Members

static tableName :String

models.js, line 840

The table name this model is associated with, excluding the service ID prefix. This is the model's class name. However, subclasses may choose to override this method and provide duplicated table name for co-existed models.

FIELDS

models.js, line 677

Defines the non-key fields. By default there are no fields. Properties are defined as a map from field names to a Todea schema:

Example

static FIELDS = {
    someNumber: S.double,
    someNumberWithOptions: S.double.optional().default(0).readOnly()
  }

INDEX_INCLUDE_KEYS

models.js, line 701

If this is enabled, we will create individual fields internally for all the field in KEY, SORT_KEY. This will incrase the storage cost but enable users to use lazy filter on individual key fields. This works only for querying via indexes, if the key field isn't part of a key in that index.

INDEXES

models.js, line 692

Defines the global secondary indexes for the table. By default there are no secondary indexes. Properties are defined as a map from index names to an object containing the keys for the index (fields defined in an index key should already exist in the table):

Example

static INDEXES = {
    index1: { KEY: [field1, field2] }
    index2: { KEY: [field3], SORT_KEY: [field4] }
  }

KEY

models.js, line 662

Defines the partition key. Every item in the database is uniquely identified by the combination of its partition and sort key. The default partition key is a UUIDv4. A key can simply be some scalar value: static KEY = { id: S.str } A key may can be "compound key", i.e., a key with one or components, each with their own name and schema: static KEY = { email: S.str, birthYear: S.int.min(1900) }

SORT_KEY

models.js, line 665

Defines the sort key, if any. Uses the compound key format from KEY.

Methods

static __computeKeyAttrMap(vals)

models.js, line 811

Returns a map containing the model's computed key values (_id, as well as _sk if model has a sort key). Verifies that the keys are valid (i.e., they match the required schema).

Parameters:

Name	Type	Description
vals	Object	map of field names to values

Returns:

- map of _id (and _sk attribute values)

static __decodeCompoundValue(keyOrder, strVal, attrName)

models.js, line 1261

Returns the map which corresponds to the given compound value string This method throws InvalidFieldError if the decoded string does not match the required schema.

Parameters:

Name	Type	Description
keyOrder	Array.<String>	order of keys in the string representation
strVal	String	the string representation of a compound value
attrName	String	which key we're parsing

static __doOneTimeModelPrep()

models.js, line 364

Check that field names don't overlap, etc.

static __encodeCompoundFieldName(fieldsopt)

models.js, line 781

Generate a compound field name given a list of fields. For compound field containing a single field that is not either a PK or SK, we use the same name as the original field to reduce data duplication. We also auto-detect if _id, or _sk can be re-used

Parameters:

Name	Type	Attributes	Description
fields		<optional>	a list of string denoting the fields

Returns:

- a string denoting the compound field's internal name

static __encodeCompoundValue(keyOrder, values)

models.js, line 1217

Returns the string representation for the given compound values. This method throws InvalidFieldError if the compound value does not match the required schema.

Parameters:

Name	Type	Description
keyOrder	Array.<String>	order of keys in the string representation
values	Object	maps component names to values; may have extra fields (they will be ignored)

package, static __getParams(encodedKeys, options) → {Object}

models.js, line 892

Parameters:

Name	Type	Description
encodedKeys	CompositeID
options	GetParams

Returns:

Object - parameters for a get request to DynamoDB

static __splitKeysAndData(data)

models.js, line 871

Given a mapping, split compositeKeys from other model fields. Return a 3-tuple, [encodedKeys, keyComponents, modelData].

Parameters:

Name	Type	Description
data	Object	data to be split

static data(vals) → {Data}

models.js, line 1320

Returns a Data fully describing a unique row in this model's DB table.

Parameters:

Name	Type	Description
vals	*	like the argument to key() but also includes non-key data

Returns:

Data - a Data object for use with tx.create() or tx.get(..., { createIfMissing: true })

static key(vals) → {Key}

models.js, line 1300

Returns a Key identifying a unique row in this model's DB table.

Parameters:

Name	Type	Description
vals	*	map of key component names to values; if there is only one partition key field (whose type is not object), then this MAY instead be just that field's value.

Returns:

Key - a Key object.

package __conditionCheckParams() → {Boolean}

models.js, line 1199

Used for optimistic locking within transactWrite requests, when the model was read in a transaction, and was subsequently used for updating other models but never written back to DB. Having conditionCheck ensures this model's data hasn't been changed so the updates to other models are also correct.

Returns:

Boolean - An Object for ConditionCheck request.

__isMutated(expectWrites)

models.js, line 1172

Indicates if any field was mutated. New models are considered to be mutated as well.

Parameters:

Name	Type	Default	Description
expectWrites	Boolean	true	whether the model will be updated, default is true.

__nonexistentModelCondition()

models.js, line 1436

Returns:

- a [ConditionExpression, ExpressionAttributeNames, ExpressionAttributeValues] tuple to make sure the model does not exist on server.

package __putParams()

models.js, line 927

Generates parameters for a put request to DynamoDB. Put overrides item entirely, removing untracked fields from DynamoDB. This library supports optimistic locking for put. Since put overrides all fields of an item, optimistic locking is performed on all fields. This means if any fields is modified after the item is read calling put would fail. Effectively the lock applies to the entire item, which may lead to more contention. Have update in most use cases is more desirable.

Returns:

- parameters for a put request to DynamoDB

package __updateParams(omitUpdates)

models.js, line 1046

Generates parameters for an update request to DynamoDB. Update only overrides fields that got updated to a different value. Untracked fields will not be removed from DynamoDB. This library supports optimistic locking for update. Since update only touches specific fields of an item, optimistic locking is only performed on fields accessed (read or write). This locking mechanism results in less likely contentions, hence is preferred over put.

Parameters:

Name	Type	Default	Description
omitUpdates	Boolean	false	(default = false) When True, generates only condition expressions for read values; skipping update expressions, related Attribute Names/Values and schema validation, with the expectation that any accessed value is either unmodified (and therefore valid) or explicitly unchecked (written but not read).

Returns:

- parameters for a update request to DynamoDB

async, package __write()

models.js, line 1363

Writes model to database. Uses DynamoDB update under the hood.

async finalize()

models.js, line 117

Hook for finalizing a model before writing to database

getField(name) → {BooleanField|ArrayField|ObjectField|NumberField|StringField}

models.js, line 827

Returns the underlying __Field associated with an attribute.

Parameters:

Name	Type	Description
name	String	the name of a field from FIELDS

Returns:

BooleanField | ArrayField | ObjectField | NumberField | StringField

getSnapshot(params)

models.js, line 1492

Return snapshot of the model, all fields included.

Parameters:

Name	Type	Description
params	Object	Properties Name Type Description initial Boolean Whether to return the initial state dbKeys Boolean Whether to return _id and _sk instead of raw key fields.

toString()

models.js, line 1473

Must be the same as NonExistentModel.toString() because it is used as the unique identifier of an item for Objects and Sets.