We Keep Coding

How can I return different values from a function depending on code inside an Axios promise? NodeJS - a

I have a block of code that calls an Api and saves results if there are differences or not. I would like to return different values for DATA as layed out on the code. But this is obviously not working since Its returning undefined.
let compare = (term) => {
let DATA;
//declare empty array where we will push every thinkpad computer for sale.
let arrayToStore = [];
//declare page variable, that will be the amount of pages based on the primary results
let pages;
//this is the Initial get request to calculate amount of iterations depending on result quantities.
axios.get('https://api.mercadolibre.com/sites/MLA/search?q='+ term +'&condition=used&category=MLA1652&offset=' + 0)
.then(function (response) {
//begin calculation of pages
let amount = response.data.paging.primary_results;
//since we only care about the primary results, this is fine. Since there are 50 items per page, we divide
//amount by 50, and round it up, since the last page can contain less than 50 items
pages = Math.ceil(amount / 50);
//here we begin the for loop.
for(i = 0; i < pages; i++) {
// So for each page we will do an axios request in order to get results
//Since each page is 50 as offset, then i should be multiplied by 50.
axios.get('https://api.mercadolibre.com/sites/MLA/search?q='+ term +'&condition=used&category=MLA1652&offset=' + i * 50)
.then((response) => {
const cleanUp = response.data.results.map((result) => {
let image = result.thumbnail.replace("I.jpg", "O.jpg");
return importante = {
id: result.id,
title: result.title,
price: result.price,
link: result.permalink,
image: image,
state: result.address.state_name,
city: result.address.city_name
}
});
arrayToStore.push(cleanUp);
console.log(pages, i)
if (i === pages) {
let path = ('./compare/yesterday-' + term +'.json');
if (fs.existsSync(path)) {
console.log("Loop Finished. Reading data from Yesterday")
fs.readFile('./compare/yesterday-' + term +'.json', (err, data) => {
if (err) throw err;
let rawDataFromYesterday = JSON.parse(data);
// test
//first convert both items to check to JSON strings in order to check them.
if(JSON.stringify(rawDataFromYesterday) !== JSON.stringify(arrayToStore)) {
//Then Check difference using id, otherwise it did not work. Using lodash to help.
let difference = _.differenceBy(arrayToStore[0], rawDataFromYesterday[0],'id');
fs.writeFileSync('./compare/New'+ term + '.json', JSON.stringify(difference));
//if they are different save the new file.
//Then send it via mail
console.log("different entries, wrote difference to JSON");
let newMail = mail(difference, term);
fs.writeFileSync('./compare/yesterday-' + term +'.json', JSON.stringify(arrayToStore));
DATA = {
content: difference,
message: "These were the differences, items could be new or deleted.",
info: "an email was sent, details are the following:"
}
return DATA;
} else {
console.log("no new entries, cleaning up JSON");
fs.writeFileSync('./compare/New'+ term + '.json', []);
DATA = {
content: null,
message: "There were no difference from last consultation",
info: "The file" + './compare/New'+ term + '.json' + ' was cleaned'
}
return DATA;
}
});
} else {
console.error("error");
console.log("file did not exist, writing new file");
fs.writeFileSync('./compare/yesterday-' + term +'.json', JSON.stringify(arrayToStore));
DATA = {
content: arrayToStore,
message: "There were no registries of the consultation",
info: "Writing new file to ' " + path + "'"
}
return DATA;
}
}
})
}
}).catch(err => console.log(err));
}
module.exports = compare
So I export this compare function, which I call on my app.js.
What I want is to make this compare function return the DATA object, so I can display the actual messages on the front end,
My hopes would be, putting this compare(term) function inside a route in app.js like so:
app.get("/api/compare/:term", (req, res) => {
let {term} = req.params
let data = compare(term);
res.send(data);
})
But as I said, Its returning undefined. I tried with async await, or returning the whole axios first axios call, but Im always returning undefined.
Thank you

