ElasticStore provides a high-level MongoDB-like API for ElasticSearch. It aims to be minimal.
$ npm install elasticstorevar elasticstore = require('elasticstore');
var store = new elasticstore.Store();
function Post(data) {
elasticstore.Node.call(this, data);
}
Post._type = 'post';
Post._schema = {
title: String,
created: {type: Date, default: Date.now}
};
require('util').inherits(Post, Node);
store.registerType(Post);
Post.insert({
title: 'Hello world'
}).then(function(post) {
console.log(post);
});A store is corresponding to an index. A model is corresponding to a document type in ElasticSearch.
connect(), connect to ElasticSearchmodel(name, schema), create a new modelregisterType(Type), to create a subclass of Node, include a call toutil.inherits(Type, Node).
Access the Node class by store.Node.
-
getContext -
getStore -
getStoreClient -
getIndex -
getType -
find(q), find all matching document. if q is null, return all documents -
findById(id), (getis aliased) find the document with the given id -
save(doc_or_docs_array), insert a new document to this model -
updateById(id, values) -
removeById(id), (removeis aliased) remove all documents in this model -
count(q), return the number of documents matching the query -
createInstance(data), create a new document of this model
To create a subclass, use the following snippet:
function Post() {
Node.call(this);
}
Node.registerType(Post);Then you need to access the methods by Post.findById etc.
You can create a new Model instance by new Model().
getContext()save()remove()update()
ORG_PKG=/Users/cllu/Projects/OrganizedApp/ ./vendor/elasticsearch-1.4.4/bin/elasticsearch -D es.config=/Users/cllu/Projects/OrganizedApp/etc/elasticsearch/elasticsearch.yml
$ gulp testAfter ElasticSearch indexes a document, it needs some time before the document is searchable.
To simplify the logic, the save() method has a refresh optional parameter which is True by default.
Ensure to set index.mapping.coerce: false so that ES will not try to coerce numerical values.