Class SafeSql

java.lang.Object
com.google.mu.safesql.SafeSql
@ThreadSafe @CheckReturnValue public final class SafeSql extends Object
An injection-safe dynamic SQL, constructed using compile-time enforced templates.

This class is intended to work with JDBC Connection API with parameters set through the setObject() method. The main use case though, is to be able to compose subqueries and leaf-level parameters with an intuitive templating API.

The syntax to create a SQL with JDBC parameters, and potentially with dynamic SQL arguments (such as column names) is as simple and intuitive as the following example: 


 List<String> groupColumns = ...;
 SafeSql sql = SafeSql.of(
     """
     SELECT `{group_columns}`, SUM(revenue) AS revenue
     FROM Sales
     WHERE sku = {sku}
     GROUP BY `{group_columns}`
     """,
     groupColumns, sku, groupColumns);
 List<RevenueRecord> results = sql.query(dataSource, RevenueRecord.class);

By default, all placeholder values are passed as JDBC parameters, unless quoted by backticks (used by databases like BigQuery, Databricks) or double quotes (used by databases like Oracle, Microsoft SQL Server or PostgreSQL). These are validated and interpreted as identifiers.

In the above example, placeholder sku will be passed as JDBC parameter, whereas the backtick-quoted groupColumns string list will be validated and then used as identifiers.

Except the placeholders, everything outside the curly braces are strictly WYSIWYG (what you see is what you get), so you can copy paste them between the Java code and your SQL console for quick testing and debugging.

The IN Operator
A common dynamic SQL use case is to use the IN SQL operator: 

   SafeSql sql = SafeSql.of(
       """
       SELECT id FROM Users
       WHERE firstName = {first_name} AND lastName IN ({last_names})
       """,
       firstName, lastNamesList);
   List<Long> ids = sql.query(dataSource, Long.class);
In the above example if firstName is "Emma" and lastNamesList is ["Watson", "Lin"], the generated SQL will be: 

   SELECT id FROM Employees
   WHERE firstName = ? AND lastName IN (?, ?)
And the parameters will be set as: 

   statement.setObject(1, "Emma");
   statement.setObject(2, "Watson");
   statement.setObject(3, "Lin");
Compile-time Protection

The templating engine uses compile-time checks to guard against accidental use of untrusted strings in the SQL, ensuring that they can only be sent as parameters of PreparedStatement: try to use a dynamically generated String as the SQL template and you'll get a compilation error.

In addition, the same set of compile-time guardrails from the StringFormat class are in effect to make sure that you don't pass lastName in the place of first_name, for example.

To enable the compile-time plugin, copy the <annotationProcessorPaths> in the "maven-compiler-plugin" section from the following pom.xml file snippet: 


 <build>
   <pluginManagement>
     <plugins>
       <plugin>
         <artifactId>maven-compiler-plugin</artifactId>
         <configuration>
           <annotationProcessorPaths>
             <path>
               <groupId>com.google.errorprone</groupId>
               <artifactId>error_prone_core</artifactId>
               <version>2.23.0</version>
             </path>
             <path>
               <groupId>com.google.mug</groupId>
               <artifactId>mug-errorprone</artifactId>
               <version>9.2</version>
             </path>
           </annotationProcessorPaths>
         </configuration>
       </plugin>
     </plugins>
   </pluginManagement>
 </build>
Conditional Subqueries
SafeSql's template syntax is designed to avoid control flows that could obfuscate SQL. Instead, complex control flow such as if-else, nested if, loops etc. should be performed in Java and passed in as subqueries.

Starting from v8.4, simple conditional subqueries (e.g. selecting a column if a flag is enabled) can use the guard operator -> inside template placeholders: 


   SafeSql sql = SafeSql.of(
       "SELECT {shows_email -> email,} name FROM Users", showsEmail());
The query text after the -> operator is the conditional subquery that's only included if showEmail() returns true. The subquery can include arbitrary characters except curly braces, so you can also have multi-line conditional subqueries.

The -> guard operator can also be used for Optional parameters such that the right-hand-side SQL will only render if the optional value is present.

For example, the following code builds SQL to query the Users table with flexible number of columns and a flexible WHERE clause depending on the UserCriteria object's state: 


   class UserCriteria {
     Optional<String> userId();
     Optional<String> firstName();
     List<String> aliases();
     ...
   }

   SafeSql usersQuery(UserCriteria criteria, @CompileTimeConstant String... columns) {
     return SafeSql.of(
         """
         SELECT `{columns}`
         FROM Users
         WHERE 1 = 1
             {user_id? -> AND id = user_id?}
             {first_name? -> AND firstName LIKE '%first_name?%'}
             {aliases? -> AND name IN (aliases?)}
         """,
         asList(columns),
         criteria.userId()),
         criteria.firstName(),
         criteria.aliases());
   }

   List<User> users = usersQuery(userCriteria, "email", "lastName")
       .query(dataSource, User.class);

The special "{foo? -> ...}" guard syntax informs the template engine that the right hand side query snippet is only rendered if the Optional parameter corresponding to the "foo?" placeholder is present, or the Collection paameter corresponding to it isn't empty, in which case the value of the Optional or Collection will be used in the right hand side snippet as if it were a regular template argument.

If UserCriteria has specified firstName() but userId() is unspecified (empty), the resulting SQL will look like: 


   SELECT `email`, `lastName` FROM Users WHERE firstName LIKE ?

