DevTools Google Chrome:
On this site (https://booyah.live/users/41874362/followers), to load the complete list of followers it is necessary to keep scrolling down the page to reload more profiles, but there comes a time when the page weighs so much that the browser crashes and it ends up needing to be closed.
Is there any way to be able to collect the follow buttons without this happening?
The current script I use is:
setInterval(function(){
document.getElementById("layout-content").scrollTo(0, 50000000000000000000000000000000000000);
document.querySelectorAll('.components-button.components-button-size-mini.components-button-type-orange.desktop.components-button-inline').forEach(btn => btn.click());
}, 10)
I use setInterval to create a loop of:
1 - Scrolling the page
2 - Loading more profiles
3 - Clicking the follow buttons
My need:
For the study I'm doing for learning, the idea is that my profile follows all profiles followers of a single most famous profile in order to analyze how many people follow back on this social media.
Additional:
In this answer provided by Leftium, it is possible to follow only one profile:
https://stackoverflow.com/a/67882688/11462274
In this answer given by KCGD, it is possible to collect the entire list of followers but during this collection the profiles are not followed, it is possible to create a list and save the data, but not follow the profiles:
https://stackoverflow.com/a/67865968/11462274
I tried to contact them both, but they haven't returned yet. It was a good way but I couldn't combine the two answers so I can follow all the profiles, I thought about the possibility according to which I would collect the profiles of the KCGD response, I would follow the profiles too, but not only the first one but also the answer of the Leftium.
Would it be possible to take advantage of the loop created by the response from KCGD and from each response, already follow all profiles instead of just the first one as in Leftium's response?
I tried to create but was unsuccessful.
The browser crashes because too much memory is used. As you scroll down the page, the HTML DOM tree is extended and more avatar images are downloaded. These HTML and image resources are not necessary for your goal.
It is possible to avoid crashing by calling the (internal) Booyah API directly. This will be much faster and consume less resources since only the text is transferred. There are two API endpoints of interest:
GET /api/v3/users/[USERID]/followers?cursor=0&count=100
Gets list of followers following a certain user:
[USERID] is the ID of the user being studied (WEEDZAO's id).
cursor is where in the list of followers to start listing. When the page first loads, this is 0. As you scroll down, the following API calls increment this (101, 201, 301...)
count is how many results to return.
Since this is a GET call, you can open this URL in your browser.
POST /api/v3/users/[USERID]/followings
Follows a user (same as clicking their 'Follow' button).
Here [USERID] is ID of the user whose follower list will be updated (your own ID).
A payload must be sent that looks like this: {followee_uid: ID, source: 43}. I'm not sure what source is.
Also a CSRF header must be included.
Because this is a POST type call, it is not possible to open this type of URL directly in your browser.
DELETE /api/v3/users/[USERID]/followings
There is also an API to unfollow a user. (Just for reference).
If you call these API's from outside the browser, you probably need to send session cookies.
This script will list WEEDZAO's first 10 followers, then follow the first one from the list:
You must replace USERID and CSRF_TOKEN with your own values.
You can copy/paste this code into the browser dev console.
Alternatively, you can use this code from a web scraping framework like Puppeteer.
// Find these values in dev console "Network" tab:
var CSRF_TOKEN, USERID, USERID_TARGET, main;
USERID_TARGET = '41874362';
USERID = '12345678';
CSRF_TOKEN = 'MTYy...p8wg';
main = async function() {
var body, followers, json, options, payload, response, url;
// Get list of first 10 followers
console.log(url = `/api/v3/users/${USERID_TARGET}/followers?cursor=0&count=10`);
response = (await fetch(url));
json = (await response.json());
followers = json.follower_list;
console.table(followers);
// Follow first member from list above
console.log(url = `/api/v3/users/${USERID}/followings`);
payload = JSON.stringify({
followee_uid: followers[0].uid,
source: 43
});
response = (await fetch(url, options = {
method: 'POST',
body: payload,
headers: {
'X-CSRF-Token': CSRF_TOKEN
}
}));
body = (await response.text());
return console.log(body);
};
main();
It crashes because the interval is too fast
setInterval(function(){}, 10)
you are trying to call a scroll and click function every 10 milliseconds (that's 100 function call every 1 second). Which also interferes with the server as they fetch new users while scrolling.
Your script could work if you will adjust the interval to atleast 1000 milliseconds (1 second). Of course, it may take a while, but it will work. You should also expect that the page may become laggy specially when the page already loaded tons of users because Virtual Scrolling is not implemented in this page.
