Getting list of likers for an instagram post - Python & Selenium

Question

I'm training to web crawling. To do so, I've challenged myself to get the list of all people having liked a post on instagram. My problem is that I'm stuck to the point where I only get the first 11 usernames of likers. I cannot find the right way to automate the scrolling process while getting the likes.

Here is my process in Jupyter Notebook (it doesn't work as a script yet):

from selenium import webdriver
import pandas as pd

driver = webdriver.Chrome()

driver.get('https://www.instagram.com/p/BuE82VfHRa6/')

userid_element = driver.find_elements_by_xpath('//*[@id="react-root"]/section/main/div/div/article/div[2]/section[2]/div/div/a')[0].click()

elems = driver.find_elements_by_xpath("//*[@id]/div/a")

users = []

for elem in elems:
    users.append(elem.get_attribute('title'))

print(users)

Do you guys have any idea?

Many thanks

KunduK · Answer 1 · 2019-02-25T18:12:57.550

Please try the following code and let me know if this work.

from selenium import webdriver
driver = webdriver.Chrome()

driver.get('https://www.instagram.com/p/BuE82VfHRa6/')

elems = driver.find_elements_by_xpath("//a[@class='FPmhX notranslate TlrDj']")

users = []

for elem in elems:
    users.append(elem.get_attribute('title'))
    print('Title : ' +elem.get_attribute('title'))

print(users)

output:-

Title : kyliejenner
Title : saturdayshade28
Title : worldmeetzboy
Title : mrokon
Title : addieisaac
Title : addieisaac
Title : amber_doerksen
Title : amber_doerksen
Title : addieisaac
Title : zayn6117
Title : amber_doerksen
Title : amber_doerksen
Title : worldmeetzboy
Title : worldmeetzboy
Title : razvanpopic1301
Title : johanna.trmn
Title : johanna.trmn
Title : johanna.trmn
Title : americ.av
Title : gabriellcostta1.0
Title : gabriellcostta1.0
Title : gabriellcostta1.0
Title : worldmeetzboy
Title : enactusepi
Title : enactusepi
[u'kyliejenner', u'saturdayshade28', u'worldmeetzboy', u'mrokon', u'addieisaac', u'addieisaac', u'amber_doerksen', u'amber_doerksen', u'addieisaac', u'zayn6117', u'amber_doerksen', u'amber_doerksen', u'worldmeetzboy', u'worldmeetzboy', u'razvanpopic1301', u'johanna.trmn', u'johanna.trmn', u'johanna.trmn', u'americ.av', u'gabriellcostta1.0', u'gabriellcostta1.0', u'gabriellcostta1.0', u'worldmeetzboy', u'enactusepi', u'enactusepi']

Thanks for your answer :) It's actually doing something but what I am looking for. When executing this script, I'm getting the usernames of the last comments. What I'm looking for is the list of all people having liked the post... I think the process is to click on this x-path (//*[@id]/div/a) and scrolling until the end of the list to get all the likes. — Grégoire de Kermel, Feb 25 '19 at 17:01
But when run this code it’s giving me all the users having linked the post. — KunduK, Feb 25 '19 at 18:06
By doing so, you get the list of the last likers (list on the right pan). But I'm looking to get the list of all 2,437,535 likers. When you click on the "2,437,535 Likes", there is a list with all likers' usernames. — Grégoire de Kermel, Feb 26 '19 at 10:03
@KunduK IS there a way to get full names of people that liked and commented on some posts on instagram? — taga, Jul 23 '19 at 09:39

score 1 · Accepted Answer · answered Feb 26 '19 at 09:34

I guess instagram site use liked user elements maximum 17.
so, this is one loop

get elements list from web
save to my list
scroll down for get new element
check, is this last scroll elements?

driver.get('https://www.instagram.com/p/BuE82VfHRa6/')

userid_element = driver.find_elements_by_xpath('//*[@id="react-root"]/section/main/div/div/article/div[2]/section[2]/div/div/a')[0].click()
time.sleep(2)

# here, you can see user list you want.
# you have to scroll down to download more data from instagram server.
# loop until last element with users table view height value.

users = []

height = driver.find_element_by_xpath("/html/body/div[3]/div/div[2]/div/div").value_of_css_property("padding-top")
match = False
while match==False:
    lastHeight = height

    # step 1
    elements = driver.find_elements_by_xpath("//*[@id]/div/a")

    # step 2
    for element in elements:
        if element.get_attribute('title') not in users:
            users.append(element.get_attribute('title'))

    # step 3
    driver.execute_script("return arguments[0].scrollIntoView();", elements[-1])
    time.sleep(1)

    # step 4
    height = driver.find_element_by_xpath("/html/body/div[3]/div/div[2]/div/div").value_of_css_property("padding-top")
    if lastHeight==height:
        match = True

print(users)
print(len(users))
driver.quit()

I test in near 100 liked post, and it worked.

