db.firmalar.find().forEach(function(obj) {
for( var i = 0; i < obj.osgbIdleri.length; i++){ //ObjectId ARRAY
obj.osgbIdleri[i] = ObjectId(obj.osgbIdleri[i]);
}
//out:result (firmaId.id & firmaId.osgbIdleri.id(ObjectId))
});
I want to save the string field "obj.osgbIdleri" in each document in the "Firmalar" collection as the ObjectId field. I think I can do this using "aggregation". But when using "aggregation", I can not return every object in foreach. I want to create "firmaId.id" and "firmaId.osgbIdleri.id (ObjectID)" while creating a new collection.
(Posted on behalf of the OP).
db.firmalar.mapReduce(
function() {
for( var i = 0; i < this.osgbIdleri.length; i++){
this.osgbIdleri[i] = ObjectId(this.osgbIdleri[i]);
}
emit(this._id, this.osgbIdleri); },
function(key, values) {return value}, {
query:{},
out:"firmalarId"
}
).find()
Related
I am new to angular/js/ionic. I have a really quick question.
How do I retrieve data from localstorage when I don't really know the index name? I just want to grab whatever ends with ".log"?
I understand that the code below will try to retrieve the data by the index 'notes'.
var logFile = angular.fromJson(window.localStorage['notes'] || '[]');
var localLogFile = function() {
for(var i = 0; i < localStorage.length; i++) {
var keyName = localStorage.key(i);
var extension = keyName.substring(keyName.lastIndexOf("."));
if(extension == ".log") {
return localStorage.getItem(keyName);
}
}
};
var logFile = angular.fromJson(localLogFile());
I have collection of data in Mongodb, i want to give best matches suggestion while user input query in our suggestion box,
when user start typing com suggestion should be:
Computer
Computer Science
something more alike
I am sorting in Node by getting all matched data from mongo first and then give a rank to each data
function rank(name, q) {
var len = name.length,
lastIndex = -1;
for(var i = 0; i < q.length; i++) {
var n = name.indexOf(q[i], (lastIndex + 1));
if(n !== -1) {
len--;
lastIndex = n;
}
}
return len;
}
var query = 'com';
// giving rank to data
data = data.map(function(v) {
v.rank = rank(v.value, query);
return v;
});
// sorting by rank
data = data.sort(function(a, b) {
return a.rank - b.rank
});
It is giving me satisfied result, but it will be too slow while dealing with large data.
I want let mongodb engine to deal with sorting and give me just limited best matches result.
Maybe you could do it through mapreduce. Map-reduce is a data processing paradigm for condensing large volumes of data into useful aggregated results.
var mapFn = function(){
var len = this.name.length,
lastIndex = -1;
var q = 'com';
for(var i = 0; i < q.length; i++) {
var n = this.name.indexOf(q[i], (lastIndex + 1));
if(n !== -1) {
len--;
lastIndex = n;
}
}
emit(len, this);
};
var reduceFn = function(key, values){
return values.sort(function(a,b){
return a.name - b.name;
});
};
db.collection.mapReduce(mapFn, reduceFn, { out: { reduce: 'result_collection'}});
If I want to query users' age > 18,
and export result to corresponding collection,
How could I do it by rewriteing the following script?
The following is psuedo code
source_collections = ["user_1", "user_2", ..., "user_999"]
output_collections = ["result_1", "result_2", ..., "result_999"]
pipeline = [
{
"$match":{"age": > 18}
}
{ "$out" : output_collections }
]
cur = db[source_collections].runCommand('aggregate',
{pipeline: pipeline,allowDiskUse: true})
The script you're looking for is something like:
var prefix_source = 'user_';
var prefix_output = 'result_';
var source_collections = [];
var output_collections = [];
var numCollections = 999;
for (var i = 1; i <= numCollections; i++) {
source_collections.push(prefix_source + i);
output_collections.push(prefix_output + i);
}
var pipeline = [{'$match': {age: {'$gt': 18}}}, {'$out': ''}]; for (var currentCollection = 0; currentCollection < source_collections.length; currentCollection++) {
pipeline[pipeline.length - 1]['$out'] = output_collections[currentCollection];
var cur = db[source_collections[currentCollection]].runCommand('aggregate', {pipeline: pipeline,allowDiskUse: true});
}
And while you're at it, the var cur = ... line could be simplified to
db[source_collections[currentCollection]].aggregate(pipeline, {allowDiskUse: true});
Note: I've added a piece that generates your arrays for you, as I'm sure you're not looking to write them by hand :D
I have mongodb in which there is 3 huge collections say 'A', 'B' and 'C'
Each collection contains about 2 million documents.
