SOQL Field Coverage Rules

SOQL Field Coverage Rules

This guide documents field coverage validation rules for SOQL queries—ensuring that all fields accessed in Apex code are actually queried. This is a common source of runtime errors, especially in LLM-generated code.

Source: LLM Mistakes in Apex & LWC - Salesforce Diaries

Table of Contents

1. The Field Coverage Problem
2. Direct Field Access
3. Relationship Field Access
4. Dynamic Field Access
5. Aggregate Queries
6. Subquery Fields
7. Validation Patterns

The Field Coverage Problem

When you query an sObject, only the fields in the SELECT clause are populated. Accessing any other field results in a runtime error:

System.SObjectException: SObject row was retrieved via SOQL without querying the requested field: Account.Industry

This error is particularly common in LLM-generated code because the LLM may:

1. Query some fields but access others in subsequent code
2. Forget to include relationship fields (e.g., Account.Name on Contact)
3. Access fields in conditional logic that weren't anticipated in the query

Direct Field Access

Rule: Every field accessed must be in the SELECT clause

❌ BAD: Accessing Unqueried Fields

// Query only includes Id and Name
List<Account> accounts = [SELECT Id, Name FROM Account];

for (Account acc : accounts) {
    // RUNTIME ERROR: Industry was not queried
    if (acc.Industry == 'Technology') {
        // RUNTIME ERROR: Description was not queried
        acc.Description = 'Tech company';
    }

    // RUNTIME ERROR: AnnualRevenue was not queried
    Decimal revenue = acc.AnnualRevenue;
}

✅ GOOD: Query All Accessed Fields

// Query ALL fields that will be accessed
List<Account> accounts = [
    SELECT Id, Name, Industry, Description, AnnualRevenue
    FROM Account
];

for (Account acc : accounts) {
    if (acc.Industry == 'Technology') {
        acc.Description = 'Tech company';  // OK - queried
    }
    Decimal revenue = acc.AnnualRevenue;  // OK - queried
}

Field Access Locations to Check

Fields can be accessed in many places—ensure coverage for all:

Relationship Field Access

Rule: Parent relationship fields require dot notation in SELECT

❌ BAD: Missing Relationship Fields

// Contact query without Account relationship fields
List<Contact> contacts = [SELECT Id, Name, AccountId FROM Contact];

for (Contact c : contacts) {
    // RUNTIME ERROR: Account.Name was not queried
    String accountName = c.Account.Name;

    // RUNTIME ERROR: Account.Industry was not queried
    if (c.Account.Industry == 'Technology') {
        // ...
    }
}

✅ GOOD: Include Relationship Fields

// Use dot notation to include parent fields
List<Contact> contacts = [
    SELECT Id, Name, AccountId,
           Account.Name,           // Parent field
           Account.Industry,       // Parent field
           Account.Owner.Name      // Grandparent field (up to 5 levels)
    FROM Contact
];

for (Contact c : contacts) {
    String accountName = c.Account.Name;  // OK - queried

    if (c.Account.Industry == 'Technology') {  // OK - queried
        String ownerName = c.Account.Owner.Name;  // OK - queried
    }
}

Relationship Traversal Limits

❌ BAD: Assuming Relationship is Populated

List<Contact> contacts = [SELECT Id, AccountId FROM Contact];

for (Contact c : contacts) {
    // AccountId is queried, but Account object is NOT populated
    // This will throw: Account.Name not queried
    if (c.Account != null) {
        String name = c.Account.Name;  // ERROR!
    }
}

✅ GOOD: Query Relationship or Use Separate Query

// Option 1: Include relationship field
List<Contact> contacts = [SELECT Id, AccountId, Account.Name FROM Contact];

for (Contact c : contacts) {
    if (c.Account != null) {
        String name = c.Account.Name;  // OK
    }
}

// Option 2: Separate query using collected IDs
List<Contact> contacts = [SELECT Id, AccountId FROM Contact];
Set<Id> accountIds = new Set<Id>();
for (Contact c : contacts) {
    if (c.AccountId != null) {
        accountIds.add(c.AccountId);
    }
}

