Matreshka is Ada framework to develop information systems

The framework includes:

• League - a rich set of reusable core components to develop Ada applications. Its main purpose is to provide a high level abstraction tool for localization, internationalization and globalization of applications, as well as a portable interface to different operating systems. It contains many other useful features, among them advanced calendrical calculations, regular expressions, and JSON support to process and generate data in JSON format;

• XML processor - provides the capability of manipulating XML streams and documents; including:

  • SAX reader to read XML streams and documents; it supports XML1.0 (Fifth Edition), XML1.1 (Second Edition), Namespaces in XML and XML Base specifications;
  • SAX writer to generate XML streams and documents from application;
  • DOM API to process XML documents 0.8;
  • XML Catalogs resolver;
  • Markup Templates Engine to process XML templates and generates XML or HTML/XHTML output 0.6.

• Web framework

  • standalone application server - SpikeDog implemented over AWS (Ada Web Server) with Servlet API 0.7
  • the FastCGI module assists with developing server side applications completely in Ada and using them with standard HTTP servers. (demo).
  • the SOAP module provides implementation of SOAP 1.2 protocol specification and assists in developing Web Services in Ada. This module includes implementation of standard security services: The WS-Security module provides SOAP message Security 1.1 (WS-Security 2004) and Web Service security: Username Token Profile 1.1.
  • WSDL to Ada translator.
  • Ada to JavaScript translator and run-time library.

• SQL database access provides a simple generic API for accessing SQL databases. Supported databases include:

  • MySQL is the world's most popular open source database software;
  • Oracle is a well known commercial database management system;
  • PostgreSQL is a powerful, open source object-relational database system;
  • SQLite3 is a in-process library that implements a self-contained, serverless, zero-configuration, transactional SQL database engine;
  • Firebird is a relational database offering many ANSI SQL standard features that runs on Linux, Windows, and a variety of Unix platforms.

• Ada Modeling Framework provides implementation of OMG's Meta Object Facility (MOF) written completely in Ada. Extension modules are provided to assist in the analysis and modification of:

  • UML models and their extensions:
    • MOF Extensions to support metamodeling
    • OCL models
    • UML Testing Profile to support Model-Driven Testing
  • Diagram Definition

Large number of specifications are used and are supported, see Specifications for complete information about conformance to specifications.

• Install
  • Dependencies
  • Linux and other POSIX environment
  • Windows 32bit
  • Windows 64bit
  • Mac OS
  • Advanced installation
• Usage
• Maintainers
• Contribute
• License

You need GNAT and gprbuild to build Matreshka from sources.

Some optional packages are used by matreshka

• AWS, required to build SpikeDog Application Server
• ASIS, required to build a2js
• MySQL, PostgreSQL, SQLite3, Oracle client libraries are required to build corresponding database driver
• Valgrind, nodejs, requirejs, js-beautify are reuired to run tests

To install Matreshka unpack its source code archive and run

make all install

Automatic configuration utility will be build and run to configure Matreshka to meet your machine and operating system.

On Windows system you need make, install and a few other commands. The easiest way to get them is to install Make and CoreUtils packages from GnuWin32 port.

Note for Windows 10 users: Windows assumes that install should ask for elevated privileges and prevents it to work as expected. To workaround that issue launch the Local Security Policy editor and

1. Go to Local policies / Security Options
2. Set User Account Control: Detect application installations and prompt for elevation to disabled

Another option - install msys2 MinGW or Cygwin environment.

Use GNAT Community 2018 Ada compiler.

Alternatively you can use GCC from msys2 distribution. Follow these installation instructions for msys2:

• Download and install msys2 as described on https://msys2.github.io/
• Install Ada compiler `pacman -S mingw-w64-gcc-ada
• Now you need to replace unworking *.dll.a with corresponding dll. Do this in msys2 console:

adalib=$(dirname `gcc -print-libgcc-file-name`)/adalib
bin_dir=$(dirname `which gcc`)
rm -f ${adalib}/libgna{t,rl}-*.dll.a
cp ${bin_dir}/libgna{t,rl}-*.dll ${adalib}/

To install gprbuild setup our repository with precompiled packages:

echo [ci.mingw] >> /etc/pacman.conf
echo SigLevel = Optional >> /etc/pacman.conf
echo Server = https://dl.bintray.com/reznikmm/msys2/ >> /etc/pacman.conf

Install gprbuild-gpl (and asis if you like a2js):

pacman -S mingw-w64-gprbuild-gpl mingw-w64-asis

Continue with

make all install

Use GNAT GPL 2017 Ada compiler. You also need make and install utilities, see Windows 64bit for details.

You may use GNAT GPL or GCC from sf.net.

On Mac OS system you need "Command Line Tools" installed. You can install this from Xcode menu > Preferences > Downloads or, alternatively, there are stand-alone installation packages

• for Mountain Lion (10.8)
• for Mavericks (10.9).

Sometimes, it is needed to have more control over configuration and build of Matreshka. One example is use Matreshka on Windows, where detection of non-standard libraries is problematic. Another example is binary packages supporters.

So, build of Matreshka consists from several steps:

• build of configuration utility;
• start configuration utility;
• package build;
• package installation.

Configuration utility can be built by using

make config

Execution of configuration utility in its simplest form looks like:

./configure

This is exactly what default make target do, so there is no reasons to use it.

Configuration utility allows to specify several additional parameters all of them are reported by

./configure --help

Below is small example of manual execution of configuration utility to install Matreshka in specified directory and to lookup SQLite's libraries in the specified directory on Windows:

configure.exe --prefix=C:\Matreshka --with-sqlite3-libdir=C:\Matreshka\lib

It is possible to disable build of some Matreshka's modules. For example, to disable AMF module run configuration utility with --disable-amf switch:

./configure --disable-amf

To build Matreshka after manual configuration it is sufficient to run

make

Several make's variables can be used to adjust build process:

• SMP_MFLAGS can be used to utilize several CPUs to compile Matreshka;
• GNAT_OPTFLAGS can be used to pass additional user defined switches for compiler.

And to install Matreshka:

make install

Install target is used DESTDIR environment variable to adjust root directory, it is very useful for binary packages supporters.

Use corresponding project file for instance

with "matreshka_league";
project My_Project is

end My_Project;

There are some examples in examples directory.

Other examples could be found on the site:

• Strings
• Codecs
• Text translator
• String Vectors
• Settings
• Calendar
• Regular Expressions
• XML SAX
• XML Engine
• Spikedog
• Database Access
• UML

• @Vadim Godunko.
• @Maxim Reznik.

Feel free to dive in! Open an issue or submit PRs.

BSD © Vadim Godunko