Even with slowing down the rate of the scrolling it still really bogs down the browser, the solution to this may be in the API the page contacts. To get the user's followers it contacts the site's V3 API
https://booyah.live/api/v3/users/41874362/followers?cursor=[LAST USER IN API RETURN]&count=100
to get all the users that would show up in the page. I wrote a script that can contact the api over and over again to get all the follower data, just run it in the page's console and use print() when you want to export the data
and copy/paste it into a .json file
//WARNING: THIS SCRIPT USES RECURSION, i have no clue how long the followers list goes so use at your own risk
var followers = []; //data collected from api
function getFollowers(cursor){
httpGet(`https://booyah.live/api/v3/users/41874362/followers?cursor=${cursor}&count=100`, function (data) { //returns data from API for given cursor (user at the end of last follower chunk)
console.log("got cursor: "+cursor);
var _followChunk = JSON.parse(String(data));
console.log(_followChunk)
followers.push(_followChunk.follower_list); //saves followers from chunk
var last_user = _followChunk.follower_list[_followChunk.follower_list.length - 1]; //gets last user of chunk (cursor for the next chunk)
setTimeout(function(){ //1 second timeout so that the API doesnt return "too many requests", not nessicary but you should probably leave this on
getFollowers(last_user.uid); //get next chunk
},1000)
})
}
var print = function(){console.log(JSON.stringify(followers))};
getFollowers(0); //get initial set of followers (cursor 0)
function httpGet(theUrl, callback) {
var xmlHttp = new XMLHttpRequest();
xmlHttp.open("GET", theUrl, false); // false for synchronous request
xmlHttp.setRequestHeader("Cache-Control", "no-store");
xmlHttp.send(null);
callback(xmlHttp.responseText);
};
if you really only need the button elements then the only way is to scroll all the way down for each time it loads new followers, as the page creates the elements as you scroll down
This is a fully working solution that I have tested in my own Chrome browser with a fresh account, successfully following all the follower accounts of the account you are targeting.
UPDATE (2021-06-18)
I've updated my solution to a drastically improved and faster function, rewritten with async/await. This new function reduces the estimated runtime from ~45min to ~10min. 10min is still a long while, but that's to be expected considering the large number of followers the user you are targeting has.
After a few iterations, the latest function not only improves speed, performance, and error reporting, but it also extends what is possible with the function. I provide several example below my solutions of how to use the function completely.
For the sake of de-cluttering my answer, I am removing my older function from this solution altogether, but you can still reference it in my solution's edit history if you like.
TL;DR
Here is the final, fastest, working solution. Make sure to replace PUT_YOUR_CSRF_TOKEN_HERE with your own CSRF token value. Detailed instructions on how to find your CSRF token are below.
You must run this in your console on the Booyah website in order to avoid CORS issues.
const csrf = 'PUT_YOUR_CSRF_TOKEN_HERE';
async function booyahGetAccounts(uid, type = 'followers', follow = 1) {
if (typeof uid !== 'undefined' && !isNaN(uid)) {
const loggedInUserID = window.localStorage?.loggedUID;
if (uid === 0) uid = loggedInUserID;
const unfollow = follow === -1;
if (unfollow) follow = 1;
if (loggedInUserID) {
if (csrf) {
async function getUserData(uid) {
const response = await fetch(`https://booyah.live/api/v3/users/${uid}`),
data = await response.json();
return data.user;
}
const loggedInUserData = await getUserData(loggedInUserID),
targetUserData = await getUserData(uid),
followUser = uid => fetch(`https://booyah.live/api/v3/users/${loggedInUserID}/followings`, { method: (unfollow ? 'DELETE' : 'POST'), headers: { 'X-CSRF-Token': csrf }, body: JSON.stringify({ followee_uid: uid, source: 43 }) }),
