IG Video Downloader

[ashutoshkrris]

Will the IG Video Downloader work if the project is deployed on Heroku?

[Anubhab2002]

Yeah it works, only thing is, during deployment you need to add two buildpacks:
$ herokubuildpacks:add --index 1 https://github.com/heroku/heroku-buildpack-chromedriver
$ heroku buildpacks:add --index 2 https://github.com/heroku/heroku-buildpack-google-chrome
as mentioned here

checkout this for details.

[Anubhab2002]

This medium blog is a good one, specifically for python projects: deploy python selenium project on heroku

[ashutoshkrris]

Thanks a bunch! @Anubhab2002
Will follow the same
If I encounter some bugs, will discuss here

[Anubhab2002]

sure thing!

[Anubhab2002]

you are welcome @ashutoshkrris!!
By the way, you could now move instagram video downloader from upcoming features section to the features section in README.md

[ashutoshkrris]

Ahh Yes!
Forgot about that, thanks

[ashutoshkrris]

@Anubhab2002
If I keep chrome_options.headless=True, it's not able to fetch the video.
Is there any solution for that, as we neee to put --headless flag on Heroku

[Anubhab2002]

Just confirming, You added the buildpacks right?

[ashutoshkrris]

Yep, those things are set up correctly.
The line which is posing the problem is :

WebDriverWait(driver, 10).until(EC.presence_of_element_located((By.CLASS_NAME, "_5wCQW")))

Looks like this doesn't work with headless chrome

[Anubhab2002]

yeah basically it is not allowing the pop-up screen to open for the scrapping to take place...anyways wait a bit. Let me try I can can refactor it a bit.

[ashutoshkrris]

Tried taking screenshot of the page in headless mode. This is what I got :

[Anubhab2002]

So is the issue fixed now?

[ashutoshkrris]

Nope
Also, I found..on local environment, it's not asking to login
But on heroku, Instagram asks to login

[Anubhab2002]

Hey @ashutoshkrris, it worked just fine on my local

[ashutoshkrris]

Yeah, that is working fine
But, on the deployed project, it's asking to login

[Anubhab2002]

but after logging in are we getting our job done?

[ashutoshkrris]

Most probably, it should
But I am not able to login programmatically

I am using this code :

if 'login' in driver.current_url:
    driver.implicitly_wait(20)
    driver.find_element_by_xpath('//*[@id="loginForm"]/div/div[1]/div/label/input').send_keys('username')
    password=driver.find_element_by_xpath('//*[@id="loginForm"]/div/div[2]/div/label/input')
    password.send_keys('pwd')
    password.send_keys(Keys.ENTER)

Error :

Message: no such element: Unable to locate element: {"method":"xpath","selector":"//*[@id="loginForm"]/div/div[1]/div/label/input"}

[Anubhab2002]

your code is alright, instead of implicit wait, try adding a sleep() before each of username, password, click. like maybe a sleep(5) before each action is performed

[Anubhab2002]

this issue arises can arise when the element has not loaded yet and you are trying to access it, so maybe try increasing the sleep time

[Anubhab2002]

@ashutoshkrris working now?

[ashutoshkrris]

No, I tried with this code, still failing

            driver.get(url)
            print("Page opened")
            print(driver.current_url)
            time.sleep(10)
            if 'login' in driver.current_url:
                driver.find_element_by_name(
                    'username').send_keys(IG_USERNAME)
                time.sleep(2)
                password = driver.find_element_by_name(
                    'password')
                time.sleep(2)
                password.send_keys(IG_PASSWORD)
                password.send_keys(Keys.ENTER)
                driver.get(url)

The first driver.get(url) can lead us to login page. So after logging in, we are using driver.get(url) again to get the correct page. But its opening the login page

[Anubhab2002]

I see. Well this is going beyond my understanding now. Best is to try asking on Stack Overflow.

[Anubhab2002]

@ashutoshkrris can stopping redirect in driver.get(url) and specifying the acceptable forms of the url help in this case?

[ashutoshkrris]

Sorry, didn't get you

[Anubhab2002]

as far as I understood the problem, when you are putting the url in the ig video downloader, the link is redirecting you to the login page. So in that case if we can somehow stop the redirect then can we get the viable results?

[ashutoshkrris]

We cannot do it using Selenium, i suppose

[Anubhab2002]

@ashutoshkrris , what is the present scenario?

[ashutoshkrris]

Couldn't figure out anything, so I have put the feature on hold.

[Anubhab2002]

I see. I am sorry, I tried as much as I could. Still I will look for alternatives.

[ashutoshkrris]

Yeah, I'm also trying.
I suppose Instagram is very strict regarding headless bots.