Javascript Promise().then to prevent re-calling the function before the first call be executed

In my node.js app, reading data from MSSQL using tedious, I'm calling the below every 1 second:
Fetch the data from the server (fetchStock function) and save it in temporary array
Send the data saved in the temporary array to the client using the Server-Sent Events (SSE) API.
It looks the 1 second is not enough to recall the fetchStock function before the previous call is completely executed, so I get execution errors from time to time.
I increased it to 5 seconds, but still get the same issue every once in a while.
How can I use Promise().then to be sure the fetchStock function is not re-called before the previouse call be completely executed?
var Request = require('tedious').Request;
var Connection = require('tedious').Connection;
var config = {
userName: 'sa',
password: 'pswd',
server: 'xx.xxx.xx.xxx',
options: {
database: 'DB',
rowCollectionOnRequestCompletion: 'true',
rowCollectionOnDone: 'true'
},
};
var sql = new Connection(config);
var addElem = (obj, elem)=> [].push.call(obj, elem);
var result = {}, tmpCol = {}, tmpRow = {};
module.exports = {
displayStock: function (es) {
var dloop = setInterval(function() {
if(result.error !== null)
if (es) es.send(JSON.stringify(result), {event: 'rmSoH', id: (new Date()).toLocaleTimeString()});
if(result.error === null)
if (es) es.send('connection is closed');
}, 1000);
},
fetchStock: function () {
request = new Request("SELECT ItemCode, WhsCode, OnHand FROM OITW where OnHand > 0 and (WhsCode ='RM' or WhsCode ='FG');", function(err, rowCount, rows) {
if (err) {
result = {'error': err};
console.log((new Date()).toLocaleTimeString()+' err : '+err);
}
if(rows)
rows.forEach(function(row){
row.forEach(function(column){
var colName = column.metadata.colName;
var value = column.value;
addElem(tmpCol, {colName: value})
});
addElem(tmpRow,{'item': tmpCol[0].colName, 'Whs': tmpCol[1].colName, 'Qty': tmpCol[2].colName});
tmpCol = {};
});
result = tmpRow;
tmpRow={}
});
sql.execSql(request);
}
}

I think what you need is a simple variable to check if there's already running request not Promise.
var latch = false;
// It will be called only if the previous call is completed
var doFetchStock = () => sql.execSql(new Request("SQL", (err, rowCount, rows) => {
// Your logic dealing with result
// Initializes the latch
latch = false;
});
module.exports = {
fetchStock: function () {
// Check if the previous request is completed or not
if (!latch) {
// Sets the latch
latch = true;
// Fetches stock
doFetchStock();
}
}
};
Actually I've used this kind of pattern a lot to allow some behavior only once.
https://github.com/cettia/cettia-javascript-client/blob/1.0.0-Beta1/cettia.js#L397-L413
https://github.com/cettia/cettia-javascript-client/blob/1.0.0-Beta1/cettia.js#L775-L797

Since javascript is mono-threaded a simple code like this should be enough on client-side
function () {
if(currentPromise != null){ // define in a closure outside
currentPromise = [..] // call to server which return a promise
currentPromise.then(function(){
currentPromise = null;
});
}
}

How can I throttle stack of api requests?

I have an array of ids, and I want to make an api request for each id, but I want to control how many requests are made per second, or better still, have only 5 open connections at any time, and when a connection is complete, fetch the next one.
Currently I have this, which just fires off all the requests at the same time:
_.each([1,2,3,4,5,6,7,8,9,10], function(issueId) {
github.fetchIssue(repo.namespace, repo.id, issueId, filters)
.then(function(response) {
console.log('Writing: ' + issueId);
writeIssueToDisk(fetchIssueCallback(response));
});
});

