Fields

Guillotina uses schemas to define content types and behaviors. These schemas consist of field definitions.

Available fields

  • guillotina.schema.Bool
  • guillotina.schema.Bytes
  • guillotina.schema.Choice: validates against vocabulary of values
  • guillotina.schema.Date
  • guillotina.schema.Datetime
  • guillotina.schema.Decimal
  • guillotina.schema.Dict
  • guillotina.schema.Float
  • guillotina.schema.Int
  • guillotina.schema.JSONField
  • guillotina.schema.List
  • guillotina.schema.Set
  • guillotina.schema.Text
  • guillotina.schema.TextLine
  • guillotina.schema.Time
  • guillotina.fields.PatchField: allow updating value without patching entire value
  • guillotina.fields.BucketListField: optimized storage for very large lists of data
  • guillotina.fields.BucketDictField: optimized storage for very large dictionaries of data
  • guillotina.files.CloudFileField: file field for storing in db or cloud storage

Patch field

Guillotina provides a PatchField which allows you to patch values of List, Dict and Int fields without having the original value. This is done doing a PATCH request to the object absolute url with the following payloads:

Patch field list

from zope.interface import Interface
from guillotina.fields import PatchField
from guillotina import schema

class IMySchema(Interface):
    field = PatchField(schema.List(
        value_type=schema.Text()
    ))

Then, payload for patching to append to this list would look like:

{
    "field": {
        "op": "append",
        "value": "foobar"
    }
}

Append if unique value only:

{
    "field": {
        "op": "appendunique",
        "value": "foobar"
    }
}

Extend:

{
    "field": {
        "op": "extend",
        "value": ["foo", "bar"]
    }
}

Extend if unique values:

{
    "field": {
        "op": "extendunique",
        "value": ["foo", "bar"]
    }
}

Delete:

{
    "field": {
        "op": "del",
        "value": 0
    }
}

Remove:

{
    "field": {
        "op": "remove",
        "value": "foobar"
    }
}

Update:

{
    "field": {
        "op": "update",
        "value": {
            "index": 0,
            "value": "Something new"
        }
    }
}

Patch dict field

from zope.interface import Interface
from guillotina.fields import PatchField
from guillotina import schema

class IMySchema(Interface):
    field = PatchField(schema.Dict(
        key_type=schema.Text(),
        value_type=schema.Text()
    ))

Then, payload for patching to add to this dict would look like:

{
    "field": {
        "op": "assign",
        "value": {
            "key": "foo",
            "value": "bar"
        }
    }
}

Delete:

{
    "field": {
        "op": "del",
        "value": "foo"
    }
}

Update:

{
    "values": {
        "op": "update",
        "field": [{
            "key": "foo",
            "value": "bar"
        }, {
            "key": "foo2",
            "value": "bar2"
        }]
    }
}

Patch int field

PatchField can also be used on Int fields to increment, decrement or reset their original value.

from zope.interface import Interface
from guillotina.fields import PatchField
from guillotina import schema

class IMySchema(Interface):
    counter = PatchField(schema.Int(
        title='My Counter',
        default=1,
    ))

The payload to increment counter by 3 units would look like:

{
    "counter": {
        "op": "inc",
        "value": 3
    }
}

Notice that, at this point, counter will be set to 4 because its default value is 1. If the default would not be set, the increment operation assumes a 0, and thus counter would be 3.

Likewise, to decrement the field, the following payload would work:

{
    "counter": {
        "op": "dec",
        "value": 4
    }
}

To reset counter to its default value, you can send the following payload without value:

{
    "counter": {
        "op": "reset"
    }
}

and counter will be set to its default value 1. Otherwise, you can also send the target reset value:

{
    "counter": {
        "op": "reset",
        "value": 0
    }
}

Notice that a reset operation on a integer without a default value is equivalent to sending a value of 0.

Bucket list field

from zope.interface import Interface
from guillotina.fields import BucketListField
from guillotina import schema

class IMySchema(Interface):
    field = BucketListField(
        value_type=schema.Text(),
        bucket_len=5000
    )

Then, payload for patching to append to this list would look like:

{
    "field": {
        "op": "append",
        "value": "foobar"
    }
}

Extend:

{
    "field": {
        "op": "extend",
        "value": ["foo", "bar"]
    }
}

Delete:

{
    "field": {
        "op": "del",
        "value": {
            "bucket_index": 0,
            "item_index": 0
        }
    }
}

Bucket dict field

from zope.interface import Interface
from guillotina.fields import BucketDictField
from guillotina import schema

class IMySchema(Interface):
    field = BucketDictField(
        key_type=schema.Text(),
        value_type=schema.Text(),
        bucket_len=5000
    )

Then, payload for patching would be…:

{
    "field": {
        "op": "assign",
        "value": {
            "key": "foo",
            "value": "bar"
        }
    }
}

Update:

{
    "field": {
        "op": "update",
        "value": [{
            "key": "foo",
            "value": "barupdated"
        }, {
            "key": "other",
            "value": "othervalue"
        }]
    }
}

Delete:

{
    "field": {
        "op": "del",
        "value": "foo"
    }
}

Field validation

constraints

To provide additional field validation, you can provide a constraint for a field.

This is a simple callable that takes an argument with the value that is being validated for the field.

The most simple type of constraint is to use a lambda function:

from zope.interface import Interface
from guillotina import schema

class IMySchema(Interface):
    field = schema.Text(constraint=lambda val: val != 'foobar')

field validators

To provide more complex field validation, you can use the validator field decorator:

from zope.interface import Interface
from guillotina import schema

class IMySchema(Interface):
    field = schema.Text()

    @field.validator
    def validate_field(field, value):
        # field is bound field so we can look at context now
        return field.context.foobar is None

invariants

You can also validate the modified objects with invariants.

from zope.interface import Interface, invariant, Invalid
from guillotina import schema

class IMySchema(Interface):
    field = schema.Text()

    @invariant
    def validate_obj(obj):
        # field is bound field so we can look at context here
        if obj.foo is None and obj.bar is None:
            raise Invalid(obj)