Thank you very much for coming back to me. It's definitely the process to scroll to a list. But it doesn't work well in instagram for me, maybe the platform is good at detecting crawlers(?). I should check if it's possible to set the speed of scrolling to see it that could affect the result. Thanks anyway :) — Grégoire de Kermel, Feb 26 '19 at 10:41
@GrégoiredeKermel at least 0.1 second is required for normal operation in my laptop environment. than 2,500,000 items need 2.9 hour...... if i use **headless chrome driver**, i do not know if it is faster. or i do not know if [**this**](https://michaeljsanders.com/2017/05/12/scrapin-and-scrollin.html#when-a-sub-section-of-a-page-has-infinite-scroll) will help — linizio, Feb 26 '19 at 11:27
It helps thanks! I've noticed that they're preventing us from getting the total list of user. I was thinking that we modify the step 3 by putting the "time.sleep(0.2)" in a variable, execute it and then, adding a step 3bis like: "if the time.sleep is higher than 0.2, then scroll up by 50 pixels and then, start the process again until next overdelay or end". By doing so, I'm pretty sure it would allow us to get the full list. — Grégoire de Kermel, Feb 27 '19 at 11:46

score 0 · Answer 3 · edited Jun 23 '19 at 22:05

I wasn't able to get the code to work as posted in predicty's answer. Therefore I made the adaptation below and it gets me now ~500 likers per post.

def get_post_likers(shortcode):
    chrome = ch.initialize()
    chrome.get('https://www.instagram.com/p/' + shortcode + '/')
    chrome.execute_script("window.scrollTo(0, 1080)") 
    url = "/p/" + shortcode + "/liked_by/"
    time.sleep(2)
    like_link = chrome.find_element_by_xpath('//a[@href="'+url+'"]')
    like_link.click()
    time.sleep(2)
    users = []
    pb = chrome.find_element_by_xpath("//div[@role = 'dialog']/div[2]/div[1]/div[1]").value_of_css_property("padding-bottom")
    match = False
    while match==False:
        lastHeight = pb

        # step 1
        elements = chrome.find_elements_by_xpath("//*[@id]/div/a")
        # step 2
        for element in elements:
            if element.get_attribute('title') not in users:
                users.append(element.get_attribute('title'))
        # step 3
        chrome.execute_script("return arguments[0].scrollIntoView();", elements[-1])
        time.sleep(1)
        # step 4
        pb = chrome.find_element_by_xpath("//div[@role = 'dialog']/div[2]/div[1]/div[1]").value_of_css_property("padding-bottom")
        if lastHeight==pb or len(users) >= 1500:
            match = True
    return users

score 0 · Answer 4 · edited Jul 11 '20 at 01:24

This worked for me:

driver.get('https://www.instagram.com/p/BuE82VfHRa6/')

time.sleep(2)

userid_element = driver.find_element_by_xpath('//*[@id="react-root"]/section/main/div/div[1]/article/div[3]/section[2]/div/div[2]/button').click()
time.sleep(2)



elems = driver.find_elements_by_xpath("//a[@class='FPmhX notranslate TlrDj']")

users = []


for i in range(10):
    i += 1
    if(i%10) == 9 :
        driver.find_element_by_xpath('/html/body/div[4]/div/div/div[2]/div').click()
        actionChain.key_down(Keys.SPACE).key_up(Keys.SPACE).perform()

    print('/html/body/div[4]/div/div/div[2]/div/div/div['+str(i)+']/div[2]/div[1]/div/a')
    Title = driver.find_element_by_xpath('/html/body/div[4]/div/div/div[2]/div/div/div['+str(i)+']/div[2]/div[1]/div/a').get_attribute('title')
    users.append(Title)
    print('Title : ' + Title)

print(users)

Kamran Gasimov · Answer 5 · 2020-11-15T17:43:37.150

I tried all the solutions above, but none of them are working. I think they are outdated.

Instead, I wrote my own. It works perfectly in 2020.

This code goes to the "username" address and take the latest post in the profile and get the liked users.

def getPosts():
    hrefs_in_view = driver.find_elements_by_tag_name('a')
    # finding relevant hrefs
    hrefs_in_view = [elem.get_attribute('href') for elem in hrefs_in_view
         if '.com/p/' in elem.get_attribute('href')]
    
    return hrefs_in_view;





def getLikers(username,limit,post=1):
    driver.get('https://www.instagram.com/' + username)
    time.sleep(1)
    users=[]

    #Get Latest Post
    driver.get(getPosts()[post])
    
    time.sleep(2)
    #Open Dialog
    followersLinkX = driver.find_element_by_xpath('//button[@class="sqdOP yWX7d     _8A5w5    "]')
    followersLinkX.click()
    time.sleep(1)
    #Get Dialog
    xxx = driver.find_element_by_xpath('//div[@role="dialog"]/div[1]/div[2]/div[1]/div[1]')
    #Focus on and Scroll
    xxx.click()
    # step 3
    actionChain = webdriver.ActionChains(driver)

  

    count = 0
    
    while(count < limit):
        for i in range(1,1000):
            try:
                users.append("https://www.instagram.com/" + driver.find_element_by_xpath('//div[@role="dialog"]/div[1]/div[2]/div[1]/div[1]/div['+ str(i) +']/div[2]/div[1]/div[1]').text) 
                count+=1
            except:
                break
        actionChain.key_down(Keys.SPACE).key_up(Keys.SPACE).perform()
        time.sleep(0.5)  

    return users

For runing likers = getLikers("deirvlon",100,1)

I tried a few solutions here and your one does seem closer to what instagram's website looks like today. I'm unable to retrieve the actual user elements though, it appears that now there is a span and then an 'a' object. Can you review and advise on how to amend? I'm puzzled, thanks! — codesaurus, Nov 12 '20 at 18:46
Hi, I have checked again it's working. If it gives errors it can be due to the internet load speed. You can increase "time.sleep(0.5)" sleep times to over come this problem — Kamran Gasimov, Nov 15 '20 at 17:39
Sorry, i have forgotten to add "getPosts()" method. I think that is the problem. Now i updated. I hope it will work) — Kamran Gasimov, Nov 15 '20 at 17:45

Getting list of likers for an instagram post - Python & Selenium

5 Answers5

Linked