Personally, I'd use Bluebird's .map() with the concurrency option since I'm already using promises and Bluebird for anything async. But, if you want to see what a hand-coded counter scheme that restricts how many concurrent requests can run at once looks like, here's one:
function limitEach(collection, max, fn, done) {
var cntr = 0, index = 0, errFlag = false;
function runMore() {
while (!errFlag && cntr < max && index < collection.length) {
++cntr;
fn(collection[index++], function(err, data) {
--cntr;
if (errFlag) return;
if (err) {
errFlag = true;
done(err);
} else {
runMore();
}
});
}
if (!errFlag && cntr === 0 && index === collection.length) {
done();
}
}
runMore();
}

With Bluebird:
function fetch(id) {
console.log("Fetching " + id);
return Promise.delay(2000, id)
.then(function(id) {
console.log(" Fetched " + id);
});
}
var ids = [1,2,3,4,5,6,7,8,9,10];
Promise.map(ids, fetch, { concurrency: 3 });
<script src="https://cdnjs.cloudflare.com/ajax/libs/bluebird/3.3.1/bluebird.min.js"></script>
<!-- results pane console output; see http://meta.stackexchange.com/a/242491 -->
<script src="http://gh-canon.github.io/stack-snippet-console/console.min.js"></script>

Divide your data into as many arrays as you want concurrent connections. Schedule with setTimeout, and have the completion callback handle the rest of the sub-array.
Wrap the setTimeout in a function of its own so that the variable values are frozen to their values at the time of delayed_fetch() invocation.
function delayed_fetch(delay, namespace, id, issueIds, filters) {
setTimeout(
function() {
var issueId=issueIds.shift();
github.fetchIssue(namespace, id, issueId, filters).then(function(response) {
console.log('Writing: ' + issueId);
writeIssueToDisk(fetchIssueCallback(response));
delayed_fetch(0, namespace, id, issueIds, filters);
});
}, delay);
}
var i=0;
_.each([ [1,2] , [3,4], [5,6], [7,8], [9,10] ], function(issueIds) {
var delay=++i*200; // millisecond
delayed_fetch(delay, repo.namespace, repo.id, issueIds, filters);
});

i'd recommend using throat just for this: https://github.com/ForbesLindesay/throat

Using Bluebird
function getUserFunc(user) {
//Get a collection of user
}
function getImageFunc(id) {
//get a collection of image profile based on id of the user
}
function search(response) {
return getUsersFunc(response).then(response => {
const promises = response.map(items => return items.id);
const images = id => {
return getImagesFunc(id).then(items => items.image);
};
return Promise.map(promises, images, { concurrency: 5 });
});
}
Previously i used ES6 function Promise.all(), but it doesn't work like what i'm expecting. Then go with third party library bluebird.js and Work like a charm.

node.js compare two arrays with objects

I need to remove all documents from my mongo db, which dont exists in new array with objects.
So I have array with objects like :
var items = [
{product_id:15, pr_name: 'a', description : 'desc'},
{product_id:44, pr_name: 'b', description : 'desc2'}
{product_id:32, pr_name: 'c', description : 'desc3'}];
and I have array with db values which I get by calling Model.find({}).
So now I do it in a 'straight' way:
async.each(products, function (dbProduct, callback) { //cycle for products removing
var equals = false;
async.each(items, function(product, callback){
if (dbProduct.product_id === product.product_id){
product.description = dbProduct.description;// I need to save desc from db product to new product
equals = true;
}
callback();
});
if (!equals) {
log.warn("REMOVE PRODUCT " + dbProduct.product_id);
Product.remove({ _id: dbProduct._id }, function (err) {
if (err) return updateDBCallback(err);
callback();
});
}
});
But its blocks the whole app and its very slow, because I have around 5000 values in my items array and in database too. So its very huge cycle numbers.
Maybe there can be a faster way?
UPDATE1
Using code below, from TbWill4321 answer:
var removeIds = [];
// cycle for products removing
async.each(products, function (dbProduct, callback) {
for ( var i = 0; i < items.length; i++ ) {
if (dbProduct.product_id === product.product_id) {
// I need to save desc from db product to new product
product.description = dbProduct.description;
// Return early for performance
return callback();
}
}
// Mark product to remove.
removeIds.push( dbProduct._id );
log.warn("REMOVE PRODUCT " + dbProduct.product_id);
return callback();
}, function() {
Product.remove({ _id: { $in: removeIds } }, function (err) {
if (err) return updateDBCallback(err);
// Continue Here.
// TODO
});
});
Its takes around 11 sec(blocks whole web-app) and takes 12 362 878 cycles for me.
So maybe somebody can advise me something?

