, gpt_answerer_component, resume_generator_manager):
         logger.debug("Setting GPT answerer and resume generator")
         self._ensure_job_profile_and_resume_set()
         gpt_answerer_component.set_job_application_profile(self.job_application_profile)
         gpt_answerer_component.set_resume(self.resume)
         self.apply_component.set_gpt_answerer(gpt_answerer_component)
+        self.apply_component.set_gpt_parser(gpt_parser_component)
         self.apply_component.set_resume_generator_manager(resume_generator_manager)
         self.state.gpt_answerer_set = True
         logger.debug("GPT answerer and resume generator set successfully")

src/ai_hawk/job_manager.py

@@ -10,7 +10,7 @@
 from inputimeout import inputimeout, TimeoutOccurred
 from selenium.common.exceptions import NoSuchElementException
 from selenium.webdriver.common.by import By
-
+from src.ai_hawk.llm.llm_manager import GPTAnswerer
 
 from ai_hawk.linkedIn_easy_applier import AIHawkEasyApplier
 from config import JOB_MAX_APPLICATIONS, JOB_MIN_APPLICATIONS, MINIMUM_WAIT_TIME_IN_SECONDS
@@ -82,6 +82,10 @@ def set_parameters(self, parameters):
     def set_gpt_answerer(self, gpt_answerer):
         logger.debug("Setting GPT answerer")
         self.gpt_answerer = gpt_answerer
+
+    def set_gpt_parser(self, gpt_parser):
+        logger.debug("Setting GPT parser")
+        self.gpt_parser = gpt_parser
 
     def set_resume_generator_manager(self, resume_generator_manager):
         logger.debug("Setting resume generator manager")
@@ -168,7 +172,7 @@ def start_applying(self):
                     try:
                         self.apply_jobs()
                     except Exception as e:
-                        logger.error(f"Error during job application: {e} {traceback.format_exc()}")
+                        logger.error(f"Error during job application: {e}")
                         continue
 
                     logger.debug("Applying to jobs on this page has been completed!")
@@ -481,18 +485,15 @@ def job_tile_to_job(self, job_tile) -> Job:
         # Extract company name and location
         try:
             # contains both with a delimter '·'
-            company_location = job_tile.find_element(
-                By.XPATH, ".//div[contains(@class, 'artdeco-entity-lockup__subtitle')]//span"
-            ).text
-            company, location = company_location.split('·')
+            company, location = self.gpt_parser.extract_company_and_title(job_tile.get_attribute("outerHTML"))
             job.company = company.strip()
             logger.debug(f"Job company extracted: {job.company}")
             job.location = location.strip()
             logger.debug(f"Job location extracted: {job.location}")
-        except ValueError:
+        except ValueError as e:
             logger.warning(f"Could not find the company and location. {e} {traceback.format.exc()}")
 
-        except NoSuchElementException:
+        except NoSuchElementException as e:
             logger.warning(f"Job comapy and location are missing. {e} {traceback.format.exc()}")
 
         # Extract job State

src/ai_hawk/llm/llm_manager.py

@@ -461,7 +461,7 @@ def parse_llmresult(self, llmresult: AIMessage) -> Dict[str, Dict]:
 class GPTAnswerer:
     def __init__(self, config, llm_api_key):
         self.ai_adapter = AIAdapter(config, llm_api_key)
-        self.llm_cheap = LoggerChatModel(self.ai_adapter)
+        self.llm = LoggerChatModel(self.ai_adapter)
 
     @property
     def job_description(self):
@@ -512,7 +512,7 @@ def summarize_job_description(self, text: str) -> str:
             prompts.summarize_prompt_template
         )
         prompt = ChatPromptTemplate.from_template(prompts.summarize_prompt_template)
-        chain = prompt | self.llm_cheap | StrOutputParser()
+        chain = prompt | self.llm | StrOutputParser()
         raw_output = chain.invoke({TEXT: text})
         output = self._clean_llm_output(raw_output)
         logger.debug(f"Summary generated: {output}")
@@ -521,7 +521,7 @@ def summarize_job_description(self, text: str) -> str:
     def _create_chain(self, template: str):
         logger.debug(f"Creating chain with template: {template}")
         prompt = ChatPromptTemplate.from_template(template)
-        return prompt | self.llm_cheap | StrOutputParser()
+        return prompt | self.llm | StrOutputParser()
 
     def answer_question_textual_wide_range(self, question: str) -> str:
         logger.debug(f"Answering textual question: {question}")
@@ -558,7 +558,7 @@ def answer_question_textual_wide_range(self, question: str) -> str:
         }
 
         prompt = ChatPromptTemplate.from_template(prompts.determine_section_template)
-        chain = prompt | self.llm_cheap | StrOutputParser()
+        chain = prompt | self.llm | StrOutputParser()
         raw_output = chain.invoke({QUESTION: question})
         output = self._clean_llm_output(raw_output)
 