logSep = (data = '', usePad = 0) => typeof data === 'string' && usePad ? console.log((data ? data + ' ' : '').padEnd(50, '━')) : console.log('━'.repeat(50),data,'━'.repeat(50));
async function getList(uid, type, follow) {
const isLoggedInUser = uid === loggedInUserID;
if (isLoggedInUser && follow && !unfollow && type === 'followings') {
follow = 0;
console.warn('You alredy follow your followings. `follow` mode switched to `false`. Followings will be retrieved instead of followed.');
}
const userData = await getUserData(uid),
totalCount = userData[type.slice(0,-1)+'_count'] || 0,
totalCountStrLength = totalCount.toString().length;
if (totalCount) {
let userIDsLength = 0;
const userIDs = [],
nickname = userData.nickname,
nicknameStr = `${nickname ? ` of ${nickname}'s ${type}` : ''}`,
alreadyFollowedStr = uid => `User ID ${uid} already followed by ${loggedInUserData.nickname} (Account #${loggedInUserID})`;
async function followerFetch(cursor = 0) {
const fetched = [];
await fetch(`https://booyah.live/api/v3/users/${uid}/${type}?cursor=${cursor}&count=100`).then(res => res.json()).then(data => {
const list = data[type.slice(0,-1)+'_list'];
if (list?.length) fetched.push(...list.map(e => e.uid));
if (fetched.length) {
userIDs.push(...fetched);
userIDsLength += fetched.length;
if (follow) followUser(uid);
console.log(`${userIDsLength.toString().padStart(totalCountStrLength)} (${(userIDsLength / totalCount * 100).toFixed(4)}%)${nicknameStr} ${follow ? 'followed' : 'retrieved'}`);
if (fetched.length === 100) {
followerFetch(data.cursor);
} else {
console.log(`END REACHED. ${userIDsLength} accounts ${follow ? 'followed' : 'retrieved'}.`);
if (!follow) logSep(targetList);
}
}
});
}
await followerFetch();
return userIDs;
} else {
console.log(`This account has no ${type}.`);
}
}
logSep(`${follow ? 'Following' : 'Retrieving'} ${targetUserData.nickname}'s ${type}`, 1);
const targetList = await getList(uid, type, follow);
} else {
console.error('Missing CSRF token. Retrieve your CSRF token from the Network tab in your inspector by clicking into the Network tab item named "bug-report-claims" and then scrolling down in the associated details window to where you see "x-csrf-token". Copy its value and store it into a variable named "csrf" which this function will reference when you execute it.');
}
} else {
console.error('You do not appear to be logged in. Please log in and try again.');
}
} else {
console.error('UID not passed. Pass the UID of the profile you are targeting to this function.');
}
}
booyahGetAccounts(41874362);
Detailed explanation of the process
As the function runs, it logs the progress to the console, both how many users have been followed so far, and how much progress has been made percentage-wise, based on the total number of followers the profile you are targeting has.
Retrieving your CSRF token
The only manual portion of this process is retrieving your CSRF token. This is rather simple though. Once you log into Booyah, navigate to the Network tab of your Chrome console and click on the item named bug-report-claims, then scroll all the way down the details window which appears on the right. There should see x-csrf-token. Store this value as a string variable in your console as csrf, which my function will reference when it runs. This is necessary in order to use the POST method to follow users.
Here is what it will look like:
The solution
The function will loop through all users the account you are targeting follows in batches of 100 (the max amount allowed per GET request) and follow them all. When the end of each batch is met, the next batch is automatically triggered recursively.
🚀 Version 3 (Fastest and most flexible, using async/await and fetch())
My previous two solution versions (🐇 …🐢) can be referenced in this answer's edit history.
Make sure to replace PUT_YOUR_CSRF_TOKEN_HERE with your own CSRF token value. Detailed instructions on how to find your CSRF token are below.
You must run this in your console on the Booyah website in order to avoid CORS issues.
