Edit on GitHub
Jump to docs navigation

Extending / Storage Layer / Fields & Field Types

Note: You are currently reading the documentation for Bolt 3.7. Looking for the documentation for Bolt 5.2 instead?

Overview

Fields are where the majority of custom functionality happen within Bolt, at their simplest they are a way to link a single part of a content record to a persisted state in the database, but for Bolt 3 the flexibility available to fields has increased greatly.

To understand the different processes that a field goes through, we'll start by looking at the simplest of the fields, the text field. Whenever you define a field in contenttypes.yml you let Bolt know what kind of behaviour you want the field to have, here's an example:

    fields:
        title:
            type: text
            group: content

In the Bolt backend a simple text field like this will show an HTML text input, and we'd expect the value to be collected, saved in a text column in the database and made available via a content record for the frontend, the normal way we access it is via the {{ record.title }} syntax in Twig.

The three stages that we talk about in all discussions of fields relate to these processes and, in turn, each field whether simple or complex will need to implement these stages as methods. They are load(), hydrate() and persist().

The Load Method

The load method is responsible for describing how the field is to be fetched from the database. Bolt uses Doctrine QueryBuilder to construct the query and by default when content records are being queried all the columns will be fetched, in SQL terms something very similar to a SELECT * from bolt_pages is taking place, so since our simple text field is just needed to be fetched as is from the database, we don't need to do anything here.

Most fields within Bolt will behave the same as this since generally they map to a simple column in the database, but having access to the QueryBuilder object here does allow you to do some advanced manipulation on the query, if you want to see the power of this then look at the source code of fields that use join tables to bring in relational data to a field.

The Hydrate Method

Hydration is the process of converting a flat array of database storage data into PHP objects. For instance a date field stored in the database as 2016-01-01 needs to be translated into a DateTime object or a JSON array stored as a string in the database needs to be converted to an array by passing through json_decode.

Any conversion that needs to take place after a record has been fetched from the database can be handled in the hydrate() method of a field. Out text field is quite simple since we don't need to do anything more complicated than taking the string from the database and set the property of the entity. So this would be all that you need to do in the hydrate() method.

    public function hydrate($data, $entity)
    {
        $key = $this->mapping['fieldname'];
        $val = $data[$key] ?: null;
        $entity->$key = $val;
    }

To walk through this, let's first look at the values that our hydrate method receives. Firstly $data is an array of the entire data array for one row from the database. In most cases we would only be interested in the specific field that we are processing, so our example field that we defined in YAML above would be available at $data['title'] again if you want to do more complex things having access here to the raw data may be useful.

The second parameter is the constructed $entity object, as part of the hydration process the entity will start off empty but as it gets hydrated by each field it will end up with all the field properties being set.

So knowing this, the three lines of our hydrate method are quite simple to understand. First of all we look up the name of our field. This is whatever we configured in our contenttypes.yml file and you can fetch all this data via $this->mapping which is an array of the configuration defined in YAML along with some other defaults.

Once we know the key, we can fetch that item from the $data array and then in the final line of code we set the entity object property with the value that came from the database. Once this is done, then Bolt will move onto the next field until the entity is fully hydrated.

The Persist Method

The persist method is the counterpart to hydrate and this performs the reverse of hydration in that it takes an entity gets the value of our field, does any conversion you want the field to do and finally adds the final value to the QueryBuilder object so it can be written back to the database.

As with the load method, having access to the entire Query stack gives us a lot of potential power that in most cases we don't need to use but should you need to then again very complex DB queries can be accomplished should you wish to do smoething more advanced in a field.

Let's run through the simple example of what our text field persist method will look like. Again because we picked the simplest of all fields we don't need to perform and translations on the values, we just want the text value from our entity to end up in the correct column in the database. So here is how the method will look:

    public function persist(QuerySet $queries, $entity)
    {
        $key = $this->mapping['fieldname'];
        $qb = $queries[0];
        $value = $entity->$key;

        $qb->set($key, ':' . $key);
        $qb->setParameter($key, $value);
    }

Again we are passed two parameters that we will need, the first $queries is a set of queries, the second is our hydrated entity. With our simple text field we just need to read the value of title from the entity and add it to the update query.

The $queries parameter is a QuerySet object, with simple field operations we are only ever concerned with the main update query, in SQL terms it will look like UPDATE bolt_pages SET title=:title, content=:content ....... WHERE id=1;

So unless you are trying to do something more complicated then the first step is to get the main query which is available via array access $queries[0]. Once we have that we are going to add the value to the query. As before we fetch the value from the entity $value = $entity->$key then we add a set command to the query $qb->set($key, ':' . $key) this makes the parameterised query for the Doctrine QueryBuilder equivalent to adding: SET title=:title and then binding the value of :title to the value of the entity.



Edit this page on GitHub
Couldn't find what you were looking for? We are happy to help you in the forum, on Slack or on Github.