This not well-known trick can solve the problem in one line

 Set<String> accountNames = new Map<String,SObject>(
    [SELECT Name Id FROM Account GROUP BY Name]).keySet();

From the Matt and Neil blog post (of bigassforce fame)

How it works

The first thing to note is that we are using an Aggregate Query and not just ordinary SOQL. We combine special behaviours available to aggregate queries to build the set:

Step 1 - GROUP BY to collect unique values Exactly per the documentation, you can use the GROUP BY option in a SOQL query to avoid iterating through individual query results. That is, you specify a group of records instead of processing many individual records.

Step 2 - Alias the field as Id The default identifier for an aggregated field is expr0 but we will alias the result as Id. It doesn't matter what the field contains, so long as it's a string. We will exploit this special alias in the next step.

Step 3 - AggregateResult inherits from sObject The methods available on sObject are also available on AggregateResult. This also applies to Lists and Maps; those same methods can be called on collections of aggregate results. Here is another example using keySet()

Set<String> emails = new Map<String,SObject>(
    [SELECT Email Id FROM Contact GROUP BY Email]).keySet();

This not well-known trick can solve the problem in one line

 Set<String> accountNames = new Map<String,SObject>(
    [SELECT Name Id FROM Account GROUP BY Name]).keySet();

From the Matt and Neil blog post

How it works

The first thing to note is that we are using an Aggregate Query and not just ordinary SOQL. We combine special behaviours available to aggregate queries to build the set:

Step 1 - GROUP BY to collect unique values Exactly per the documentation, you can use the GROUP BY option in a SOQL query to avoid iterating through individual query results. That is, you specify a group of records instead of processing many individual records.

Step 2 - Alias the field as Id The default identifier for an aggregated field is expr0 but we will alias the result as Id. It doesn't matter what the field contains, so long as it's a string. We will exploit this special alias in the next step.

Step 3 - AggregateResult inherits from sObject The methods available on sObject are also available on AggregateResult. This also applies to Lists and Maps; those same methods can be called on collections of aggregate results. Here is another example using keySet()

Set<String> emails = new Map<String,SObject>(
    [SELECT Email Id FROM Contact GROUP BY Email]).keySet();

This not well-known trick can solve the problem in one line

 Set<String> accountNames = new Map<String,SObject>(
    [SELECT Name Id FROM Account GROUP BY Name]).keySet();

From the Matt and Neil blog post (of bigassforce fame)

How it works

The first thing to note is that we are using an Aggregate Query and not just ordinary SOQL. We combine special behaviours available to aggregate queries to build the set:

Step 1 - GROUP BY to collect unique values Exactly per the documentation, you can use the GROUP BY option in a SOQL query to avoid iterating through individual query results. That is, you specify a group of records instead of processing many individual records.

Step 2 - Alias the field as Id The default identifier for an aggregated field is expr0 but we will alias the result as Id. It doesn't matter what the field contains, so long as it's a string. We will exploit this special alias in the next step.

Step 3 - AggregateResult inherits from sObject The methods available on sObject are also available on AggregateResult. This also applies to Lists and Maps; those same methods can be called on collections of aggregate results. Here is another example using keySet()

Set<String> emails = new Map<String,SObject>(
    [SELECT Email Id FROM Contact GROUP BY Email]).keySet();

This not well-known trick can solve the problem in one line

 Set<String> accountNames = new Map<String,SObject>(
    [SELECT Name Id FROM Account GROUP BY Name]).keySet();

From the Matt and Neil blog post (of bigassforce fame)

How it works

The first thing to note is that we are using an Aggregate Query and not just ordinary SOQL. We combine special behaviours available to aggregate queries to build the set:

Step 1 - GROUP BY to collect unique values Exactly per the documentation, you can use the GROUP BY option in a SOQL query to avoid iterating through individual query results. That is, you specify a group of records instead of processing many individual records.

Step 2 - Alias the field as Id The default identifier for an aggregated field is expr0 but we will alias the result as Id. It doesn't matter what the field contains, so long as it's a string. We will exploit this special alias in the next step.

Step 3 - AggregateResult inherits from sObject The methods available on sObject are also available on AggregateResult. This also applies to Lists and Maps; those same methods can be called on collections of aggregate results. Here is another example using keySet()

Set<String> emails = new Map<String,SObject>(
    [SELECT Email Id FROM Contact GROUP BY Email]).keySet();