And when you call usersQuery.prepareStatement(connection) or one of the similar convenience methods, statement.setObject(1, "%" + criteria.firstName().get() + "%") will be called to populate the PreparedStatement.

Complex Dynamic Subqueries
By composing SafeSql objects that encapsulate subqueries, you can parameterize by arbitrary sub-queries that are computed dynamically.

Imagine if you need to translate a user-facing structured search expression like location:US AND name:jing OR status:active into SQL. And you already have the search expression parser that turns the search expression into an AST (abstract syntax tree). The following code uses SafeSql template to turn it into SQL where clause that can be used to query the database for the results: 



 // The AST
 interface Expression permits AndExpression, OrExpression, HasExpression {}

 record AndExpression(Expression left, Expression right) implements Expression {}
 record OrExpression(Expression left, Expression right) implements Expression {}
 record HasExpression(String field, String text) implements Expression {}

 // AST -> SafeSql
 SafeSql toSqlFilter(Expression expression) {
   return switch (expression) {
     case HasExpression(String field, String text) ->
         SafeSql.of("`{field}` LIKE '%{text}%'", field, text);
     case AndExpression(Expression left, Expression right) ->
         SafeSql.of("({left}) AND ({right})", toSqlFilter(left), toSqlFilter(right));
     case OrExpression(Expression left, Expression right) ->
         SafeSql.of("({left}) OR ({right})", toSqlFilter(left), toSqlFilter(right));
   };
 }

 SafeSql query = SafeSql.of("SELECT * FROM Foos WHERE {filter}", toSqlFilter(expression));
Parameterize by Column Names or Table Names
Sometimes you may wish to parameterize by table names, column names etc. for which JDBC parameterization has no support.

If the identifiers are compile-time string literals, you can wrap them using SafeSql.of(COLUMN_NAME), which can then be composed as subqueries. But what if the identifier string is loaded from a resource file, or is specified by a request field?

Passing the string directly as a template parameter will only generate the JDBC '?' parameter in its place, which won't work (PreparedStatement can't parameterize identifiers); SafeSql.of(theString) will fail to compile because such strings are inherently dynamic and untrusted.

The safe way to parameterize dynamic strings as identifiers is to backtick-quote their placeholders in the SQL template (if you use Oracle, PostgreSQL that use double quotes for identifier, use double quotes instead). For example: 


   SafeSql.of("SELECT `{columns}` FROM Users", request.getColumns())
The backticks tell SafeSql that the string is supposed to be an identifier (or a list of identifiers). SafeSql will sanity-check the string(s) to ensure injection safety.

In the above example, if getColumns() returns ["id", "age"], the genereated SQL will be: 


   SELECT `id`, `age` FROM Users

That is, each individual string will be backtick-quoted and then joined by ", ".

The LIKE Operator

Note that with straight JDBC API, if you try to use the LIKE operator to match a user-provided substring, i.e. using LIKE '%foo%' to search for "foo", this seemingly intuitive syntax is actually incorect: 


   String searchTerm = ...;
   PreparedStatement statement =
       connection.prepareStatement("SELECT id FROM Users WHERE firstName LIKE '%?%'");
   statement.setString(1, searchTerm);
JDBC PreparedStatement considers the quoted question mark as a literal so the setString() call will fail. You'll need to use the following workaround: 

   PreparedStatement statement =
       connection.prepareStatement("SELECT id FROM Users WHERE firstName LIKE ?");
   statement.setString(1, "%" + searchTerm + "%");
And even then, if the searchTerm includes special characters like '%' or backslash ('\'), they'll be interepreted as wildcards and escape characters, opening it up to a form of minor SQL injection despite already using the parameterized SQL.

The SafeSql template protects you from this caveat. The most intuitive syntax does exactly what you'd expect (and it escapes special characters too): 


   String searchTerm = ...;
   SafeSql sql = SafeSql.of(
       "SELECT id FROM Users WHERE firstName LIKE '%{search_term}%'", searchTerm);
   List<Long> ids = sql.query(dataSource, Long.class);

Automatic Escaping: No Need for ESCAPE Clause

This means you do not need to (and in fact, must not) write SQL using ESCAPE clauses after LIKE '%{foo}%'. Any such attempt, as in: 


   SELECT name FROM Users WHERE name LIKE '%{term}%' ESCAPE '\'

...will be rejected, because SafeSql already performs all necessary escaping internally and automatically uses ESCAPE '^'. In other words, LIKE just works.

This eliminates the need for developers to deal with brittle double-escaping (like '\\'), or any cross-dialect compatibility issues. The template is also more readable.

If you find yourself wanting to use ESCAPE, consider whether you are manually escaping strings that could instead be safely passed as-is to SafeSql's template system.

That said, this only applies when template placeholder is used in the LIKE string. You can use any valid SQL ESCAPE syntax if placeholder isn't used in the LIKE expression.

Quote String Placeholders
Even when you don't use the LIKE operator or the percent sign (%), it may still be more readable to quote the string placeholders just so the SQL template explicitly tells readers that the parameter is a string. The following template works with or without the quotes around the {id} placeholder: 

   // Reads more clearly that the {id} is a string
   SafeSql sql = SafeSql.of("SELECT * FROM Users WHERE id = '{id}'", userId);
Enforce Identical Parameter

The compile-time check tries to be helpful and checks that if you use the same parameter name more than once in the template, the same value must be used for it.