const csrf = 'PUT_YOUR_CSRF_TOKEN_HERE';
async function booyahGetAccounts(uid, type = 'followers', follow = 1) {
if (typeof uid !== 'undefined' && !isNaN(uid)) {
const loggedInUserID = window.localStorage?.loggedUID;
if (uid === 0) uid = loggedInUserID;
const unfollow = follow === -1;
if (unfollow) follow = 1;
if (loggedInUserID) {
if (csrf) {
async function getUserData(uid) {
const response = await fetch(`https://booyah.live/api/v3/users/${uid}`),
data = await response.json();
return data.user;
}
const loggedInUserData = await getUserData(loggedInUserID),
targetUserData = await getUserData(uid),
followUser = uid => fetch(`https://booyah.live/api/v3/users/${loggedInUserID}/followings`, { method: (unfollow ? 'DELETE' : 'POST'), headers: { 'X-CSRF-Token': csrf }, body: JSON.stringify({ followee_uid: uid, source: 43 }) }),
logSep = (data = '', usePad = 0) => typeof data === 'string' && usePad ? console.log((data ? data + ' ' : '').padEnd(50, '━')) : console.log('━'.repeat(50),data,'━'.repeat(50));
async function getList(uid, type, follow) {
const isLoggedInUser = uid === loggedInUserID;
if (isLoggedInUser && follow && !unfollow && type === 'followings') {
follow = 0;
console.warn('You alredy follow your followings. `follow` mode switched to `false`. Followings will be retrieved instead of followed.');
}
const userData = await getUserData(uid),
totalCount = userData[type.slice(0,-1)+'_count'] || 0,
totalCountStrLength = totalCount.toString().length;
if (totalCount) {
let userIDsLength = 0;
const userIDs = [],
nickname = userData.nickname,
nicknameStr = `${nickname ? ` of ${nickname}'s ${type}` : ''}`,
alreadyFollowedStr = uid => `User ID ${uid} already followed by ${loggedInUserData.nickname} (Account #${loggedInUserID})`;
async function followerFetch(cursor = 0) {
const fetched = [];
await fetch(`https://booyah.live/api/v3/users/${uid}/${type}?cursor=${cursor}&count=100`).then(res => res.json()).then(data => {
const list = data[type.slice(0,-1)+'_list'];
if (list?.length) fetched.push(...list.map(e => e.uid));
if (fetched.length) {
userIDs.push(...fetched);
userIDsLength += fetched.length;
if (follow) followUser(uid);
console.log(`${userIDsLength.toString().padStart(totalCountStrLength)} (${(userIDsLength / totalCount * 100).toFixed(4)}%)${nicknameStr} ${follow ? 'followed' : 'retrieved'}`);
if (fetched.length === 100) {
followerFetch(data.cursor);
} else {
console.log(`END REACHED. ${userIDsLength} accounts ${follow ? 'followed' : 'retrieved'}.`);
if (!follow) logSep(targetList);
}
}
});
}
await followerFetch();
return userIDs;
} else {
console.log(`This account has no ${type}.`);
}
}
logSep(`${follow ? 'Following' : 'Retrieving'} ${targetUserData.nickname}'s ${type}`, 1);
const targetList = await getList(uid, type, follow);
} else {
console.error('Missing CSRF token. Retrieve your CSRF token from the Network tab in your inspector by clicking into the Network tab item named "bug-report-claims" and then scrolling down in the associated details window to where you see "x-csrf-token". Copy its value and store it into a variable named "csrf" which this function will reference when you execute it.');
}
} else {
console.error('You do not appear to be logged in. Please log in and try again.');
}
} else {
console.error('UID not passed. Pass the UID of the profile you are targeting to this function.');
}
}
Usage
To run the function (for either of the above solutions), just call the function name with the desired User ID name as an argument, in your example case, 41874362. The function call would look like this:
booyahGetAccounts(41874362);
The function is quite flexible in its abilities though. booyahGetAccounts() accepts three parameters, but only the first is required.
booyahGetAccounts(
uid, // required, no default
type = 'followers', // optional, must be 'followers' or 'followings' -> default: 'followers'
follow = 1 // optional, must be 0, 1, or -1, -> default: 1 (boolean true)
)
The second parameter, type, allows you to choose whether you would like to process the targeted user's followers or followings (the users which that user follows).
The third parameter allows you to choose whether you would like to follow/unfollow the returned users or only retrieve their User IDs. This defaults to 1 (boolean true) which will follow the users returned, but if you only want to test the function and not actually follow the returned users, set this to a falsy value such as 0 or false. Using -1 will unfollow the users returned.
This function intelligently retrieves your own User ID for you from the window.localStorage object, so you don't need to retrieve that yourself. If you would like to process your own followers or followings, simply pass 0 as the main uid parameter value, and the function will default the uid to your own User ID.
Because you can't re-follow users you already follow, if you try to follow your followings, the function will produce the warning You already follow your followings. 'follow' mode switched to 'false'. Followings will be retrieved instead of followed. and instead return them as if you had set the follow parameter to false.
However, it can be very useful to process your own list. For example, if you want to follow all of your own followers back, you could do so like this:
booyahGetAccounts(0); // `type` and `follow` parameters already default to the correct values here
On the other hand, if you were strategically using a follow/unfollow technique in order to increase your number of followers and needed to unfollow all of your followers, you could do so like this:
booyahGetAccounts(0, 'followers', -1);
By setting the follow parameter value to -1, you instruct the function to run its followUser function on all returned User IDs using the DELETE method instead of the POST method, thereby unfollowing those users returned instead of following them.