The Async library does not execute synchronous code in an asynchronous fashion.
5000 items is not a huge number for JavaScript, as I've worked on Big Data set's with 5 million+ points and it doesn't take long. You can get better performance by structuring like this:
var removeIds = [];
// cycle for products removing
async.each(products, function (dbProduct, callback) {
for ( var i = 0; i < items.length; i++ ) {
if (dbProduct.product_id === product.product_id) {
// I need to save desc from db product to new product
product.description = dbProduct.description;
// Return early for performance
return callback();
}
}
// Mark product to remove.
removeIds.push( dbProduct._id );
log.warn("REMOVE PRODUCT " + dbProduct.product_id);
return callback();
}, function() {
Product.remove({ _id: { $in: removeIds } }, function (err) {
if (err) return updateDBCallback(err);
// Continue Here.
// TODO
});
});

Among the many problems you may have, off the top of my head you may want to start off by changing this bit:
Product.remove({ _id: dbProduct._id }, function (err) {
if (err) return updateDBCallback(err);
callback();
});
Being within a .each() call, you'll make one call to the database for each element you want to delete. It's better to store all the ids in one array and then make a single query to delete all elements that have an _id that is in that array. Like this
Product.remove({ _id: {$in: myArrayWithIds} }, function (err) {
if (err) return updateDBCallback(err);
callback();
});
On another note, since async will execute synchronously, node.js does offer setImmediate() (docs here), that will execute the function from within the event loop. So basically you can "pause" execution of new elements and serve any incoming requests to simulate "non-blocking" processing.

Bulk upsert in MongoDB using mongoose

Is there any option to perform bulk upserts with mongoose? So basically having an array and insert each element if it not exists or update it if it exists? (I am using customs _ids)
When I do use .insert MongoDB returns an error E11000 for duplicate keys (which should be updated). Inserting multiple new document works fine though:
var Users = self.db.collection('Users');
Users.insert(data, function(err){
if (err) {
callback(err);
}
else {
callback(null);
}
});
Using .save returns an error that the parameter must be a single document:
Users.save(data, function(err){
...
}
This answer suggest there is no such option, however it is specific for C# and also already 3 years old. So I was wondering if there is any option to do that using mongoose?
Thank you!

Not in "mongoose" specifically, or at least not yet as of writing. The MongoDB shell as of the 2.6 release actually uses the "Bulk operations API" "under the hood" as it were for all of the general helper methods. In it's implementation, it tries to do this first, and if an older version server is detected then there is a "fallback" to the legacy implementation.
All of the mongoose methods "currently" use the "legacy" implementation or the write concern response and the basic legacy methods. But there is a .collection accessor from any given mongoose model that essentially accesses the "collection object" from the underlying "node native driver" on which mongoose is implemented itself:
var mongoose = require('mongoose'),
Schema = mongoose.Schema;
mongoose.connect('mongodb://localhost/test');
var sampleSchema = new Schema({},{ "strict": false });
var Sample = mongoose.model( "Sample", sampleSchema, "sample" );
mongoose.connection.on("open", function(err,conn) {
var bulk = Sample.collection.initializeOrderedBulkOp();
var counter = 0;
// representing a long loop
for ( var x = 0; x < 100000; x++ ) {
bulk.find(/* some search */).upsert().updateOne(
/* update conditions */
});
counter++;
if ( counter % 1000 == 0 )
bulk.execute(function(err,result) {
bulk = Sample.collection.initializeOrderedBulkOp();
});
}
if ( counter % 1000 != 0 )
bulk.execute(function(err,result) {
// maybe do something with result
});
});
The main catch there being that "mongoose methods" are actually aware that a connection may not actually be made yet and "queue" until this is complete. The native driver you are "digging into" does not make this distinction.
So you really have to be aware that the connection is established in some way or form. But you can use the native driver methods as long as you are careful with what you are doing.

