Tables

---

NOTE: Examples work with gun version 0.2.x. Will be upgrading soon for 0.3.x

---

So you want Tabular data with GUN? Huh? You can do that too!

localStorage.clear(); // refresh the storage for this example.

var gun = Gun();
var table = gun.get('my/table').set();

function Person(name, age){
	this.name = name;
	this.age = age;
}

table.set(new Person('Alice', 24));

table.set(new Person('Bob', 25));

table.set(new Person('Carl', 24));

table.map(function(person){
	alert(person.name + " of age " + person.age + " is in the table!");
});

(try "run with js" on http://jsbin.com/fijajojala/edit?html,js,output)

What is happening?

First we need to instantiate GUN. Then we need to select which table we want to load FROM (actual SQL support coming soon!) by doing gun.get('my/table'). You also notice the .set()? Tables (or collections or lists or 'arrays') are called 'sets' in GUN, akin to a mathematical set. Calling .set() instantiates an empty set (or the null set) in case the table doesn't already exist.

Next up is the schema for the table! GUN does not require a schema, but using schemas is recommended - if you are interested in a schema extension for GUN check out RangerMauve's gun-schema. For this example we're going to simply use a class instead, thus the function Person(){}.

Now we want to insert records into our table! In GUN lingo, all we now have to do is add an item into the set. Thus table.set({name: 'Alice', age: 24});, but since we want a schema we call new Person('Alice', 24) instead. Let's repeat this a couple times adding records into our set.

Finally we want to read each record in the table. All we have to do is call table.map(function(record){})! However WARNING this also subscribes to all realtime updates too! That means an individual record might be called multiple times so do not be alarmed if you get multiple repeats. GUN by default is functional and reactive in design.

How do you check for uniqueness then? GUN has a unique primary ID called a 'soul' on every record in the database, however you can add your own ID as well and use that to check. But this leads to a bigger question...

How do Tables work with Graphs?

You can actually think of a graph as table with every record in it. Any other 'table' you create then has a record that just points to primary ID in the graph table. Pretty simple, huh? But it is not just tables that work with graphs, documents do too! With a graph you don't actually store the entire document as a single record, instead every sub object in the document is stored in the graph table and the document is reconstructed with pointers. Woohoo!

What is the whole Tuple thing?

People often throw around the word "tuple" in graph databases, however tuples are not necessary for graph databases. For instance, Neo4J uses a property graph model - and GUN is similar. That doesn't mean you cannot do tuples though, you can. But that is for a different section.