Skip to content

library-architecture-and-functionality.md

Latest commit

 

History

History
83 lines (59 loc) · 3.1 KB

library-architecture-and-functionality.md

File metadata and controls

83 lines (59 loc) · 3.1 KB
Raw

Library Architecture and Functionality

The project consists of two main packages.

The dbaccess package contains a set of core classes that enable database access. It is a proprietary, very simple yet effective solution for building a data access layer without using external frameworks—only JDBC and a database driver are required. This package includes classes responsible for connection handling, connection pooling, a custom transaction abstraction, base classes for repositories, SQL statement handling, and a set of simple helper utilities.

The ragdeterm package contains a collection of entity and repository definitions created to support the RAGdeterm tables. The structure diagram of these tables is shown below:

Image description

Entity classes use the Entity suffix, while basic CRUD repositories are classes with the Repo suffix (these classes are not implemented manually—they are generated automatically). Additional methods (most often custom queries) that go beyond the standard functionality are placed in classes with the Repox suffix.

The ragdeterm package also includes low-level tools for building RAGdeterm-type contexts. These are the classes CooperationCtxRepox, FullCtxRepox, and StructureCtxRepox.

Configuration and Initial Startup

Before starting work with RAGdeterm, a database must be created. The project uses a PostgreSQL database. The scripts for creating the database are located in the src/main/resources/sql folder of the project. First, the database itself is created (_create_db.txt), followed by the tables (create_table.sql).

After creating the database, the correct URL should look as follows (assuming the default port):

jdbc:postgresql://localhost:5432/ragdeterm?currentSchema=rag

Typical Usage Scenario

The db-lib project is a library used by other modules. A typical usage scenario can be found in several test classes included in the project.

First, database access must be defined as a set of connection pool parameters (the Hikari library is used). This is done using the following code:

DbConnContainer.addDbConn(new DbConnConfig("Default",
        Consta.CONN_URL, Consta.CONN_USER, Consta.CONN_PSW, Consta.CONN_DRIVER,
        false, 20, 180000
));

where the following constants are used:

static public final String CONN_URL = "jdbc:postgresql://localhost:5432/ragdeterm?currentSchema=rag";
static public final String CONN_USER = "ragdeterm";
static public final String CONN_PSW = "ragdeterm";
static public final String CONN_DRIVER = "org.postgresql.ds.PGSimpleDataSource";

The DbConnContainer.addDbConn method is static and defines the default connection pool (the library supports multiple pools, but the RAGdeterm project uses only one).

Database operations are always executed within a block similar to the one below:

try (var trx = DbConnContainer.newTrx()) {

    ...
    ...

    trx.commit();
} catch (Exception ex) {
    throw new DbException(ex.getMessage(), ex);
}

The DbConnContainer.newTrx() method creates and retrieves a new database connection from the default pool.