Map<Id, Account> accountMap = new Map<Id, Account>(
    [SELECT Id, Name FROM Account WHERE Id IN :accountIds]
);

for (Contact c : contacts) {
    Account acc = accountMap.get(c.AccountId);
    if (acc != null) {
        String name = acc.Name;  // OK
    }
}

Dynamic Field Access

Rule: Dynamic field access (using get()) also requires queried fields

❌ BAD: Dynamic Access to Unqueried Field

List<Account> accounts = [SELECT Id, Name FROM Account];
String fieldName = 'Industry';  // Dynamic field name

for (Account acc : accounts) {
    // RUNTIME ERROR: Industry was not queried
    Object value = acc.get(fieldName);
}

✅ GOOD: Query Fields Used Dynamically

// If you know which fields will be accessed dynamically, query them
List<Account> accounts = [SELECT Id, Name, Industry FROM Account];
String fieldName = 'Industry';

for (Account acc : accounts) {
    Object value = acc.get(fieldName);  // OK - Industry is queried
}

✅ GOOD: Build Dynamic Query

// For truly dynamic scenarios, build the query dynamically
Set<String> fieldsToQuery = new Set<String>{'Id', 'Name'};
fieldsToQuery.addAll(dynamicFieldList);  // Add dynamic fields

String query = 'SELECT ' + String.join(new List<String>(fieldsToQuery), ', ') +
               ' FROM Account WHERE Id IN :accountIds';

List<Account> accounts = Database.query(query);

Aggregate Queries

Rule: Aggregate queries return AggregateResult, not sObjects

❌ BAD: Treating Aggregate as sObject

// This returns AggregateResult, not Account
List<Account> accounts = [
    SELECT Industry, COUNT(Id) cnt
    FROM Account
    GROUP BY Industry
];  // COMPILE ERROR - wrong type

// Even with correct type, can't access normal fields
AggregateResult[] results = [
    SELECT Industry, COUNT(Id) cnt
    FROM Account
    GROUP BY Industry
];

for (AggregateResult ar : results) {
    // Cannot access like sObject fields
    String industry = ar.Industry;  // COMPILE ERROR
}

✅ GOOD: Use get() for Aggregate Results

AggregateResult[] results = [
    SELECT Industry, COUNT(Id) cnt, SUM(AnnualRevenue) totalRevenue
    FROM Account
    GROUP BY Industry
];

for (AggregateResult ar : results) {
    // Use get() with field alias
    String industry = (String) ar.get('Industry');
    Integer count = (Integer) ar.get('cnt');
    Decimal totalRevenue = (Decimal) ar.get('totalRevenue');

    System.debug(industry + ': ' + count + ' accounts, $' + totalRevenue);
}

Aggregate Field Aliases

Subquery Fields

Rule: Child relationship subqueries create nested lists

❌ BAD: Accessing Subquery Fields Incorrectly

// Query with contact subquery
List<Account> accounts = [
    SELECT Id, Name,
           (SELECT Id, Name FROM Contacts)
    FROM Account
];

for (Account acc : accounts) {
    // ERROR: Contacts is a List, not a single Contact
    String contactName = acc.Contacts.Name;

    // ERROR: Cannot access unqueried field from subquery
    for (Contact c : acc.Contacts) {
        String email = c.Email;  // Email not in subquery SELECT!
    }
}

✅ GOOD: Proper Subquery Field Access

// Query all needed fields in subquery
List<Account> accounts = [
    SELECT Id, Name,
           (SELECT Id, Name, Email, Phone FROM Contacts)
    FROM Account
];

for (Account acc : accounts) {
    // Contacts is a List<Contact>
    List<Contact> contacts = acc.Contacts;

    if (contacts != null && !contacts.isEmpty()) {
        for (Contact c : contacts) {
            String name = c.Name;    // OK - in subquery SELECT
            String email = c.Email;  // OK - in subquery SELECT
            String phone = c.Phone;  // OK - in subquery SELECT
        }
    }
}

Subquery Null Safety

List<Account> accounts = [
    SELECT Id, (SELECT Id FROM Contacts)
    FROM Account
];

for (Account acc : accounts) {
    // Subquery result can be null if no child records
    if (acc.Contacts != null) {
        for (Contact c : acc.Contacts) {
            // Process contact
        }
    }