You don't need to manage limit (1000) as #neil-lunn suggested. Mongoose does this already. I used his great answer as a basis for this complete Promise-based implementation & example:
var Promise = require('bluebird');
var mongoose = require('mongoose');
var Show = mongoose.model('Show', {
"id": Number,
"title": String,
"provider": {'type':String, 'default':'eztv'}
});
/**
* Atomic connect Promise - not sure if I need this, might be in mongoose already..
* #return {Priomise}
*/
function connect(uri, options){
return new Promise(function(resolve, reject){
mongoose.connect(uri, options, function(err){
if (err) return reject(err);
resolve(mongoose.connection);
});
});
}
/**
* Bulk-upsert an array of records
* #param {Array} records List of records to update
* #param {Model} Model Mongoose model to update
* #param {Object} match Database field to match
* #return {Promise} always resolves a BulkWriteResult
*/
function save(records, Model, match){
match = match || 'id';
return new Promise(function(resolve, reject){
var bulk = Model.collection.initializeUnorderedBulkOp();
records.forEach(function(record){
var query = {};
query[match] = record[match];
bulk.find(query).upsert().updateOne( record );
});
bulk.execute(function(err, bulkres){
if (err) return reject(err);
resolve(bulkres);
});
});
}
/**
* Map function for EZTV-to-Show
* #param {Object} show EZTV show
* #return {Object} Mongoose Show object
*/
function mapEZ(show){
return {
title: show.title,
id: Number(show.id),
provider: 'eztv'
};
}
// if you are not using EZTV, put shows in here
var shows = []; // giant array of {id: X, title: "X"}
// var eztv = require('eztv');
// eztv.getShows({}, function(err, shows){
// if(err) return console.log('EZ Error:', err);
// var shows = shows.map(mapEZ);
console.log('found', shows.length, 'shows.');
connect('mongodb://localhost/tv', {}).then(function(db){
save(shows, Show).then(function(bulkRes){
console.log('Bulk complete.', bulkRes);
db.close();
}, function(err){
console.log('Bulk Error:', err);
db.close();
});
}, function(err){
console.log('DB Error:', err);
});
// });
This has the bonus of closing the connection when it's done, displaying any errors if you care, but ignoring them if not (error callbacks in Promises are optional.) It's also very fast. Just leaving this here to share my findings. You can uncomment the eztv stuff if you want to save all eztv shows to a database, as an example.

await Model.bulkWrite(docs.map(doc => ({
updateOne: {
filter: {id: doc.id},
update: doc,
upsert: true
}
})))
Or more verbose:
const bulkOps = docs.map(doc => ({
updateOne: {
filter: {id: doc.id},
update: doc,
upsert: true
}
}))
Model.bulkWrite(bulkOps)
.then(bulkWriteOpResult => console.log('BULK update OK:', bulkWriteOpResult))
.catch(err => console.error('BULK update error:', err))
https://stackoverflow.com/a/60330161/5318303

I have released a plugin for Mongoose that exposes a static upsertMany method to perform bulk upsert operations with a promise interface.
An added benefit of using this plugin over initializing your own bulk op on the underlying collection, is that this plugin converts your data to Mongoose model's first, and then back to plain objects before the upsert. This ensures Mongoose schema validation is applied, and data is depopulated and fit for raw insertion.
https://github.com/meanie/mongoose-upsert-many
https://www.npmjs.com/package/#meanie/mongoose-upsert-many
Hope it helps!

