quick fix for LinkedIn Automation

[cjbbb]

The changed "job_manager.py" can work well to "applied job" pages.
Still need to change "linkedin_easy_applier.py". There is too much work. If someone has time today, they can continue working on this basis.

[AvizRepo]

I mean can you use sonnet or chatgpt do to this faster? I need some explaination, can help you with this

[cjbbb]

@surapuramakhil could you please help me test this version. Right now, it works well on my device.

[PetrShchukin]

Hey, thanks for the work! I tested it using python main.py --collect and encountered some issues. Please review the comments for further details.

[PetrShchukin]

While the code works, the scaffold-layout__list class appears to be closer to the target ul element than scaffold-layout__list-detail-container, making it more appropriate for selection.

[PetrShchukin]

There’s a potential unhandled exception here. If the previos line raised an exception and title_element is not found, this line will fail and raise a different exception rather than NoSuchElementException, causing the entire program to stop abruptly.
Suggestion: job.link = job_tile.find_element(By.CLASS_NAME, 'job-card-list__title--link').get_attribute('href').split('?')[0]

[PetrShchukin]

I've got job titles duplicated with this code: "job_title": "Senior Hardware Experience Designer\nSenior Hardware Experience Designer",
Suggestion: job.title = job_tile.find_element(By.CLASS_NAME, 'job-card-list__title--link').find_element(By.TAG_NAME, 'strong').text

[PetrShchukin]

for me job-card-container__metadata-wrapper is empty tag.

[PetrShchukin]

Company name and location are splited by a dot now. I've got Company name along with the Location when I did --collect.
Suggestion:

# Extract Company Name and Location
        try:
            full_text = job_tile.find_element(By.XPATH, ".//div[contains(@class, 'artdeco-entity-lockup__subtitle')]//span").text
            company, location = full_text.split('·')
            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 NoSuchElementException as e:
            logger.warning(f'Job company and location are missing. {e} {traceback.format_exc()}')

[cjbbb]

Thanks for all of your advices!!! I will try and apply

[cjbbb]

bad news.. my linkedin account is blocked

[feder-cr]

@cjbbb are you serious? Why?

[cjbbb]

@cjbbb are you serious? Why?

I have no idea. let me try to connect with them and release my account.

[feder-cr]

@cjbbb @surapuramakhil any update for this Pr?

[BeniaminC]

Needs a needs a filename sanitization method in job_application_saver.py

# job_application_saver.py
# Used in the event that the job title has invalid characters
def sanitize_filename(filename):
    # Remove invalid characters
    return re.sub(r'[<>:"/\\|?*\n]', '_', filename)

# job_application_saver.py
       dir_name = sanitize_filename(f"{job.id} - {job.company} {job.title}")
       dir_path = os.path.join(BASE_DIR, dir_name)
...

Job name and location are in the same element. Split the string and pick the respective element from the list

# job_manager.py
        # Extract Company Name
        try:
            job.company = job_tile.find_element(
                By.XPATH, ".//div[contains(@class, 'artdeco-entity-lockup__subtitle')]//span"
            ).text.split('·')[0].strip()
            logger.debug(f"Job company extracted: {job.company}")
        except NoSuchElementException as e:
            logger.warning(f"Job company is missing. {e} {traceback.format_exc()}")

        # Extract job Location
        try:
            job.location = job_tile.find_element(
                By.XPATH, ".//div[contains(@class, 'artdeco-entity-lockup__subtitle')]//span"
            ).text.split('·')[1].strip()
            logger.debug(f"Job location extracted: {job.location}")
        except NoSuchElementException:
            logger.warning("Job location is missing.")
...

Use a nested try block to first find text in the section. If no question is found, go to grandparent's span and try to find the question there.

# linkedIn_easy_applier.py
    def _find_and_handle_dropdown_question(self, job_context : JobContext, section: WebElement) -> bool:
        job_application = job_context.job_application
        # in the event that there is one question, the question is outside the subsection, so we need to find the
        # parents' text

        try:
            try:
                question = section.find_element(By.CLASS_NAME, 'fb-dash-form-element__label')
                question_text = question.text.lower()
            except NoSuchElementException:
                logger.debug(f"Unable to find subsection question, trying parent class...")
                # parent = hash, grandparent = hash + span texts
                grand_parent = section.find_element(By.XPATH, "../..")
                # find the elements with texts
                question = grand_parent.find_elements(By.TAG_NAME, "span")
                # combine the texts
                question = '\n'.join([question.text for question in question])
                question_text = question.lower()
...

It seems to be working perfectly fine now.

[BeniaminC]

There's an issue with unfollowing. Change the query.

    def _unfollow_company(self) -> None:
        try:
            logger.debug("Unfollowing company")
            follow_checkbox = self.driver.find_element(
                By.XPATH, "//label[@for='follow-company-checkbox']"
            )
            follow_checkbox.click()
        except Exception as e:
            logger.debug(f"Failed to unfollow company: {e}")

[jayceftg]

not usable again? linkedin changed ui again?

[surapuramakhil]

closing this PR