Code & Data: from Assembler to RVL

June 7th, 2017 by Adam Sandman

In this short blog post, our test automation guru - Denis Markovtsev explains the ideas and principles behind the design of Inflectra's new Rapise Visual Language (RVL) introduced in Rapise 5.1. RVL is a spreadsheet-based approach to UI test automation implemented in Rapise to help domain specialists and test analysts, who are not programmers, participate in test automation projects.

Introduction

Since the beginning of computer era, programmers divided code and static data. Let's briefly trace history starting with x86 assembler:

        .model  small
        .stack  128

        .code
start:  mov     ax, @data
        mov     ds, ax
        mov     ah, 9
        lea     dx, msg
        int     21h
        mov     ah, 4ch
        int     21h

        .data
msg     byte    'Hello World!', 13, 10, '$'

        end     start

Notice those .code, .data segments and msg variable containing Hello World! string.

In C language the example turns into.

#include <stdio.h>

static const char* msg = "Hello World!\r\n";

int main()
{
    printf(msg);
    return 0;
}

And here is the one of the most modern successors of C - C#.

using System;

namespace CSharpConsoleApplication
{
    class Program
    {
        private static string msg = "Hello World!";
        static void Main(string[] args)
        {
            Console.WriteLine(msg);
        }
    }
}

Separation of code and data is a natural approach providing many benefits.

Same data variable may be used in several places in the code.
If variable value is changed no need to update the code.
Data variables may be grouped together for better visibility.
Data may be defined in external sources which makes it easier to maintain and update.

Notice that writing Hello World! example in any of the languages mentioned above requires knowledge of

Keywords (like .model, .stack, char, main, namespace, public).
Syntax rules for using symbols like {, }, (, ), [, ].
Rules for definition of string constants.

RVL Maps

At Inflectra we kept in mind all these considerations while designing Maps feature for RVL and tried to simplify syntax as much as possible. Here is how Hello World! looks like in RVL:

Actually, this example does a more complex thing than printing Hello World! to a report. It has a loop that iterates through a data set (map) and prints all defined values. Adding more values

and see the execution result:

Let's iterate through a list of great capabilities expressed by RVL in this example:

Data and code (logic) are separated.
Data and code are in the same document which makes it easy to bootstrap a working prototype of the test.
No special knowledge of string formatting is required. One can just write values in spreadsheet cells.

Want to place data values in rows rather than columns. Excellent:

Now imagine you want to work with a particular row in the map. Just put this on a desired row and the loop will execute just one iteration for it.

Easy and simple.

Conclusion

RVL is an effort to easy life for non-programmers and enable them to participate in UI test automation projects more effectively. RVL is based on many proven concepts developed for programming languages since assembler birth times and is build around simple spreadsheet-inspired notation.