So for example, if you are trying to generate a SQL that looks like: 


   SELECT u.firstName, p.profileId
   FROM (SELECT firstName FROM Users WHERE id = 'foo') u,
        (SELECT profileId FROM Profiles WHERE userId = 'foo') p
It'll be important to use the same user id for both subqueries. And you can use the following template to make sure of it at compile time: 

   SafeSql sql = SafeSql.of(
       """
       SELECT u.firstName, p.profileId
       FROM (SELECT firstName FROM Users WHERE id = '{user_id}') u,
            (SELECT profileId FROM Profiles WHERE userId = '{user_id}') p
       """,
       userId, userId);
If someone mistakenly passes in inconsistent ids, they'll get a compilation error.

For Spring users: in order to participate in Spring declarative transaction (methods annotated with @Transactional), you need to call one of the methods that accept a Connection, such as the update(Connection) method. In a nutshell, it takes calling DataSourceUtils.getConnection(dataSource) to get the connection in the current transaction, then passing it to update(connection).

Note that you will also need to catch SQLException and turn it into a Spring DataAccessException. For that you need Spring's SQLExceptionTranslator, which has some quirks to use.

At this point, it may be easier to create a small wrapper class to execute SafeSql from within a Spring transaction. And while you are there, might as well make it safer to also support calling from outside of a transaction, by using try-with-resources to close the connection: 


 // Use Java 16 record for brevity. You can use a regular class too.
 @Component
 public record SafeSqlBridge(DataSource dataSource, SQLExceptionTranslator translator) {
   public int executeUpdate(SafeSql sql) {
     try {
       if (TransactionSynchronizationManager.isActualTransactionActive()) {
         // in an active transaction, don't close or release the connection.
         return sql.update(DataSourceUtils.getConnection(dataSource()));
       } else {
         // not in active transaction, should close the connection.
         try (Connection connection = dataSource().getConnection()) {
           return sql.update(connection);
         }
       }
     } catch (SQLException e) {
       DataAccessException dae =
           translator().translate("executeUpdate(SafeSql)", sql.debugString(), e);
       throw dae == null ? throw new UncheckedSqlException(e) : dae;
     }
   }
 }
You can then dependency-inject SafeSqlBridge to execute SafeSql queries: 

 // Use Java 16 record for brevity. You can use a regular class too.
 @Service
 record MyService(SafeSqlBridge bridge) {
   @Transactional void transferCredit(String fromAccount, String toAccount) {
     SafeSql sql = SafeSql.of("INSERT INTO(...)...'{from}'...'{to}'", fromAccount(), toAccount());
     bridge().executeUpdate(sql);
   }
 }

Immutable if the template parameters you pass to it are immutable.

Starting from v9.0, SafeSql is moved to the mug-safesql artifact, and no longer requires Guava as a dependency.

Since:
8.2

  • Field Details

    • EMPTY

      public static final SafeSql EMPTY
      An empty SQL

  • Method Details

    • of

      @TemplateFormatMethod public static SafeSql of(@TemplateString @CompileTimeConstant String template, Object... params)
      Returns SafeSql using template and params.

      For example: 

      
 List<Long> jobIds = SafeSql.of(
         "SELECT id FROM Jobs WHERE timestamp BETWEEN {start} AND {end}",
         startTime, endTime)
     .query(dataSource, Long.class);

      Note that if you plan to create a PreparedStatement and use it multiple times with different sets of parameters, it's more efficient to use prepareToQuery() or prepareToUpdate(), which will reuse the same PreparedStatement for multiple calls. The returned StringFormat.Templates are protected at compile-time against incorrect varargs.

      Parameters:
      template - the sql template
      params - The template parameters. Parameters that are themselves SafeSql are considered trusted subqueries and are appended directly. Other types are passed through JDBC PreparedStatement.setObject(int, java.lang.Object, int), with one exception: when the corresponding placeholder is quoted by backticks like `{columns}`, its string parameter value (or Iterable<String> parameter value) are directly appended (quotes, backticks, backslash and other special characters are disallowed). This makes it easy to parameterize by table names, column names etc.

    • nonNegativeLiteral

      @Deprecated public static SafeSql nonNegativeLiteral(int number)
      Deprecated.
      Prefer OFFSET-FETCH clause, which is parameterizable
      Wraps non-negative number as a literal SQL snippet in a SafeSql object.

      For example, the following SQL Server query allows parameterization by the TOP n number: 

      
   SafeSql.of("SELECT TOP {page_size} UserId FROM Users", nonNegativeLiteral(pageSize))

      Needed because the SQL Server JDBC driver doesn't support parameterizing the TOP number through PreparedStatement API.

      Although, you should generally use the OFFSET-FETCH syntax, which supports parameterization.

      Throws:
      IllegalArgumentException - if number is negative

    • optionally

      @TemplateFormatMethod @Deprecated public static SafeSql optionally(@TemplateString @CompileTimeConstant String query, Optional<?> param)
      Deprecated.
      Use SafeSql.of("{foo? -> OR foo?}", optionalFoo) instead of optionally("or {foo}", optionalFoo) because the former allows you to reference foo? multiple times in the right hand side snippet.
      An optional query that's only rendered if param is present; otherwise returns EMPTY. It's for use cases where a subquery is only added when present, for example the following query will add the WHERE clause if the filter is present: 
      
 SafeSql query = SafeSql.of(
     "SELECT * FROM jobs {where}",
     SafeSql.optionally("WHERE {filter}", getOptionalWhereClause()));

