Design of a New Language Seeks Literature Survey

In a scientific study, computing is a must-have tool. In general, scientists have various difficulties, requirements, and views when it comes to computation, which need to be addressed by the programming language that they use, this cannot be satisfied by general-purpose languages. Also, researchers need to concentrate on the issue they are working on rather than the optimizations for the calculations, so instead of using a general-purpose language, if there exists a language whose compiler would take care of those optimizations, it would make their work easier and faster. This is a survey of the work undertaken to design the programming language and its compiler. The primary goal of this research is to examine the function of work, implementation strategy, steps taken for improving the performance, the procedure of benchmarking, and finally, the outcome of the papers studied. The survey's main conclusions are that: the most common language mentioned among the papers was Python which appears to be more popular among developers due to its simple syntax and library support for computing. On the other hand, Python lacks performance, to compensate for this performance issue, the community has developed tools like Cython, Numba, Pythran, etc, which can be used to speed up Python. Domain-specific languages such as Wolfram, Seq, and ELI highlighted various methods for overcoming problems. Some languages like Wolfram and ELI moved from interpreter to compiler to get the performance boost. Most of the compilers use LLVM as the backend for optimizations and code generation. Keywords: scientific computation, compiler, programming language