Get started

For the impatient, bellow is a complete example.

import java.util.List;

import java.sql.DriverManager;
import java.sql.Connection;
import java.sql.Date;

import acolyte.jdbc.ConnectionHandler;
import acolyte.jdbc.StatementHandler;
import acolyte.jdbc.CompositeHandler;
import acolyte.jdbc.RowList3;
import acolyte.jdbc.Result;

import acolyte.jdbc.StatementHandler.Parameter;

import static acolyte.jdbc.RowLists.rowList3;

// ...

// Configure in anyway JDBC with following url,
// declaring handler registered with 'my-unique-id' will be used.
final String jdbcUrl = "jdbc:acolyte:anything-you-want?handler=my-unique-id"

// Prepare handler
StatementHandler handler = new CompositeHandler().
  withQueryDetection("^SELECT "). // regex test from beginning
  withQueryDetection("EXEC that_proc"). // second detection regex
  withUpdateHandler(new CompositeHandler.UpdateHandler() {
    // Handle execution of update statement (not query)
    public UpdateResult apply(String sql, List<Parameter> parameters) {
      // ...
      return UpdateResult.Nothing;
  }).withQueryHandler(new CompositeHandler.QueryHandler () {
    public QueryResult apply(String sql, List<Parameter> parameters) {
      // ...

      // Prepare list of 2 rows
      // with 3 columns of types String, Float, Date
      RowList3<String, Float, Date> rows = 
        rowList3(String.class, Float.class, Date.class).
        withLabel(1, "String").withLabel(3, "Date"). // Optional: set labels
        append("str", 1.2f, new Date(1, 2, 3)). // values append
        append("val", 2.34f, null);

      return rows.asResult();

// Register prepared handler with expected ID 'my-unique-id'
acolyte.jdbc.Driver.register("my-unique-id", handler);

// then ...
Connection con = DriverManager.getConnection(jdbcUrl);

// ... Connection |con| is managed through |handler|

You can see more use cases whose expectations are visible in specifications.

Setup in your project

With Maven 2/3+, the dependency can be configured as bellow.

<!-- ... -->
    <!-- ... -->

The dependency will usually be set with the scope test.

Maven Central

It can be done similarly using SBT:

libraryDependencies ++= Seq(
  "" % "jdbc-driver" % "1.2.7" % Test


As soon as you register Acolyte handler with a unique ID, corresponding connection can be resolved using JDBC URL including this ID as parameter.

// Register prepared handler with expected ID 'my-unique-id'
// handler: acolyte.jdbc.ConnectionHandler or acolyte.jdbc.StatementHandler instance
acolyte.jdbc.Driver.register("my-unique-id", handler);

// then ...
// ... later as handler has registered with 'my-unique-id'
final String jdbcUrl = "jdbc:acolyte:anything-you-want?handler=my-unique-id";

Connection con = DriverManager.getConnection(jdbcUrl);
// ... Connection |con| is managed through |handler|

If you just want to directly get connection from acolyte.jdbc.Driver, without using JDBC driver registry, you can use Acolyte direct connection:

// handler: acolyte.jdbc.ConnectionHandler or acolyte.jdbc.StatementHandler instance
Connection con = acolyte.jdbc.Driver.connection(handler);

Connection properties

JDBC allows to pass properties to driver to customize connection creation:

Connection con = DriverManager.getConnection(jdbcUrl, someJavaUtilProps);
Connection con = acolyte.jdbc.Driver.connection(handler, someJavaUtilProps);

Acolyte specific properties are:

  • acolyte.parameter.untypedNull: If "true", Acolyte fallbacks untyped null from statement.setObject(p, null) to null string (default: false).
  • acolyte.batch.continueOnError: If "true", Acolyte doesn’t stop executing batch on statement, but continue processing and finally throw BatchUpdateException with update counts of successfully executed elements (see java.sql.Statement#executeBatch).
  • acolyte.resultSet.initOnFirstRow: If "true", Acolyte will degrade JDBC compliance by positioning cursor of result sets initially on the first row, rather than before (as specified by JDBC ResultSet class. It makes Acolyte behaves has Oracle JDBC driver.

Query result creation

Acolyte provides Row and RowList classes (and their sub-classes) to allow easy and typesafe creation of result.

Row lists can be built as following using RowList factory.

import acolyte.jdbc.RowList1;
import acolyte.jdbc.RowList3;

import static acolyte.jdbc.RowLists.rowList1;
import static acolyte.jdbc.RowLists.rowList3; 

// ...

RowList1<String> list1 = RowLists.rowList1(String.class);

RowList3<Integer, Float, Character> list2 = RowLists.
  rowList3(Integer.class, Float.class, Character.class)

In previous example, list1 is a list of row with 1 column whose class is String (VARCHAR as for JDBC/SQL type). Considering list2, it is a list of row with 3 columns, whose classes are Integer, Float and Character.

Column names/labels can also be setup (column first index is 1):

// ...

list1 = list1.withLabel(1, "first label");

list2 = list2.withLabel(2, "first label").withLabel(3, "third name");

Both column classes and names can be declared in bulk way, using definition class:

import static acolyte.jdbc.RowList.Column;

// ...