    • when

      @TemplateFormatMethod public static SafeSql when(boolean condition, @TemplateString @CompileTimeConstant String template, Object... params)
      An optional query that's only rendered if condition is true; otherwise returns EMPTY. It's for use cases where a subquery is only conditionally added, for example the following query will only include the userEmail column under super user mode: 
      
 SafeSql query = SafeSql.of(
     "SELECT job_id, start_timestamp {user_email} FROM jobs",
     SafeSql.when(isSuperUser, ", user_email"));
      Parameters:
      condition - the guard condition to determine if template should be renderd
      template - the template to render if condition is true
      params - see of(String, Object...) for discussion on the template arguments

    • when

      public SafeSql when(boolean condition)
      Returns this SafeSql if condition is true; otherwise returns EMPTY.
      Since:
      8.4

    • template

      public static StringFormat.Template<SafeSql> template(@CompileTimeConstant String template)
      Returns a StringFormat.Template of SafeSql based on the template string. Useful for creating a constant to be reused with different parameters.

      For example: 

      
 private static final Template<SafeSql> FIND_USERS_BY_NAME =
     SafeSql.template("SELECT `{columns}` FROM Users WHERE name LIKE '%{name}%'");

 String searchBy = ...;
 List<User> userIds = FIND_USERS_BY_NAME.with(asList("id", "name"), searchBy)
     .query(dataSource, User.class);

      If you don't need a reusable template, consider using of(java.lang.String, java.lang.Object...) instead, which is simpler.

      The template arguments follow the same rules as discussed in of(String, Object...) and receives the same compile-time protection against mismatch or out-of-order human mistakes, so it's safe to use the template as a constant.

      The returned template is immutable and thread safe.

    • and

      public static Collector<SafeSql,?,SafeSql> and()
      A collector that joins boolean query snippet using AND operator. The AND'ed sub-queries will be enclosed in pairs of parenthesis to avoid ambiguity. If the input is empty, the result will be "(1 = 1)".

      Empty SafeSql elements are ignored and not joined.

    • or

      public static Collector<SafeSql,?,SafeSql> or()
      A collector that joins boolean query snippet using OR operator. The OR'ed sub-queries will be enclosed in pairs of parenthesis to avoid ambiguity. If the input is empty, the result will be "(1 = 0)".

      Empty SafeSql elements are ignored and not joined.

    • joining

      public static Collector<SafeSql,?,SafeSql> joining(@CompileTimeConstant String delimiter)
      Returns a collector that joins SafeSql elements using delimiter.

      Empty SafeSql elements are ignored and not joined.

    • orElse

      @TemplateFormatMethod public SafeSql orElse(@TemplateString @CompileTimeConstant String fallback, Object... params)
      If this query is empty (likely from a call to optionally(java.lang.String, java.util.Optional<?>) or when(boolean, java.lang.String, java.lang.Object...)), returns the SafeSql produced from the fallback template and args.

      Using this method, you can create a chain of optional queries like: 

      
 SafeSql.of(
     """
     CREATE TABLE ...
     {cluster_by}
     """,
     when(enableCluster, "CLUSTER BY (`{cluster_columns}`)", clusterColumns)
         .orElse("-- no cluster"));
      Since:
      8.3

    • orElse

      public SafeSql orElse(SafeSql fallback)
      If this query is empty (likely from a call to optionally(java.lang.String, java.util.Optional<?>) or when(boolean, java.lang.String, java.lang.Object...)), returns the fallback query.
      Since:
      8.3

    • orElse

      public SafeSql orElse(Supplier<SafeSql> fallback)
      If this query is empty (likely from a call to optionally(java.lang.String, java.util.Optional<?>) or when(boolean, java.lang.String, java.lang.Object...)), returns the query produced by the fallback supplier.
      Since:
      8.3

    • query

      public <T> List<T> query(DataSource dataSource, Class<? extends T> resultType)
      Executes the encapsulated SQL as a query against dataSource.getConnection(). The ResultSet will be consumed, transformed to a list of T and then closed before returning.

      For example: 

      
 List<User> users = SafeSql.of("SELECT id, name FROM Users WHERE name LIKE '%{name}%'", name)
     .query(dataSource, User.class);

 record User(long id, String name) {...}

      The class of resultType must define a non-private constructor that accepts the same number of parameters as returned by the query. The parameter order doesn't matter but the parameter names and types must match.

      Note that if you've enabled the -parameters javac flag, the above example code will just work. If you can't enable -parameters, consider explicitly annotating the constructor parameters as in: 

      
 record User(@SqlName("id") long id, @SqlName("name") String name) {...}

      Alternatively, if your query only selects one column, you could also use this method to read the results: 

      
 List<String> names = SafeSql.of("SELECT name FROM Users WHERE name LIKE '%{name}%'", name)
     .query(dataSource, String.class);

      You can also map the result rows to Java Beans, for example: 

      
 List<UserBean> users =
     SafeSql.of("SELECT id, name FROM Users WHERE name LIKE '%{name}%'", name)
         .query(dataSource, UserBean.class);

 public class UserBean {
   public void setId(long id) {...}
   public void setName(String name) {...}
 }

      The rules of mapping query columns to Java Bean properties are:

      • Case doesn't matter. job_id will match jobId or JOB_ID.
      • The column names can be a superset, or a subset of the bean property names. This allows you to use "select *" in the query (when performance isn't a concern), or use a generic Java bean that may have more properties than some individual queries.
      • If a bean property is of primitive type, and the corresponding query column value is null, the property will be left as is.
      • If you can't make a bean property match a query column, consider annotating the setter method with the @SqlName annotation to customize the column name.
      • Exception will be thrown if a column doesn't map to a settable property, and the columns aren't a superset. For example, you may have renamed a property but forgot to rename the corresponding query column. In such case, failing loudly and clearly is safer than letting the program silently run with corrupted state.
      Throws:
      UncheckedSqlException - wraps SQLException if failed
      Since:
      9.0

    • query

      public <T> List<T> query(Connection connection, Class<? extends T> resultType) throws SQLException
      Similar to query(DataSource, Class), but uses an existing connection.

