set will write information to the data store regardless of whether the key for the value already exists or not. The method is destructive; if the key exists, it will overwrite any existing value. Typically you want to use set in cases where you do not care whether or not you overwrite an existing value, nor do you care if the key already exists or not. This method is similar to an INSERT statement in SQL.

For instance, if you have a player location document in a game, you might not care whether you overwrite the location with a new value; it is however important that you quickly create a location document if it does not already exist. In the case of this type of application logic, you might not want to waste any code to check if a player location exists; performing rapid read/writes of the player location and creating the initial score document may be more important than performing any checks in your application logic. In this case, using set would be suitable.

Another scenario is when you populate a database with initial values. This can be a production or development database. In this case, you are creating all the initial values for an entire set of keys. Since you are starting out with an empty database, and have no risk of overwriting useful data, you would use set here as well. For instance, if you are populating your new test database with documents that represent different planets, you could follow this approach:

Figure 2.1. Creating Items with Set

Another scenario that is appropriate for using set is another scenario where you do not care about overwriting the last value for a key. For instance if you want to item the last visitor to your site, you would store that as a key and update it each time a new visitor is at your site. You might not care who the previous visitors are; in this case, you use set to overwrite anything that exists and replace it with the latest visit information.

This method is the functional equivalent of a RDMS commit/insert. Set will support the following parameters which are used during the operation:

The following shows a simple example of using set using the Ruby SDK:

c.set("foo", "bar", :ttl => 2)

This operation takes the key foo and sets the string 'bar' for the key which will expire after 2 seconds. This next example is part of a data loader script in PHP which reads in different JSON files in a specified directory. It then sends requests to write each file to Couchbase Server:

function import($cb, $dir) {
    $d = dir($dir);
    while (false !== ($file = $d->read())) {
        if (substr($file, -5) != '.json') continue;
            echo "adding $file\n";
        $json = json_decode(file_get_contents($dir . $file), true);
        unset($json["_id"]);
        echo $cb->set(substr($file, 0, -5), json_encode($json));
        echo "\n";
    }
}

The first part of this function takes a Couchbase client instance as a parameter and a directory. It then assigns the directory to a local variable $d and opens it. The while loop will continue reading each file in the directory so long as we have not finished reading all the files. In the loop we detect if the file has the .json file type or not so we know to handle it, and store it. If the file is json we decode it, remove the attribute _id and then set the key as the filename with the other file contents as value. We choose this different key for better identification in our system. The following illustrates a sample JSON file which you can use with this loader:

{
  "_id":"beer_#40_Golden_Lager",
  "brewery":"Minocqua Brewing Company",
  "name":"#40 Golden Lager",
  "updated":"2010-07-22 20:00:20"
}

To view the complete loader application and sample data available on GitHub, see Couchbase Labs: Beer Sample

In Couchbase SDKs you can store a value with set while simultaneously providing a document expiration.

Set will return a boolean of true if it is able to successfully commit data to the databases; it can also return a CAS value, which is a unique identifier for the item, and is used for optimistic locking.

The associated memcached protocol in ASCII is set which stores a key. For more information, see memcached protocol

The types of errors that can occur during this operation include 1) inability to connect to a node, or 2) some error exists while attempting to format a value being set. If you have a connection-level error you may need to reattempt connection, and possibly check the status of the server. If you have an error with the size of your value or formatting, you need to check the value itself, and how it is encoded and see if there are any issues that make the item incompatible with Couchbase Server.

For more information about connections and connection-level settings, see Section 6.4.4, “Optimizing Client Instances” and Section 6.6.1, “Client-Side Timeouts”