Desired outcome
Function call
Follow all your own followers
booyahGetAccounts(0, 'followers');
Unfollow all your own followers
booyahGetAccounts(0, 'followers', -1);
Unfollow all your own followings
booyahGetAccounts(0, 'followings', -1);
Follow users that follow User ID #12345
booyahGetAccounts(12345, 'followers');
Follow users followed by User ID #12345
booyahGetAccounts(12345, 'followings');
Retrieve User IDs of accounts following User ID #12345
booyahGetAccounts(12345, 'followers', 0);
Retrieve User IDs of accounts followed by User ID #12345
booyahGetAccounts(12345, 'followings', 0);
Other notes
To improve the performance of this function, as it's very heavy, I've replaced all calls to userIDs.length with a dedicated userIDsLength variable which I add to using += with each iteration rather than calling length each time. Similarly, I store the length of the stringified followerCount in the variable followerCountStrLength rather than calling followerCount.toString().length with each iteration. Because this is a rather heavy function, it is possible for your browser window to crash. However, it should eventually complete.
If the page appears to crash by flickering and auto-closing the console, FIRST try to re-open the console without refreshing the page at all. In my case, the inspector occasionally closed on its own, likely due to the exhaustion from the function, but when I opened the inspector's console again, the function was still running.
How can I set the badge number on a specific tab only? So far I have a code that sets the badge number on all the tabs.. I've been reading around A LOT, but there doesn't seem to be a whole lot of information about this, so perhaps I will find a solution for this here.
I would like something like Adblock Plus, which sets the badge number for a specific tab. This is pretty easy in Chrome etc, but doesn't seem to be the case in Safari.
Does anyone know how extensions like Adblock plus shows the badge number on a specific tab?
So far I only have this code, but as mentioned, it sets the badge on all the tabs, which is not the result I want.
safari.extension.toolbarItems[0].badge = 2;
Edit:
I have been looking at the source code of Adblock plus, and a few other extensions that had this function. And it seems it is using some prototype.
Adblock plus background snippet:
BrowserAction.prototype = {
_set: function(name, value)
{
var toolbarItem = getToolbarItemForWindow(this._page._tab.browserWindow);
if (!toolbarItem)
{
return;
}
var property = toolbarItemProperties[name];
if (!property)
{
property = toolbarItemProperties[name] = {
pages: new ext.PageMap(),
global: toolbarItem[name]
};
}
property.pages.set(this._page, value);
if (isPageActive(this._page))
{
toolbarItem[name] = value;
}
},
setIcon: function(path)
{
this._set("image", safari.extension.baseURI + path.replace("$size", "16"));
},
setBadge: function(badge)
{
if (!badge)
{
this._set("badge", 0);
}
else if ("number" in badge)
{
this._set("badge", badge.number);
}
}
};
Content script (adblockplus.js)
FilterNotifier.on("filter.hitCount", function(filter, newValue, oldValue, page)
{
if (!(filter instanceof BlockingFilter) || !page)
{
return;
}
Prefs.blocked_total++;
var blocked = blockedPerPage.get(page) || 0;
blockedPerPage.set(page, ++blocked);
if (Prefs.show_statsinicon)
{
page.browserAction.setBadge(
{
color: badgeColor,
number: blocked
});
}
});
It seems this is how Adblock plus does it, but so far I haven't been able to replicate it. Still trying though..
Okay, so I finally found a solution for this, and thought I would share what I did, in case somebody else is in the same situation.
This morning I got the idea of storing the data in an array, when the user visits one of the websites I want to display the badge number on (doesn't store all websites the user visits), only if it matched one of the websites I wanted to target. I stored the following data in the array: root domain (example.com) and the badgeNumber.
For this to work, you need to make an array of the root domain of the websites you want to target, and then only execute the following when it matches, otherwise the array would fill up very quickly, and we don't want too much data in it.