      It's usually more convenient to use the DataSource overload because you won't have to manage the JDBC resources. Use this method if you need to reuse a connection (mostly for multi-statement transactions).

      Throws:
      SQLException
      Since:
      8.7

    • query

      public <T> List<T> query(DataSource dataSource, SqlFunction<? super ResultSet, ? extends T> rowMapper)
      Executes the encapsulated SQL as a query against dataSource.getConnection(). The ResultSet will be consumed, transformed by rowMapper and then closed before returning.

      For example: 

      
 List<Long> ids = SafeSql.of("SELECT id FROM Users WHERE name LIKE '%{name}%'", name)
     .query(dataSource, row -> row.getLong("id"));

      Internally it delegates to PreparedStatement.executeQuery() or Statement.executeQuery(java.lang.String) if this sql contains no JDBC binding parameters.

      Throws:
      UncheckedSqlException - wraps SQLException if failed
      Since:
      9.0

    • query

      public <T> List<T> query(DataSource dataSource, StatementSettings settings, Class<? extends T> resultType)
      Similar to query(DataSource, Class), but with settings (can be set via lambda like stmt -> stmt.setMaxRows(100)) to allow customization.
      Throws:
      UncheckedSqlException - wraps SQLException if failed
      Since:
      9.0

    • query

      public <T> List<T> query(Connection connection, StatementSettings settings, Class<? extends T> resultType) throws SQLException
      Similar to query(DataSource, StatementSettings, Class), but uses an existing connection.

      It's usually more convenient to use the DataSource overload because you won't have to manage the JDBC resources. Use this method if you need to reuse a connection (mostly for multi-statement transactions).

      Throws:
      SQLException
      Since:
      9.0

    • query

      public <T> List<T> query(DataSource dataSource, StatementSettings settings, SqlFunction<? super ResultSet, ? extends T> rowMapper)
      Executes the encapsulated SQL as a query against dataSource.getConnection(), using settings (can be set via lambda like stmt -> stmt.setMaxRows(100)). The ResultSet will be consumed, transformed by rowMapper and then closed before returning.

      For example: 

      
 List<Long> ids = SafeSql.of("SELECT id FROM Users WHERE name LIKE '%{name}%'", name)
     .query(dataSource, stmt -> stmt.setMaxRows(100000), row -> row.getLong("id"));

      Internally it delegates to PreparedStatement.executeQuery() or Statement.executeQuery(java.lang.String) if this sql contains no JDBC binding parameters.

      Throws:
      UncheckedSqlException - wraps SQLException if failed
      Since:
      9.0

    • query

      public <T> List<T> query(Connection connection, StatementSettings settings, SqlFunction<? super ResultSet, ? extends T> rowMapper) throws SQLException
      Similar to query(DataSource, StatementSettings, SqlFunction), but uses an existing connection.

      It's usually more convenient to use the DataSource overload because you won't have to manage the JDBC resources. Use this method if you need to reuse a connection (mostly for multi-statement transactions).

      For example: 

      
 List<Long> ids = SafeSql.of("SELECT id FROM Users WHERE name LIKE '%{name}%'", name)
     .query(connection, stmt -> stmt.setMaxRows(100000), row -> row.getLong("id"));

      Internally it delegates to PreparedStatement.executeQuery() or Statement.executeQuery(java.lang.String) if this sql contains no JDBC binding parameters.

      Throws:
      SQLException
      Since:
      9.0

    • query

      public <T> List<T> query(Connection connection, SqlFunction<? super ResultSet, ? extends T> rowMapper) throws SQLException
      Similar to query(DataSource, SqlFunction), but uses an existing connection.

      It's usually more convenient to use the DataSource overload because you won't have to manage the JDBC resources. Use this method if you need to reuse a connection (mostly for multi-statement transactions).

      For example: 

      
 List<Long> ids = SafeSql.of("SELECT id FROM Users WHERE name LIKE '%{name}%'", name)
     .query(connection, row -> row.getLong("id"));

      Internally it delegates to PreparedStatement.executeQuery() or Statement.executeQuery(java.lang.String) if this sql contains no JDBC binding parameters.

      Throws:
      SQLException

    • queryForOne

      public <T> Optional<T> queryForOne(DataSource dataSource, Class<? extends T> resultType)
      Similar to query(DataSource, Class), but only fetches one row if the query result includes at least one rows, or else returns Optional.empty().

      Suitable for queries that search by the primary key, for example: 

      
 Optional<User> user =
     SafeSql.of("select id, name from Users where id = {id}", userId)
         .queryForOne(dataSource, User.class);
      Throws:
      UncheckedSqlException - wraps SQLException if failed
      NullPointerException - if the first column value is null, and resultType isn't a record or Java Bean.
      Since:
      9.2

    • queryForOne

      public <T> Optional<T> queryForOne(Connection connection, Class<? extends T> resultType) throws SQLException
      Similar to query(Connection, Class), but only fetches one row if the query result includes at least one rows, or else returns Optional.empty().