There are certain properties for each of the document.
Each document need to be updated based on those values of certain properties, from which i can determine what should be the '$set' to that document.
currently i am using the same approach for each collection.
that to find all documents in batches. collection them in memory (which i think the culprit for the current approach), then one by one update them all.
For the first collection(that have similar data as in other collections), it takes 10 minutes to get completed. then the next two collections taking 2 hours approx to get the task done or mongodb client get crashed earlier.
There is something wrong and no desired in the current approach.
Model.collection.find({}).batchSize(BATCH).toArray(function(err, docs){
if(err || !docs || !docs.length)
return afterCompleteOneCollection(err);
var spec = function(index) {
if(index % 1000 === 0) console.log('at index : ' + index);
var toSet = { };
var toUnset = { };
var over = function(){
var afterOver = function(err){
if(err) return afterCompleteOneCollection(err);
if(index < docs.length - 1) spec(index+1);
else afterCompleteOneCollection(null);
};
var sb = Object.keys(toSet).length;
var ub = Object.keys(toUnset).length;
if(sb || ub) {
var all = {};
if(sb) all.$set = toSet;
if(ub) all.$unset = toUnset;
Model.collection.update({ _id : docs[index]._id }, all, {}, afterOver);
} else afterOver(null);
};
forEachOfDocument(docs[index], toSet, toUnset, over);
};
spec(0);
});
Is there any better solution for the same.?
The streaming approach from here http://mongodb.github.io/node-mongodb-native/api-generated/cursor.html#stream worked for me
This is what i am doing :
var stream = Model.collection.find().stream();
stream.on('data', function(data){
if(data){
var toSet = { };
var toUnset = { };
var over = function(){
var afterOver = function(err){
if(err) console.log(err);
};
var sb = Object.keys(toSet).length;
var ub = Object.keys(toUnset).length;
if(sb || ub) {
var all = {};
if(sb) all.$set = toSet;
if(ub) all.$unset = toUnset;
Model.collection.update({ _id : data._id }, all, {}, afterOver);
} else afterOver(null);
};
forEachOfDocument(data, toSet, toUnset, over);
}
});
stream.on('close', function() {
afterCompleteOneCollection();
});
I have a document which includes a subdocument:
{
"_id" : ObjectId("XXXXX"),
"SearchKey" : "1234",
"SearchTerms" : {
"STA" : ["1"],
"STB" : ["asdfasdf"],
"STC" : ["another"]
}
}
The SearchTerm elements are not fixed - sometimes we'll have STA without STC, for example.
I can do this:
var map = function() {
for (key in this.SearchTerms)
{
emit(key, 1);
}
}
but I can't do this:
var map = function() {
for (var i=0; i< this.SearchTerms.length; i++)
{
emit(this.SearchTerms[i], 1)
}
}
because the latter doesn't produce any results after the reduce. Why not?
As an aside - what I need to do is count the cross-product of the search terms over all documents, that is, find the incidence of (STA and STB) and (STA and STC) and (STB and STC) in the case above. If someone knows how to do that right away, that works even better.
As always, thanks for the help
The key that you emit should be a composite of both keys.
var map = function() {
if(!this.SearchTerms) return;
for(var i = 0 ; i < this.SearchTerms.length; i++){
var outerKey = this.SearchTerms[i];
for(var j = i + 1; j < this.SearchTerms.length; j++){
var innerKey = this.SearchTerms[j];
// assuming you don't care about counting (STA and STC) separately from (STC and STA),
// then order doesn't matter, lets make sure both occurences have the same key.
var compositeKey = (outerKey < innerKey) ? outerKey+"_"+innerKey : innerKey+"_"+outerKey;
emit(compositeKey, 1);
}
}
}
This is because this.SearchTerms is a dictionary/subdocument and not an array. this.SearchTerms[0] doesn't exist.
For the second question: something like this should work:
for (key1 in this.SearchTerms)
{
for (key2 in this.SearchTerms)
{
if (key1 < key2)
{
key = [key1, key2];
emit(key, 1);
}
}
}