In the global page, start by making an empty array to store the data
var badgeUpdateArray = [];
You then need to set up message handling in your global page as well.
safari.application.addEventListener('message', handleMessage, false);
function handleMessage(event) {
if(event.name === "setBadgeText"){
var id = badgeUpdateArray.length + 1;
var isFound = 0;
var found = badgeUpdateArray.some(function (el) {
if(el.identifier === event.message.identifier){
// Was found
isFound = 1;
}
});
if (isFound == 0) {
// Not found, add to the array
badgeUpdateArray.push({identifier:event.message.identifier,badgeNumber:event.message.badgeNumber});
}
// Set the badge number
safari.extension.toolbarItems[0].badge = event.message.badgeNumber;
}
}
Now we need to send the message from the content script to the global page. You need to get the root domain (example.com), I'm not getting into that here, as it's pretty easy. You will also need the badgeNumber value, this can be gathered from wherever (GET request, or elsewhere..)
Remember, only execute this code if the website matches your target domains.
var message = {
identifier: domain,
badgeNumber: rows.length
}
safari.self.tab.dispatchMessage("setBadgeText", message);
This will send the message, and store the data in the array, it will also set the badge number.
Now, for this to be working on different tabs, you will need to make an event handler for "activate" on the global page, this will run whenever a tab is active.
safari.application.addEventListener("activate", updateBadge, true);
function updateBadge(){
var cDomain = safari.application.activeBrowserWindow.activeTab.url;
cDomain = cDomain.replace("www3.","");
cDomain = cDomain.replace("www2.","");
cDomain = cDomain.replace("www1.","");
cDomain = cDomain.replace("www.","");
cDomain = new URL(cDomain);
cDomain = cDomain.hostname;
var id = badgeUpdateArray.length + 1;
var isFound = 0;
var badgeNumber = 0;
var found = badgeUpdateArray.some(function (el) {
badgeNumber = el.badgeNumber;
if(el.identifier === cDomain){
// Was found, set the badge number
isFound = 1;
safari.extension.toolbarItems[0].badge = el.badgeNumber;
}
});
if (isFound == 0) {
// Was not found
safari.extension.toolbarItems[0].badge = 0;
}
}
Hopefully I've got it all in here, and at least something that works, though I have to say that I would prefer an easier way of storing it.. like Chrome etc does it, with the tab API.
I'm trying to write a simple PhantomJS script where I find an element by ID and determine if it is empty of not. I've tried a few suggested things such as .childNodes.length, .textContent, etc.
These either result in a null error:
TypeError: null is not an object (evaluating 'document.getElementById('idname').childNodes')
Or phantom just crashes and refuses to check the links at all, usually this happens if I run my script twice in a row without much pause. And it will sometimes sit and do nothing.
I've written other scrapers that effectively used getElementById in this way, and they were successful, although there I was just checking if the element existed by checking if it was !== null. Checking manually, this element does exist in all the pages I'm checking, it's just that it sometimes has content and sometimes doesn't (it's a div). Anyway, here is my code:
var fs = require('fs')
var urls = fs.read('urls.txt').split('\n');
var page;
page = require('webpage').create();
console.log('The default user agent is ' + page.settings.userAgent);
page.settings.userAgent = 'SpecialAgent';
function check_link(url){
page = require('webpage').create();
page.open(url, function(status){
if (status !== 'success') {
console.log('Unable to access network');
} else {
var error = page.evaluate(function() {
return document.getElementById('error-message');
});
console.log(error.childNodes.length);
fs.write('results.csv', error.childNodes.length + ', ' + url + '\n', 'a');
page.release();
setTimeout(next_link, 1000);
}
});
}
function next_link(){
var url = urls.shift();
console.log(url);
if(!urls){
phantom.exit(0);
} else{
check_link(url);
}
}
next_link();
PhantomJS provides access to the sandboxed page context (DOM context) through page.evaluate() with the following note:
Note: The arguments and the return value to the evaluate function must be a simple primitive object. The rule of thumb: if it can be serialized via JSON, then it is fine.
Closures, functions, DOM nodes, etc. will not work!
So you cannot pass the DOM node out of the page context, but you can do everything you want with it in the page context and then pass out the result.
var errors = page.evaluate(function() {
var e = document.getElementById('error-message');
return (e && e.childNodes) ? e.childNodes.length : -1
});
console.log(errors);
I am using the spotify-web-api-js node module, JavaScript, and jQuery to log into a user's Spotify account and save an array of tracks to a newly created playlist, but so far I can only get a blank playlist created.
