Regular expressions are a powerful tool for searching words in a text corpus by a specific pattern. Regular expressions are represented in many programming languages with different implementations. These implementations often have varying syntax and mathematical bases, which can create issues with backward compatibility and optimization of the expressions themselves. To address this, we propose a universal translator that can translate one syntax to another with the minimization of intermediate states to speed up the regular expression's operation. Optimization techniques such as naive replacement of fuzzy boundaries, refinement of search patterns, and finding more efficient expressions using genetic evolution algorithms are suggested.