    // Or use null-safe size check
    Integer contactCount = acc.Contacts?.size() ?? 0;
}

Validation Patterns

Pattern 1: Field-to-Query Mapping

Create a systematic approach to track field usage:

public class AccountProcessor {
    // Document required fields at the top
    private static final Set<String> REQUIRED_FIELDS = new Set<String>{
        'Id', 'Name', 'Industry', 'Description', 'AnnualRevenue',
        'OwnerId', 'Owner.Name', 'Owner.Email'
    };

    // Single method for consistent querying
    public static List<Account> queryAccounts(Set<Id> accountIds) {
        return [
            SELECT Id, Name, Industry, Description, AnnualRevenue,
                   OwnerId, Owner.Name, Owner.Email
            FROM Account
            WHERE Id IN :accountIds
        ];
    }

    public static void processAccounts(List<Account> accounts) {
        for (Account acc : accounts) {
            // All fields in REQUIRED_FIELDS are safe to access
            if (acc.Industry == 'Technology') {
                acc.Description = 'Tech: ' + acc.Name;
            }
        }
    }
}

Pattern 2: Selector Layer

Use a selector pattern to centralize query field management:

public class AccountSelector {

    // Default fields for most operations
    private static final List<String> DEFAULT_FIELDS = new List<String>{
        'Id', 'Name', 'Industry', 'Type', 'OwnerId'
    };

    // Extended fields for detailed views
    private static final List<String> DETAIL_FIELDS = new List<String>{
        'Id', 'Name', 'Industry', 'Type', 'OwnerId',
        'Description', 'AnnualRevenue', 'NumberOfEmployees',
        'BillingCity', 'BillingState', 'BillingCountry',
        'Owner.Name', 'Owner.Email'
    };

    public List<Account> selectById(Set<Id> ids) {
        return selectByIdWithFields(ids, DEFAULT_FIELDS);
    }

    public List<Account> selectByIdDetailed(Set<Id> ids) {
        return selectByIdWithFields(ids, DETAIL_FIELDS);
    }

    private List<Account> selectByIdWithFields(Set<Id> ids, List<String> fields) {
        String query = 'SELECT ' + String.join(fields, ', ') +
                       ' FROM Account WHERE Id IN :ids';
        return Database.query(query);
    }
}

Pattern 3: Field Validation Helper

public class SObjectFieldValidator {

    /**
     * Check if a field was queried on an sObject
     * @param obj The sObject to check
     * @param fieldName The API name of the field
     * @return true if the field is populated (was queried)
     */
    public static Boolean isFieldPopulated(SObject obj, String fieldName) {
        try {
            obj.get(fieldName);
            return true;
        } catch (SObjectException e) {
            return false;
        }
    }

    /**
     * Get field value with default if not queried
     * @param obj The sObject
     * @param fieldName The field API name
     * @param defaultValue Value to return if field not queried
     * @return The field value or default
     */
    public static Object getFieldOrDefault(SObject obj, String fieldName, Object defaultValue) {
        try {
            Object value = obj.get(fieldName);
            return value != null ? value : defaultValue;
        } catch (SObjectException e) {
            return defaultValue;
        }
    }
}

Quick Reference: Field Coverage Checklist

Before running code that processes SOQL results:

Direct Fields

• All fields in if conditions are queried
• All fields on left side of assignments are queried
• All fields passed to methods are queried
• All fields used in map keys/values are queried

Relationship Fields

• Parent fields use dot notation (e.g., Account.Name)
• Parent object null checks before field access
• Relationship traversal doesn't exceed 5 levels

Subqueries

• Child records accessed as List, not single record
• Subquery SELECT includes all accessed child fields
• Null check before iterating subquery results

Dynamic Access

• Fields accessed via get(fieldName) are queried
• Dynamic queries include all needed fields

Common LLM Mistakes Summary

Reference

• SOQL Anti-Patterns: See docs/anti-patterns.md for general SOQL mistakes
• Selector Patterns: See docs/selector-patterns.md for query organization
• Source: Salesforce Diaries - LLM Mistakes