I gather all of the tracks through the API and put them in a globally defined array as I show the list to the user. This array is defined on line 39: var song_uris = [];
console.log shows that this array is filled with the URIs I need after this function, which occurs upon initial search:
function searchArtists(originalArtist, callback) {
$(window).on('hashchange', function() {
console.log(window.location.hash);
});
console.log('originalArtist', originalArtist);
$.getJSON("https://api.spotify.com/v1/search?type=artist&q=" + originalArtist, function(json) {
$('#artist').html('<p>'+ '<img src="' + json.artists.items[0].images[2].url + '" height="100" width="100" /> ' + json.artists.items[0].name +'</p>');
var originalArtistId = json.artists.items[0].id;
s.getArtistRelatedArtists(originalArtistId, function(err, data) {
relatedArtists = {};
for (var i = 0; i < data.artists.length; i++) {
relatedArtists[data.artists[i].id] = {};
relatedArtists[data.artists[i].id].name = data.artists[i].name;
relatedArtists[data.artists[i].id].id = data.artists[i].id;
}
var counter = 0;
for (var id in relatedArtists) {
relatedArtists[counter] = relatedArtists[id];
delete relatedArtists[id];
counter++;
}
async.times(counter, function(n, next) {
console.log(n);
console.log(relatedArtists[n].id);
s.getArtistTopTracks(relatedArtists[n].id, "US", function (err, data2) {
relatedArtists[n].song = data2.tracks[0].name; //sometimes this is a TypeError? idk
relatedArtists[n].uri = data2.tracks[0].uri;
$('#related-artist').append('<p><strong>' + relatedArtists[n].name + '</strong> -- \"' + relatedArtists[n].song + '\"</p>');
song_uris.push(relatedArtists[n].uri);
next(null, relatedArtists[n].uri);
});
}, function(err, song_uris) {
console.log(song_uris); //array is full here
});
});
});
}
However, as soon as people go through the optional login/callback process, I get all of the user's data (including the new playlist URI), but the song_uris array I had before is now empty.
if (params.access_token) {
s.setAccessToken(params.access_token);
s.getMe().then(function(data) {
// and here it goes the user's data!!!
console.log(data);
console.log(data.id);
user_id = data.id;
s.createPlaylist(user_id, {name: 'Related Artist Playlist'}).then(function(data3) {
console.log(data3);
playlist_id = data3.uri;
playlist_id = playlist_id.substring(33);
console.log(playlist_id);
console.log(user_id);
console.log(song_uris); //array is empty here
s.addTracksToPlaylist(user_id, playlist_id, song_uris).then(function(data){
console.log(data);
});
});
});
}
Since this is one of the parameters of the addTracksToPlaylist function, the XML request doesn't work.
Here is a screenshot of my console, where I see a failed POST request and an uncaught exception: object XMLHttpRequest:
There is obviously an issue with the scope and/or async, but I can't seem to figure out where to put addTracksToPlaylist instead. Putting it immediately after the line where console.log outputs a full array either doesn't do anything (if it's within a window.onhashchange = function ()), or gives an error because the user and playlist are undefined before the callback (if the addTracksToPlaylist function is there alone).
Is the data being overwritten when the site refreshes post-callback? If so, how can I stop that so I have a full array of URIs that fit in the URL string? (According to the Spotify Developer docs, the URL should have the comma-separated list of track URIs (which was in my array) passed to it through the uri parameter, but because of the empty array it breaks and I get https://api.spotify.com/v1/users/account/playlists/5kxeeKym1tpEx8Trj3qkd5/tracks?uris= (with the empty parameter). How can I solve this? Or is there an alternate way to keep the desired user flow of Search for Artist -> See List of Songs from Related Artists -> (Optional) Login To Spotify Account -> (Optional) Save List of Songs from Related Artists to New Playlist?
I've created a jsfiddle with the full JS and HTML code as well: https://jsfiddle.net/37Lkrcb1/1/
Notes: This is running on a Node.js server and I use bundle and browserify to compile the script.
It sounds like at least one path in your app will involve a page refresh you can't avoid (going to Spotify and coming back). In that case, all your variables will be reset by the page load.
If you have song_uris you want to persist across that reload, you can store them in sessionStorage (or localStorage) prior to the Spotify authentication, and retrieve them when the page loads.
Storing it:
sessionStorage.setItem("song_uris", JSON.stringify(song_uris));
// Or localStorage.setItem...
Retrieving it on page load:
var song_uris = JSON.parse(sessionStorage.getItem("song_uris") || "null");
if (!song_uris) {
// There weren't any in storage, populate in another way or set dfeault
}
Web storage is nearly universally supported.