      Suitable for queries that search by the primary key, for example: 

      
 Optional<User> user =
     SafeSql.of("select id, name from Users where id = {id}", userId)
         .queryForOne(connection, User.class);
      Throws:
      NullPointerException - if the first column value is null, and resultType isn't a record or Java Bean.
      SQLException
      Since:
      9.2

    • queryForOne

      public <T> Optional<T> queryForOne(DataSource dataSource, SqlFunction<? super ResultSet, ? extends T> rowMapper)
      Similar to query(DataSource, SqlFunction), but only fetches one row if the query result includes at least one rows, or else returns Optional.empty().

      Suitable for queries that search by the primary key.

      Throws:
      UncheckedSqlException - wraps SQLException if failed
      NullPointerException - if rowMapper returns null
      Since:
      9.2

    • queryForOne

      public <T> Optional<T> queryForOne(Connection connection, SqlFunction<? super ResultSet, ? extends T> rowMapper) throws SQLException
      Similar to query(Connection, SqlFunction), but only fetches one row if the query result includes at least one rows, or else returns Optional.empty().

      Suitable for queries that search by the primary key.

      Throws:
      NullPointerException - if rowMapper returns null
      SQLException
      Since:
      9.2

    • queryLazily

      @MustBeClosed public <T> Stream<T> queryLazily(DataSource dataSource, Class<? extends T> resultType)
      Executes the encapsulated SQL as a query against connection, and then fetches the results lazily in a stream.

      Each result row is transformed into resultType.

      The caller must close it using try-with-resources idiom, which will close the associated Statement and ResultSet.

      For example: 

      
 SafeSql sql = SafeSql.of("SELECT id, name FROM Users WHERE name LIKE '%{name}%'", name);
 try (Stream<User> users = sql.queryLazily(connection, User.class)) {
   return users.findFirst();
 }

 record User(long id, String name) {...}

      The class of resultType must define a non-private constructor that accepts the same number of parameters as returned by the query. The parameter order doesn't matter but the parameter names and types must match.

      Note that if you've enabled the -parameters javac flag, the above example code will just work. If you can't enable -parameters, consider explicitly annotating the constructor parameters as in: 

      
 record User(@SqlName("id") long id, @SqlName("name") String name) {...}

      Alternatively, if your query only selects one column, you could also use this method to read the results: 

      
 SafeSql sql = SafeSql.of("SELECT id FROM Users WHERE name LIKE '%{name}%'", name);
 try (Stream<Long> ids = sql.queryLazily(dataSource, Long.class)) {
   return ids.findFirst();
 }

      You can also map the result rows to Java Beans, similar to query(Connection, Class).

      Throws:
      UncheckedSqlException - wraps SQLException if failed
      Since:
      9.2

    • queryLazily

      @MustBeClosed public <T> Stream<T> queryLazily(DataSource dataSource, SqlFunction<? super ResultSet, ? extends T> rowMapper)
      Executes the encapsulated SQL as a query against connection, and then fetches the results lazily in a stream.

      The returned Stream includes results transformed by rowMapper. The caller must close it using try-with-resources idiom, which will close the associated Connection, Statement and ResultSet.

      For example: 

      
 SafeSql sql = SafeSql.of("SELECT name FROM Users WHERE name LIKE '%{name}%'", name);
 try (Stream<String> names = sql.queryLazily(dataSource, row -> row.getString("name"))) {
   return names.findFirst();
 }

      Internally it delegates to PreparedStatement.executeQuery() or Statement.executeQuery(java.lang.String) if this sql contains no JDBC binding parameters.

      Throws:
      UncheckedSqlException - wraps SQLException if failed
      Since:
      9.2

    • queryLazily

      @MustBeClosed public <T> Stream<T> queryLazily(DataSource dataSource, StatementSettings settings, Class<? extends T> resultType)
      Executes the encapsulated SQL as a query against connection, with settings (can be set via lambda like stmt -> stmt.setFetchSize(100)), and then fetches the results lazily in a stream.

      Each result row is transformed into resultType.

      For example: 

      
 SafeSql sql = SafeSql.of("SELECT name FROM Users WHERE name LIKE '%{name}%'", name);
 try (Stream<String> names = sql.queryLazily(
     dataSource, stmt -> stmt.setFetchSize(100), String.class)) {
   return names.findFirst();
 }

      Internally it delegates to PreparedStatement.executeQuery() or Statement.executeQuery(java.lang.String) if this sql contains no JDBC binding parameters.

      Throws:
      UncheckedSqlException - wraps SQLException if failed
      Since:
      9.2

    • queryLazily

      @MustBeClosed public <T> Stream<T> queryLazily(DataSource dataSource, StatementSettings settings, SqlFunction<? super ResultSet, ? extends T> rowMapper)
      Executes the encapsulated SQL as a query against dataSource, with settings (can be set via lambda like stmt -> stmt.setFetchSize(100), and then fetches the results lazily in a stream.

      The returned Stream includes results transformed by rowMapper. The caller must close it using try-with-resources idiom, which will close the associated Connection, Statement and ResultSet.

      For example: 

      
 SafeSql sql = SafeSql.of("SELECT name FROM Users WHERE name LIKE '%{name}%'", name);
 try (Stream<String> names = sql.queryLazily(
     dataSource, stmt -> stmt.setFetchSize(100), row -> row.getString("name"))) {
   return names.findFirst();
 }

      Internally it delegates to PreparedStatement.executeQuery() or Statement.executeQuery(java.lang.String) if this sql contains no JDBC binding parameters.

