Cartoons are an important part of our entertainment culture. Though drawing a cartoon is not for everyone, creating it using an arrangement of basic geometric primitives that approximates that character is a fairly frequent technique in art. The key motivation behind this technique is that human bodies - as well as cartoon figures - can be split down into various basic geometric primitives. Numerous tutorials are available that demonstrate how to draw figures using an appropriate arrangement of fundamental shapes, thus assisting us in creating cartoon characters. This technique is very beneficial for children in terms of teaching them how to draw cartoons. In this paper, we develop a tool - shape2toon - that aims to automate this approach by utilizing a generative adversarial network which combines geometric primitives (i.e. circles) and generate a cartoon figure (i.e. Mickey Mouse) depending on the given approximation. For this purpose, we created a dataset of geometrically represented cartoon characters. We apply an image-to-image translation technique on our dataset and report the results in this paper. The experimental results show that our system can generate cartoon characters from input layout of geometric shapes. In addition, we demonstrate a web-based tool as a practical implication of our work.