RowList1<String> list1 = RowLists.
  rowList1(Column(String.class, "first label"));

RowList3<Integer, Float, Character> list2 = RowLists.
  rowList3(Column(Integer.class, "1st"), 
           Column(Float.class, "2nd"), 
           Column(Character.class, "3rd"));

Once you have declared your row list, and before turning it as result set, you can either add rows to it, or leave it empty.

import java.sql.ResultSet;

import static acolyte.jdbc.Rows.row1;

// ...

// we have declared list1 and list2 (see previous example)

list1 = list1.append("str");

ResultSet rs1 = list1.resultSet();
ResultSet rs2 = list2.resultSet();

From previous example, result set rs1 will contain 1 row, whereas rs2 is empty.

Take care to list1 = list1.append("str");. As provided RowList classes are immutable, you should get updated instance from append to work on the list containing added row. This is more safe, and allow to rewrite previous example like:

ResultSet rs1 = list1.append("str").resultSet();
ResultSet rs2 = list2.resultSet();

RowList factory also provide convenience constructor for single column row list:

// Instead of RowLists.rowList1(String.class).append("string") ...
RowLists.stringList("string"/* ... */); // init with value(s)

// Instead of RowLists.rowList1(Boolean.TYPE).append(true) ...
RowLists.booleanList(true/* ... */);

// Instead of RowLists.rowList1(Byte.TYPE).append((byte)1) ...
RowLists.byteList((byte)1/* ... */);

// Instead of RowLists.rowList1(Short.TYPE).append((short)1) ...
RowLists.shortList((short)1/* ... */);

// Instead of RowLists.rowList1(Integer.TYPE).append(1) ...
RowLists.intList(1/* ... */);

// Instead of RowLists.rowList1(Long.TYPE).append(1l) ...
RowLists.longList(1l/* ... */);

// Instead of RowLists.rowList1(Float.TYPE).append(1f) ...
RowLists.floatList(1f/* ... */);

// Instead of RowLists.rowList1(Double.TYPE).append(1d) ...
RowLists.doubleList(1d/* ... */);

// Instead of RowLists.rowList1(BigDecimal.class).append(bigDecimal) ...
RowLists.bigDecimalList(bigDecimal/* ... */);

// Instead of RowLists.rowList1(Date.class).append(date) ...
RowLists.dateList(date/* ... */);

// Instead of RowLists.rowList1(Time.class).append(time) ...
RowLists.timeList(time/* ... */);

// Instead of RowLists.rowList1(Timestamp.class).append(ts) ...
RowLists.timestampList(tsRow/* ... */);

SQL Warnings

Acolyte can also mock up SQL warnings, on update or query, so that java.sql.Statement.getWarnings() will returned expected instance.

import acolyte.jdbc.UpdateResult;
import acolyte.jdbc.QueryResult;

// ...

// Update results to be returned from an acolyte.jdbc.UpdateHandler
UpdateResult upNothingWarn = UpdateResult.Nothing.withWarning("Nothing");
UpdateResult up1ResWithWarn = UpdateResult.One.withWarning("Warning 1");
UpdateResult up10ResWithWarn = new UpdateResult(10).
  withWarning("updateCount = 10 with warning");

// Query result (wrapping row list) to be returned from acolyte.jdbc.QueryHandler
QueryResult nilWithWarning = QueryResult.Nil.withWarning("Nil with warning");
QueryResult resWithWarning = aRowList.asResult().
  withWarning("Row list result with warning");

Generated keys

Update case not only returning update count but also generated keys can be represented with UpdateResult:

import acolyte.jdbc.UpdateResult;
import acolyte.jdbc.RowLists;

// Result with update count == 1 and a generated key 2L

Keys specified on result will be given to JDBC statement .getGeneratedKeys().

Java 8

The module jdbc-java8 provides a JDBC DSL benefiting from Java 8 features.

import acolyte.jdbc.Java8CompositeHandler;
import acolyte.jdbc.RowLists;
import acolyte.jdbc.RowList3;

import acolyte.jdbc.AcolyteDSL.handleStatement;
import acolyte.jdbc.AcolyteDSL.handleQuery2;
import acolyte.jdbc.AcolyteDSL.connection;

connection(handleQuery2((sql, params) -> true));

Java8CompositeHandler handler = handleStatement.
  withQueryDetection("^SELECT ", // regex test from beginning
                     "EXEC that_proc"). // second detection regex
  withUpdateHandler1((sql, ps) -> {
    if (sql.startsWith("DELETE ")) {
      /* Process deletion ... deleted = */ return 2;
    } else {
      /* ... Process ... count = */ return 1;
  withQueryHandler((sql, ps) -> {
    if (sql.startsWith("SELECT ")) {
      return RowLists.rowList1(String.class).asResult();
    } else {
      // ... EXEC that_proc 
      // (see previous withQueryDetection)

      // Prepare list of 2 rows
      // with 3 columns of types String, Float, Date
      RowList3.Impl<String, Float, Date> rows =
        RowLists.rowList3(String.class, Float.class, Date.class).
                          // Optional: set labels
                          withLabel(1, "String").withLabel(3, "Date").
                          append("str", 1.2F, new Date(1l)).
                          append("val", 2.34F, null);

        return rows.asResult();

It can be added to your project with the following dependency.


See online API documentation.

Next: Acolyte for Scala