      Throws:
      UncheckedSqlException - wraps SQLException if failed
      Since:
      9.2

    • queryLazily

      @MustBeClosed public <T> Stream<T> queryLazily(Connection connection, Class<? extends T> resultType) throws SQLException
      Executes the encapsulated SQL as a query against connection, and then fetches the results lazily in a stream.

      Each result row is transformed into resultType.

      The caller must close it using try-with-resources idiom, which will close the associated Statement and ResultSet.

      For example: 

      
 SafeSql sql = SafeSql.of("SELECT id, name FROM Users WHERE name LIKE '%{name}%'", name);
 try (Stream<User> users = sql.queryLazily(connection, User.class)) {
   return users.findFirst();
 }

 record User(long id, String name) {...}

      The class of resultType must define a non-private constructor that accepts the same number of parameters as returned by the query. The parameter order doesn't matter but the parameter names and types must match.

      Note that if you've enabled the -parameters javac flag, the above example code will just work. If you can't enable -parameters, consider explicitly annotating the constructor parameters as in: 

      
 record User(@SqlName("id") long id, @SqlName("name") String name) {...}

      Alternatively, if your query only selects one column, you could also use this method to read the results: 

      
 SafeSql sql = SafeSql.of("SELECT id FROM Users WHERE name LIKE '%{name}%'", name);
 try (Stream<Long> ids = sql.queryLazily(connection, Long.class)) {
   return ids.findFirst();
 }

      You can also map the result rows to Java Beans, similar to query(Connection, Class).

      Throws:
      SQLException
      Since:
      8.7

    • queryLazily

      @MustBeClosed public <T> Stream<T> queryLazily(Connection connection, SqlFunction<? super ResultSet, ? extends T> rowMapper) throws SQLException
      Executes the encapsulated SQL as a query against connection, and then fetches the results lazily in a stream.

      The returned Stream includes results transformed by rowMapper. The caller must close it using try-with-resources idiom, which will close the associated Statement and ResultSet.

      For example: 

      
 SafeSql sql = SafeSql.of("SELECT name FROM Users WHERE name LIKE '%{name}%'", name);
 try (Stream<String> names = sql.queryLazily(connection, row -> row.getString("name"))) {
   return names.findFirst();
 }

      Internally it delegates to PreparedStatement.executeQuery() or Statement.executeQuery(java.lang.String) if this sql contains no JDBC binding parameters.

      Throws:
      SQLException
      Since:
      8.4

    • queryLazily

      @MustBeClosed public <T> Stream<T> queryLazily(Connection connection, StatementSettings settings, Class<? extends T> resultType) throws SQLException
      Executes the encapsulated SQL as a query against connection, with settings (can be set via lambda like stmt -> stmt.setFetchSize(100)), and then fetches the results lazily in a stream.

      Each result row is transformed into resultType.

      For example: 

      
 SafeSql sql = SafeSql.of("SELECT name FROM Users WHERE name LIKE '%{name}%'", name);
 try (Stream<String> names = sql.queryLazily(
     connection, stmt -> stmt.setFetchSize(100), String.class)) {
   return names.findFirst();
 }

      Internally it delegates to PreparedStatement.executeQuery() or Statement.executeQuery(java.lang.String) if this sql contains no JDBC binding parameters.

      Throws:
      SQLException
      Since:
      9.0

    • queryLazily

      @MustBeClosed public <T> Stream<T> queryLazily(Connection connection, StatementSettings settings, SqlFunction<? super ResultSet, ? extends T> rowMapper) throws SQLException
      Executes the encapsulated SQL as a query against connection, with settings (can be set via lambda like stmt -> stmt.setFetchSize(100), and then fetches the results lazily in a stream.

      The returned Stream includes results transformed by rowMapper. The caller must close it using try-with-resources idiom, which will close the associated Statement and ResultSet.

      For example: 

      
 SafeSql sql = SafeSql.of("SELECT name FROM Users WHERE name LIKE '%{name}%'", name);
 try (Stream<String> names = sql.queryLazily(
     connection, stmt -> stmt.setFetchSize(100), row -> row.getString("name"))) {
   return names.findFirst();
 }

      Internally it delegates to PreparedStatement.executeQuery() or Statement.executeQuery(java.lang.String) if this sql contains no JDBC binding parameters.

      Throws:
      SQLException
      Since:
      9.0

    • update

      @CanIgnoreReturnValue public int update(DataSource dataSource)
      Executes the encapsulated DML (create, update, delete statements) against dataSource and returns the number of affected rows.

      For example: 

      
 SafeSql.of("INSERT INTO Users(id, name) VALUES({id}, '{name}')", id, name)
     .update(dataSource);

      Internally it delegates to PreparedStatement.executeUpdate().

      Throws:
      UncheckedSqlException - wraps SQLException if failed
      Since:
      9.2

    • update

      @CanIgnoreReturnValue public int update(DataSource dataSource, StatementSettings settings)
      Similar to update(DataSource), but with settings (can be set via lambda like stmt -> stmt.setQueryTimeout(100)) to allow customization.
      Throws:
      UncheckedSqlException - wraps SQLException if failed
      Since:
      9.2

    • update

      @CanIgnoreReturnValue public int update(Connection connection) throws SQLException
      Executes the encapsulated DML (create, update, delete statements) against connection and returns the number of affected rows.

