Skip to main content

Key Concepts

Assumptions

  1. Small Selector Classes - Selector class should be small and contains ONLY query base configuration (fields, sharing settings) and very generic methods (getById, getByRecordType). Why?
    • Huge classes are hard to manage.
    • A lot of merge conflicts.
    • Problems with methods naming.
  2. Build SOQL inline in a place of need - Business specific SOQLs should be build inline via SOQL builder in a place of need.
    • Most of the queries on the project are case specific and are not generic. There is no need to keep them in Selector class.
  3. Build SOQL dynamically via builder - Developer should be able to adjust query with specific fields, conditions, and other SOQL clauses.
  4. Do not spend time on selector methods naming - It can be difficult to find a proper name for method that builds a query. Selector class contains methods like selectByFieldAAndFieldBWithDescOrder. It can be avoided by building SOQL inline in a place of need.
  5. Controll FLS and sharing settings - Selector should allow to control Field Level Security and sharing settings by the simple methods like .systemMode(), .withSharing(), .withoutSharing().
  6. Auto binding - Selector should be able to bind variables dynamically without additional effort from developer side.
  7. Mock results in Unit Tests - Selector should allow for mocking data in unit tests.

Concepts

SOQL Library consist of:

  • SOQL Builder
  • SOQL Selector

SOQL Builder

SOQL Builder allows to build query dynamically and execute it.

// SELECT Id, Name, Industry FROM Account
List<Account> accounts = SOQL.of(Account.SObjectType)
.with(new List<SObjectField>{
Account.Id, Account.Name, Account.Industry
}).toList();

SOQL Selector

A selector layer contains code responsible for querying records from the database. Although you can place SOQL queries in other layers, a few things can happen as the complexity of your code grows. ~ Salesforce

SOQL Lib provides the whole new concept for Selectors usage.

Old Approach

FFLIB Selector concept assumes that all queries are be stored in Selector class.

  • To avoid duplicates.
  • One place to manage all queries.

Selector Issues:

  • One-time queries (like aggregation, case specific) added to Selector.
  • Huge class with a lot of methods.
  • Queries are difficult for reuse.
  • Similar methods with small differences like limit, offset.
  • Problem with naming methods.
  • Merge conflicts.

New Approach

SOQL Lib has slightly different approach.

Assumption: Most of the SOQLs on the project are one-time queries executed for specific business case.

Solution:

  1. Small Selector Classes - Selector class should be small and contains ONLY query base configuration (fields, sharing settings) and very generic methods (getById, getByRecordType)
  2. Build SOQL inline in a place of need - Business specific SOQLs should be build inline via SOQL builder in a place of need.
  3. Do not spend time on selector methods naming - Queries are created inline, so not need to find a name.
  4. Keep Selector Strengths - Set default Selector configuration (default fields, sharing settings), keep generic methods.
public with sharing class AccountSelector implements SOQL.Selector {
public static SOQL query() {
return SOQL.of(Account.SObjectType)
.with(new List<SObjectField>{
Account.Name,
Account.AccountNumber
})
.systemMode()
.withoutSharing();
}

public static SOQL getByRecordType(String rt) {
return query.with(new List<SObjectField>{
Account.BillingCity,
Account.BillingCountry
}).whereAre(SOQL.Filter.recordType().equal(rt));
}
}
public with sharing class ExampleController {

public static List<Account> getAccounts(String accountName) {
return AccountSelector.query()
.with(Account.BillingCity)
.with(Account.BillingCountry)
.whereAre(SOQL.Filter.with(Account.Name).contains(accountName))
.toList();
}

public static List<Account> getAccountsByRecordType(String recordType) {
return AccountSelector.getByRecordType(recordType)
.with(Account.ParentId)
.toList();
}
}