---
section: 32
x-masysma-name: uni/dke/u03/11
title: "UNI DKE U03/11: Relational Algebra Interpreter"
date: 2016/05/05 23:55:56
lang: en-US
author: ["Linux-Fan, Ma_Sys.ma (Ma_Sys.ma@web.de)"]
keywords: ["uni", "dke", "relation", "algebra", "interpreter", "ra"]
x-masysma-version: 1.0.0
x-masysma-repository: https://www.github.com/m7a/bp-uni-dke-u03-11-ra-interpreter
x-masysma-website: https://masysma.net/32/uni_dke_u03_11.xhtml
x-masysma-owned: 1
x-masysma-copyright: |
  Ma_Sys.ma Relation Algebra Interpreter 1.0.0.0, Copyright (c) 2016 Ma_Sys.ma.
  For further info send an e-mail to Ma_Sys.ma@web.de.
---
Introduction
============

The _Relation Algebra Interpreter_ (or short RA Interpreter) is a program
which can read CSV tables and perform some basic relation algebra on it, that
is to say: Project, Join and Select operations.

This program was written for learning purposes. If you find it useful for
anything else, keep in mind, that it was probably not designed to do what you
want.

Compilation
===========

If you are on a Linux system with the GNU Ada Compiler GNAT installed, it should
be enough to type `make` to build this program. If you want to use Docker, you
may either build this image with `docker build .` or run it directly with

	docker run -it --rm masysma/uni-dke-u03-11-ra-interpreter

Otherwise, you will need to compile the program with any Ada 2012 compiler.
The main file is `ra_interpreter.adb`.

Usage
=====

The program runs interactively on the console and takes commands as input
and prints resulting tables directly to the console.

Format your files as comma separated value, for example like `test/t1.csv`

	File,Lines
	gpl.ads,24
	ra_interpreter.adb,26
	csv.ads,32
	func.ads,37

The first row is considered the schema of this relation which you can
interactively load as follows:

	RA> read(t1, test/t1.csv)

You will not get an answer, because you just load the relation into memory
under the name `t1` (first parameter to `read`).

In order to print a table, select all of its rows:

	RA> SEL[1](t1)
	ooooooooooooooooooooooooooooo
	  File                Lines
	+++++++++++++++++++++++++++++
	  gpl.ads             24
	  ra_interpreter.adb  26
	  csv.ads             32
	  func.ads            37
	ooooooooooooooooooooooooooooo

It pretty-printed the relation.

Now, let's add a condition to the select statement:

	RA> SEL[(File = "csv.ads")](t1)
	oooooooooooooooooo
	  File     Lines
	++++++++++++++++++
	  csv.ads  32
	oooooooooooooooooo

Note the parentheses around the condition, we wrote `SEL[(...)]`. If you consult
what `help` has to say about this:

	RA> help
	[...]
	Bool:  (v = v)    Compares values.
	[...]

It says, that you have to put a comparison in parentheses This simplifies
parsing but is a bit more to write. Also, comparison only works for values.
You cannot compare Booleans with this program.

Finally, let's have a look at an example for Natural Joining. Load the relation
from `test/t2.csv` as `t2` and perform a join on the two tables:

	RA> load(t2, test/t2.csv)
	RA> JOIN(t1,t2)
	ooooooooooooooooooooooooooooooooooooooooo
	  File                Lines  Size/Bytes
	+++++++++++++++++++++++++++++++++++++++++
	  gpl.ads             24     858
	  ra_interpreter.adb  26     1038
	  csv.ads             32     1126
	  func.ads            37     1400
	ooooooooooooooooooooooooooooooooooooooooo

... It works as expected joining the two relations on the common field `File`.

Explore the other options by viewing `help`. Most of them work as expected.

Issues
======

As this was created for learning purposes, it's features are only rudimentary.
Especially, you can not

 * Recall commands via a history
 * Use another syntax than given there, spaces are sometimes relevant!
 * Do a lot of boolean operations
 * Use fields longer than 256 characters :(

Programming Language
:   This was written in Ada as a means of getting to know that language.
    Thus, there are probably a lot of beginner's mistakes in there...