JavaScript is eating the world. It has become the most popular programming language on GitHub. Since it is widely used in both client-side and server-side programming in web applications, the robustness and performance of JavaScript programs become vital. Unfortunately, there is limited study on JavaScript issues. Web developers rarely have access to useful tools that can help produce high-quality JavaScript code. In this work, we present nine issue patterns derived from open-source projects and propose a general static analysis framework, JSOptimizer, to help detect such patterns of issues and optimize the code accordingly. We applied our framework to seven real open-source JavaScript projects, and five bugs detected by our framework have been confirmed by developers. Besides, we conducted a case study based on a popular project, which shows that addressing the issues detected by our framework can speed up the original project by over 300%.