TypeError: Cannot read property 'user' of undefined about user-instagram

[SuspiciousLookingOwl]

Seems like you need to be logged in to an account if you want this to work in VPN / VPS.

As mentioned by @Jacodom , if you are using VPN / VPS you also need to send a cookie header, the required cookie is sessionid, which can be obtained by logging in.

So I made a simple function to login using username and password, you can create a dummy account for this

const puppeteer = require("puppeteer");

const getSessionId = async (username: string, password: string): Promise<string> => {
	return new Promise((resolve, reject) => {
		(async () => {
			// Open Browser and Page, then go to login page
			const browser = await puppeteer.launch();
			const page = await browser.newPage();            
            await page.goto("https://www.instagram.com/accounts/login/");
            
            // Check login response
			page.on("response", (response) => {
				if (response.url() === "https://www.instagram.com/accounts/login/ajax/" && response.status() !== 200) reject(new Error("Failed to Login"));
			});
            
			//  Create listener and listen for received cookie, `page.on("response")` doesn't show `set-cookie` header for some reason
			const client = await page.target().createCDPSession();
			await client.send("Network.enable");
			client.on("Network.responseReceivedExtraInfo", async (response) => {
				if (!response || !response.headers) return;
				if (!("set-cookie" in response.headers) || !response.headers["set-cookie"].includes("sessionid")) return;
				const cookie = response.headers["set-cookie"];
				const sessionId = cookie.split("sessionid=")[1].split(";")[0];
				await browser.close();
				return resolve(sessionId);
            });
            
			// Wait for form to load, fill in username & password, then click Login button
			await page.waitForSelector(".Igw0E > .-MzZI:nth-child(1) > ._9GP1n > .f0n8F > ._2hvTZ");
			await page.click(".Igw0E > .-MzZI:nth-child(1) > ._9GP1n > .f0n8F > ._2hvTZ");
			await page.type(".Igw0E > .-MzZI:nth-child(1) > ._9GP1n > .f0n8F > ._2hvTZ", username);
			await page.type(".Igw0E > .-MzZI:nth-child(2) > ._9GP1n > .f0n8F > ._2hvTZ", password);
			await page.waitForSelector("#loginForm > .Igw0E > .Igw0E > .sqdOP > .Igw0E");
			await page.click("#loginForm > .Igw0E > .Igw0E > .sqdOP > .Igw0E");
		})();
	});  
};

You can probably just use axios and send request manually, but I'm too lazy to tinker with it (it also requires some headers and password encryption, etc.), so I just used puppeteer to simulate login, just to get the sessionid.

And when sending the request

const REQUEST_PARAMETERS = {
           method: 'GET',
           url: normalizePostUrl(shortcode),
           headers: {
               'user-agent': 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/83.0.4103.116 Safari/537.36',
               'authority': 'www.instagram.com',
               'cache-control': 'max-age=0',
               'Cookie': `sessionid=${await getSessionId(instagramUsername, instagramPassword)};`
           }
       };
       const GQL = await axios(REQUEST_PARAMETERS)
         .catch(reject);

Keep in mind that if you are trying to login too many times (even though it's a correct attempt with correct username & password), Instagram might lock you from login for few minutes, so it's better to store your sessionid in a persistent storage, probably .json file, database, or somewhere and take the value from your persistent storage if it exists instead of sending a request to instagram everytime you need to get sessionid.

Also sessionid seems to be lasts for 31536000 seconds (1 year)
, so if you don't want to deal with all of the hassle above, you can just login to your Instagram account on your current PC, get the sessionid, put it as a env variable or something, and use that as the sessionid cookie value when sending the request, and change it every year.

At the end, my code that I used for my Discord bot looks like this (it's in Typescript)

[daaain]

Yeah, on further testing and even trying scraping the HTML, it seems that Instagram has probably rate-limited these shared hosts as it's impossible to get a useful response back. So it's not an issue with this library per se, but I guess a more informative error message would be useful to implement.

[EdouardCourty]

Can you send me the username or post data you want to scrape ?

