enhanced function of getting active users from the latest n posts for @ortetgafernando

get_active_users.py

def get_active_users(browser, username, posts, boundary, logger):
    """Returns a list with usernames who liked the latest n posts"""

    browser.get('https://www.instagram.com/' + username)
    sleep(2)

    total_posts = format_number(browser.find_element_by_xpath(
        "//span[contains(@class,'_t98z6')]//span").text)

    # if posts > total user posts, assume total posts
    if posts >= total_posts:
        # reaches all user posts
        posts = total_posts

    # click latest post
    browser.find_element_by_xpath(
        "(//div[contains(@class, '_si7dy')])[1]").click()

    active_users = []
    sc_rolled = 0
    start_time = time.time()
    too_many_requests = 0   #this will help to prevent misbehaviours when you request the list of active users repeatedly within less than 10 min of breaks

    message = (("~collecting the entire usernames from posts without a boundary!\n") if boundary is None else
                ("~collecting only the visible usernames from posts without scrolling at the boundary of zero..\n") if boundary==0 else
                 ("~collecting the usernames from posts with the boundary of {}\n".format(boundary)))
    # posts argument is the number of posts to collect usernames
    logger.info("Getting active users who liked the latest {} posts:\n  {}".format(posts, message))

    for count in range(0, posts):
        try:
            likes_count = format_number(browser.find_element_by_xpath(
                "//a[contains(@class, '_nzn1h')]/span").text)

            browser.find_element_by_xpath(
                "//a[contains(@class, '_nzn1h')]").click()
            sleep(4)

            dialog = browser.find_element_by_xpath(
                "//div[text()='Likes']/following-sibling::div")

            scroll_it = True
            try_again = 0

            while scroll_it != False and boundary!=0:
                scroll_it = browser.execute_script('''
                    var div = arguments[0];
                    if (div.offsetHeight + div.scrollTop < div.scrollHeight) {
                        div.scrollTop = div.scrollHeight;
                        return true;}
                    else {
                        return false;}
                    ''', dialog)
                
                if sc_rolled > 91 or too_many_requests>1:   #old value 100
                    logger.info("Too Many Requests sent! ~will sleep some :>")
                    sleep(600)
                    sc_rolled = 0
                    too_many_requests = 0 if too_many_requests>=1 else too_many_requests
                else:
                    sleep(4.2)   #old value 5.6
                    sc_rolled += 1

                tmp_list = browser.find_elements_by_xpath(
                                    "//a[contains(@class, '_2g7d5')]")
                if boundary is not None:
                    if len(tmp_list) >= boundary:
                        break

                if (scroll_it == False and
                        likes_count-1 > len(tmp_list)):
                    if ((boundary is not None and likes_count-1>boundary) or
                            boundary is None):
                            if try_again <= 1:   #you can increase the amount of tries here
                                logger.info("Cor! ~failed to get the desired amount of usernames, trying again!  |  post:{}  |  attempt: {}".format(posts, try_again+1))
                                try_again += 1
                                too_many_requests += 1
                                scroll_it = True
                                nap_it = 4 if try_again==0 else 7
                                sleep(nap_it)

            tmp_list = browser.find_elements_by_xpath(
                                "//a[contains(@class, '_2g7d5')]")
            logger.info("Post {}  |  Likers: found {}, catched {}".format(count+1, likes_count, len(tmp_list)))

        except NoSuchElementException:
            try:
                tmp_list = browser.find_elements_by_xpath(
                    "//div[contains(@class, '_3gwk6')]/a")
            except NoSuchElementException:
                logger.error('There is some error searching active users')

        if len(tmp_list) is not 0:
            for user in tmp_list:
                active_users.append(user.text)

        sleep(1)
        # if not reached posts(parameter) value, continue
        if count+1 != posts:
            try:
                # click next button
                browser.find_element_by_xpath(
                    "//a[@class='_3a693 coreSpriteRightPaginationArrow']"
                    "[text()='Next']").click()
            except:
                logger.error('Unable to go to next profile post')

    real_time = time.time()
    diff_in_minutes = int((real_time - start_time)/60)
    # delete duplicated users
    active_users = list(set(active_users))
    logger.info("Gathered total of {} unique active followers from the latest {} posts in {} minutes".format(len(active_users), posts, diff_in_minutes))

    return active_users