@@ -615,7 +615,7 @@ def answer_question_numeric(
             prompts.numeric_question_template
         )
         prompt = ChatPromptTemplate.from_template(func_template)
-        chain = prompt | self.llm_cheap | StrOutputParser()
+        chain = prompt | self.llm | StrOutputParser()
         raw_output_str = chain.invoke(
             {
                 RESUME_EDUCATIONS: self.resume.education_details,
@@ -650,7 +650,7 @@ def answer_question_from_options(self, question: str, options: list[str]) -> str
         logger.debug(f"Answering question from options: {question}")
         func_template = self._preprocess_template_string(prompts.options_template)
         prompt = ChatPromptTemplate.from_template(func_template)
-        chain = prompt | self.llm_cheap | StrOutputParser()
+        chain = prompt | self.llm | StrOutputParser()
         raw_output_str = chain.invoke(
             {
                 RESUME: self.resume,
@@ -672,7 +672,7 @@ def resume_or_cover(self, phrase: str) -> str:
         prompt = ChatPromptTemplate.from_template(
             prompts.resume_or_cover_letter_template
         )
-        chain = prompt | self.llm_cheap | StrOutputParser()
+        chain = prompt | self.llm | StrOutputParser()
         raw_response = chain.invoke({PHRASE: phrase})
         response = self._clean_llm_output(raw_response)
         logger.debug(f"Response for resume_or_cover: {response}")
@@ -686,7 +686,7 @@ def resume_or_cover(self, phrase: str) -> str:
     def is_job_suitable(self):
         logger.info("Checking if job is suitable")
         prompt = ChatPromptTemplate.from_template(prompts.is_relavant_position_template)
-        chain = prompt | self.llm_cheap | StrOutputParser()
+        chain = prompt | self.llm | StrOutputParser()
         raw_output = chain.invoke(
             {
                 RESUME: self.resume,
@@ -707,3 +707,74 @@ def is_job_suitable(self):
         if int(score) < JOB_SUITABILITY_SCORE:
             logger.debug(f"Job is not suitable: {reasoning}")
         return int(score) >= JOB_SUITABILITY_SCORE
+
+import re
+import json
+import logging
+
+logger = logging.getLogger(__name__)
+
+class GPTParser:
+    def __init__(self, config, llm_api_key):
+        self.ai_adapter = AIAdapter(config, llm_api_key)
+        self.llm = LoggerChatModel(self.ai_adapter)
+
+    @staticmethod
+    def _clean_llm_output(output: str) -> str:
+        return output.replace("*", "").replace("#", "").strip()
+
+    @staticmethod
+    def _preprocess_template_string(template: str) -> str:
+        return textwrap.dedent(template)
+
+    def extract_company_and_title(self, html_content: str) -> dict:
+        """
+        Uses AI to extract the company name and job title from HTML code.
+
+        Args:
+            html_content (str): The HTML code to analyze.
+
+        Returns:
+            dict: A dictionary with 'company' and 'title' as keys.
+        """
+        logger.debug("Extracting company and title from HTML content.")
+
+        # AI prompt template
+        extract_prompt_template = """
+        You are an AI assistant extracting information from HTML code.
+        Extract the company name and job title from the following HTML code:
+
+        {html_content}
+
+        Provide the response in JSON format with keys "company" and "title", Provide only the exact JSON without any explanations or additional text and also without ```json ```
+
+        """
+
+        # Preprocess the template
+        extract_prompt_template = self._preprocess_template_string(extract_prompt_template)
+
+        # Create the prompt
+        prompt = ChatPromptTemplate.from_template(extract_prompt_template)
+
+        # Create the chain
+        chain = prompt | self.llm | StrOutputParser()
+
+        # Invoke the chain with the HTML
+        raw_output = chain.invoke({"html_content": html_content})
+
+        # Clean the output
+        output = self._clean_llm_output(raw_output)
+        logger.debug(f"Raw output from AI: {output}")
+
+        # Parse the JSON output
+        try:
+            result = json.loads(output)
+            company = result.get('company', 'Company not found')
+            title = result.get('title', 'Title not found')
+        except json.JSONDecodeError as e:
+            logger.error(f"JSON decoding failed: {e}")
+            company = 'Company not found'
+            title = 'Title not found'
+
+        logger.debug(f"Extracted company: {company}, title: {title}")
+        return company, title

src/utils/chrome_utils.py

@@ -1,5 +1,5 @@
 import os
-from selenium import webdriver
+import undetected_chromedriver as uc
 from src.logging import logger
 
 chromeProfilePath = os.path.join(os.getcwd(), "chrome_profile", "linkedin_profile")
@@ -18,7 +18,7 @@ def ensure_chrome_profile():
 def chrome_browser_options():
     logger.debug("Setting Chrome browser options")
     ensure_chrome_profile()
-    options = webdriver.ChromeOptions()
+    options = uc.ChromeOptions()
     options.add_argument("--start-maximized")
     options.add_argument("--no-sandbox")
     options.add_argument("--disable-dev-shm-usage")
@@ -37,7 +37,6 @@ def chrome_browser_options():
     options.add_argument("--disable-plugins")
     options.add_argument("--disable-animations")
     options.add_argument("--disable-cache")
-    options.add_experimental_option("excludeSwitches", ["enable-automation", "enable-logging"])
 
     prefs = {
         "profile.default_content_setting_values.images": 2,
@@ -57,4 +56,9 @@ def chrome_browser_options():
 
     return options
 
-
+def init_browser() -> uc.Chrome:
+    try:
+        options = chrome_browser_options()
+        return uc.Chrome(options=options)
+    except Exception as e:
+        raise RuntimeError(f"Failed to initialize browser: {str(e)}")