If you're not seeing the bulk methods in your db.collection ie you're getting a error to the effect of
xxx variable has no method: initializeOrderedBulkOp()
Try updating your mongoose version. Apparently older mongoose versions don't pass through all of the underlying mongo db.collection methods.
npm install mongoose
took care of it for me.

I had to achieve this recently while storing products in my ecommerce app. My database used to timeout as I had to upsert 10000 items every 4 hours. One option for me was to set the socketTimeoutMS and connectTimeoutMS in mongoose while connecting to the database but it sorta felt hacky and I did not want to manipulate connection timeout defaults of the database. I also see that the solution by #neil lunn takes a simple sync approach of taking a modulus inside the for loop. Here is an async version of mine that I believe does the job much better
let BATCH_SIZE = 500
Array.prototype.chunk = function (groupsize) {
var sets = [];
var chunks = this.length / groupsize;
for (var i = 0, j = 0; i < chunks; i++ , j += groupsize) {
sets[i] = this.slice(j, j + groupsize);
}
return sets;
}
function upsertDiscountedProducts(products) {
//Take the input array of products and divide it into chunks of BATCH_SIZE
let chunks = products.chunk(BATCH_SIZE), current = 0
console.log('Number of chunks ', chunks.length)
let bulk = models.Product.collection.initializeUnorderedBulkOp();
//Get the current time as timestamp
let timestamp = new Date(),
//Keep track of the number of items being looped
pendingCount = 0,
inserted = 0,
upserted = 0,
matched = 0,
modified = 0,
removed = 0,
//If atleast one upsert was performed
upsertHappened = false;
//Call the load function to get started
load()
function load() {
//If we have a chunk to process
if (current < chunks.length) {
console.log('Current value ', current)
for (let i = 0; i < chunks[current].length; i++) {
//For each item set the updated timestamp to the current time
let item = chunks[current][i]
//Set the updated timestamp on each item
item.updatedAt = timestamp;
bulk.find({ _id: item._id })
.upsert()
.updateOne({
"$set": item,
//If the item is being newly inserted, set a created timestamp on it
"$setOnInsert": {
"createdAt": timestamp
}
})
}
//Execute the bulk operation for the current chunk
bulk.execute((error, result) => {
if (error) {
console.error('Error while inserting products' + JSON.stringify(error))
next()
}
else {
//Atleast one upsert has happened
upsertHappened = true;
inserted += result.nInserted
upserted += result.nUpserted
matched += result.nMatched
modified += result.nModified
removed += result.nRemoved
//Move to the next chunk
next()
}
})
}
else {
console.log("Calling finish")
finish()
}
}
function next() {
current++;
//Reassign bulk to a new object and call load once again on the new object after incrementing chunk
bulk = models.Product.collection.initializeUnorderedBulkOp();
setTimeout(load, 0)
}
function finish() {
console.log('Inserted ', inserted + ' Upserted ', upserted, ' Matched ', matched, ' Modified ', modified, ' Removed ', removed)
//If atleast one chunk was inserted, remove all items with a 0% discount or not updated in the latest upsert
if (upsertHappened) {
console.log("Calling remove")
remove()
}
}
/**
* Remove all the items that were not updated in the recent upsert or those items with a discount of 0
*/
function remove() {
models.Product.remove(
{
"$or":
[{
"updatedAt": { "$lt": timestamp }
},
{
"discount": { "$eq": 0 }
}]
}, (error, obj) => {
if (error) {
console.log('Error while removing', JSON.stringify(error))
}
else {
if (obj.result.n === 0) {
console.log('Nothing was removed')
} else {
console.log('Removed ' + obj.result.n + ' documents')
}
}
}
)
}
}

You can use mongoose's Model.bulkWrite()
const res = await Character.bulkWrite([
{
updateOne: {
filter: { name: 'Will Riker' },
update: { age: 29 },
upsert: true
}
},
{
updateOne: {
filter: { name: 'Geordi La Forge' },
update: { age: 29 },
upsert: true
}
}
]);
reference : https://masteringjs.io/tutorials/mongoose/upsert