      For example: 

      
 SafeSql.of("INSERT INTO Users(id, name) VALUES({id}, '{name}')", id, name)
     .update(connection);

      Internally it delegates to PreparedStatement.executeUpdate().

      Throws:
      SQLException

    • update

      @CanIgnoreReturnValue public int update(Connection connection, StatementSettings settings) throws SQLException
      Similar to update(Connection), but with settings (can be set via lambda like stmt -> stmt.setQueryTimeout(100)) to allow customization.
      Throws:
      SQLException
      Since:
      9.0

    • prepareStatement

      @MustBeClosed public PreparedStatement prepareStatement(Connection connection) throws SQLException
      Returns a PreparedStatement with the encapsulated sql and parameters.

      It's often more convenient to use query(javax.sql.DataSource, java.lang.Class<? extends T>) or update(javax.sql.DataSource) unless you need to directly operate on the PreparedStatement.

      Throws:
      SQLException

    • prepareToQuery

      public static <T> StringFormat.Template<List<T>> prepareToQuery(Connection connection, @CompileTimeConstant String template, Class<? extends T> resultType)
      Returns a query template that will reuse the same cached PreparedStatement for repeated calls of StringFormat.Template.with(java.lang.Object...) using different parameters.

      Allows callers to take advantage of the performance benefit of PreparedStatement without having to re-create the statement for each call. For example in: 

      
   try (var connection = ...) {
     var queryByName = SafeSql.prepareToQuery(
         connection, "SELECT id, name FROM Users WHERE name LIKE '%{name}%'",
         User.class);
     for (String name : names) {
       for (User user : queryByName.with(name))) {
         ...
       }
     }
   }

   record User(long id, String name) {...}
      Each time queryByName.with(name) is called, it executes the same query template against the connection, but with a different name parameter. Internally it reuses the cached PreparedStatement object and just calls PreparedStatement.setObject(int, Object) with the new set of parameters before calling PreparedStatement.executeQuery().

      The template arguments follow the same rules as discussed in of(String, Object...) and receives the same compile-time protection against mismatch or out-of-order human mistakes.

      The returned Template is not thread safe.

      The caller is expected to close the connection after done, which will close the cached PreparedStatement. The ResultSet objects are guaranteed to be closed after each use before the PreparedStatement is closed.

      Since:
      8.7

    • prepareToQuery

      public static <T> StringFormat.Template<List<T>> prepareToQuery(Connection connection, @CompileTimeConstant String template, SqlFunction<? super ResultSet, ? extends T> rowMapper)
      Returns a query template that will reuse the same cached PreparedStatement for repeated calls of StringFormat.Template.with(java.lang.Object...) using different parameters.

      Allows callers to take advantage of the performance benefit of PreparedStatement without having to re-create the statement for each call. For example in: 

      
   try (var connection = ...) {
     var queryByName = SafeSql.prepareToQuery(
         connection, "SELECT id FROM Users WHERE name LIKE '%{name}%'",
         row -> row.getLong("id"));
     for (String name : names) {
       for (long id : queryByName.with(name))) {
         ...
       }
     }
   }
      Each time queryByName.with(name) is called, it executes the same query template against the connection, but with a different name parameter. Internally it reuses the cached PreparedStatement object and just calls PreparedStatement.setObject(int, Object) with the new set of parameters before calling PreparedStatement.executeQuery().

      The template arguments follow the same rules as discussed in of(String, Object...) and receives the same compile-time protection against mismatch or out-of-order human mistakes.

      The returned Template is not thread safe.

      The caller is expected to close the connection after done, which will close the cached PreparedStatement. The ResultSet objects are guaranteed to be closed after each use before the PreparedStatement is closed.

    • prepareToUpdate

      public static StringFormat.Template<Integer> prepareToUpdate(Connection connection, @CompileTimeConstant String template)
      Returns a DML (create, update, delete) template that will reuse the same cached PreparedStatement for repeated calls of StringFormat.Template.with(java.lang.Object...) using different parameters.

      Allows callers to take advantage of the performance benefit of PreparedStatement without having to re-create the statement for each call. For example in: 

      
   try (var connection = ...) {
     var insertUser = SafeSql.prepareToUpdate(
         connection, "INSERT INTO Users(id, name) VALUES({id}, '{name}')");
     int totalRowsAffected = insertUser.with(1, "Tom") + insertUser.with(2, "Emma");
   }
      Each time insertUser.with(...) is called, it executes the same DML template against the connection, but with different id and name parameters. Internally it reuses the cached PreparedStatement object and just calls PreparedStatement.setObject(int, Object) with the new set of parameters before calling PreparedStatement.executeUpdate().

      The template arguments follow the same rules as discussed in of(String, Object...) and receives the same compile-time protection against mismatch or out-of-order human mistakes.

      The returned Template is not thread safe because the cached PreparedStatement objects aren't.

      The caller is expected to close the connection after done, which will close the cached PreparedStatement.

    • debugString

      public String debugString()
      Returns a query string with the parameter values embedded for easier debugging (logging, testing, golden file etc.). DO NOT use it as the production SQL query because embedding the parameter values isn't safe from SQL injection.
      Since:
      9.0

    • toString

      public String toString()
      Returns the SQL text with the template parameters translated to the JDBC '?' placeholders.
      Overrides:
      toString in class Object

    • hashCode

      public int hashCode()
      Overrides:
      hashCode in class Object

    • equals

      public boolean equals(Object obj)
      Overrides:
      equals in class Object