[EhsanFox]

ofcurse:

userInstagram("boycode1") // Same as getUserData()
        .then(result => {
            console.log(result);
            if(result.isPrivate) pageStatus = "<:lock:750353629046767728>";
            else pageStatus = "<:unlock:750353629092904970>";
            let rsMessaeg = new MessageEmbed()
                .setAuthor(result.fullName, result.profilePic, result.link)
                .setDescription("**Bio: **\n```"+result.biography+"```\n\n")
                .addField("Followers", result.subscribersCount, true)
                .addField("Post Counter", result.postsCount.toString(), true)
                .addField("Page Status", pageStatus, true);
            message.channel.send(rsMessaeg);
        })
        .catch(console.error);

i also tried another usernames but didn't work and i have to say it was working like for 1 hour and out of no where this error is now keep logging in console!

[EhsanFox]

also, I tried your code that u posted on the intro, it has the same error.... !

[EdouardCourty]

I think you hit some kind of rate limit with Instagram, or the IP range owned by your VPS provider is limited, I don't really know, but the only people who sent me this error used VPS.

[EhsanFox]

I Tried using on my computer and also using vpn, didn't changed for me

[EhsanFox]

also, just to remind you that i checked with diffrent vps and computers......

i don't think it's kind fo rate limit or something.....

[EdouardCourty]

Try without the VPN, it should work !

[EdouardCourty]

https://www.reddit.com/r/Windscribe/comments/dc6rnr/issue_with_instagram_when_connected_to_windscribe/
I think this is an IP restriction about VPNs, the only people that reported this issue used a VPN.

[EhsanFox]

it's still the same, I tried with vpn and with out that.....

[daaain]

This is weirdly also happening to me on Netlify when it's trying to build a Gatsby site, but works perfectly on my laptop on my home internet 🤷

3:27:36 PM: (node:3654) UnhandledPromiseRejectionWarning: TypeError: Cannot read property 'user' of undefined
3:27:36 PM:     at Promise (/opt/build/repo/node_modules/user-instagram/src/scrape.js:23:37)
3:27:36 PM:     at process._tickCallback (internal/process/next_tick.js:68:7)
3:27:36 PM: (node:3654) UnhandledPromiseRejectionWarning: Unhandled promise rejection. This error originated either by throwing inside of an async function without a catch block, or by rejecting a promise which was not handled with .catch(). (rejection id: 1)
3:27:36 PM: (node:3654) [DEP0018] DeprecationWarning: Unhandled promise rejections are deprecated. In the future, promise rejections that are not handled will terminate the Node.js process with a non-zero exit code.

[EdouardCourty]

The request simply does not return a valid JSON object and cannot be used by the module, weirdly this only happens to people that use VPNs or VPS

[arshizx]

need to login when i try using node-fetch, maybe instagram making redirect or something.

[luminous8]

I have the same issue, works well on my computer but doesn't work on my server.
Any idea where to host it? lambda?

[EdouardCourty]

Buy a Raspberry Pi xD

[EdouardCourty]

Yeah, on further testing and even trying scraping the HTML, it seems that Instagram has probably rate-limited these shared hosts as it's impossible to get a useful response back. So it's not an issue with this library per se, but I guess a more informative error message would be useful to implement.

You're right, I'll implement this asap !

[Jacodom]

Hello @EdouardCourty !

First of all thank you for this super useful library!

I was having the same issue that the other folks have been addressing with the deployable version of my site which is located in a vps server.
I tried to emulate the request that the library makes in the scrape.js file (line 14) from within the vps server with a curl command and I got no response whatsoever. Then I navigate to instagram.com with my chrome browser and I looked on if there was any other relevant header. I found a cookie header that was present in the browser but not in the scrape.js file so I added it to my curl command and finally I got a decent response.
So...A workaround could be adding this cookie header to the scrape.js file but I honestly don't believe is a sustainable solution. Also I found that the response I got has a bit different structure regarding the JSON fields.
I hope this information may be useful to address the issue.
Regards

[ali200465]

hi
your code is very nice and clear and useability, but i get this Error
" ERROR Cannot read property 'shortcode_media' or 'user' of undefined ".
i think instagram block my isp ip, so i suggest you that add api proxy to your project.

so thanks,

[ertemishakk]

Same error here. Anyone found an answer?