Putra, Habridio Kurniawan (2019) Pengembangan Sistem Aplikasi Pendeteksi Long Method Smell Berdasarkan Refactoring Filtering Metrics. Sarjana thesis, Universitas Brawijaya.
Abstract
Dalam pengembangan atau perubahan suatu perangkat lunak perlu memperhatikan nilai maintainability agar perangkat lunak tersebut tidak memiliki nilai maintainability rendah. Nilai maintainability yang rendah dapat membuat perangkat lunak tersebut susah dimengerti, mahal untuk diperbaiki dan kembangkan atau diubah dikemudian hari. Salah satu penyebab rendahnya nilai maintainability adalah code smell. Tidak semua developer memiliki pengetahuan yang cukup mengenai code smell, sekalipun mengerti belum tentu dapat menemukan solusi yang harus diambil untuk mengatasi code smell tersebut. Sistem deteksi long method smell ini memiliki fitur untuk membantu developer dalam mendeteksi long method smell serta cara refactoring yang dapat dilakukan dari source code java sehingga dapat mengurangi effort developer dalam mengatasi code smell khususnya long method. Sistem ini dikembangkan berdasarkan Refactoring Filtering Metrics(RFM) yang didalamnya menghitung nilai switch, edge, node, def, c-use, p-use dan complexity of condition. Nilai RFM tersebut dijadikan dasar mendeteksi long method smell menggunakan Refactoring Filtering Condition. Terdapat lima kondisi pada Refactoring Filtering Condition dan masing masing kondisi memiliki solusi berupa teknik refactoring yang berbeda beda. Sistem ini telah diuji dengan menggunakan pengujian unit yang menggunakan metode Whitebox serta pengujian validasi menggunakan metode Blackbox. Sistem ini dapat melakukan deteksi dalam kurun waktu kurang dari 10 detik serta memiliki akurasi hasil deteksi sebesar 100 persen.
English Abstract
In developing or changing a software, it is necessary to observe the maintainability so that the software does not have low maintainability. Low maintainability can make the software difficult to understand, expensive to repair and develop or change in the future. One of the causes of low maintainability is code smell. Not all developers have enough knowledge about code smell, even though they understand about it, it is uncertain if they can necessarily find a solution that must be taken to overcome the code smell or not. The long method smell detection system has a feature to help developers detect it, and effective methods and refactoring methods that can be done from java source code so that it can reduce the effort of developers to overcome code smell, especially long methods. This system developed based on Refactoring Filtering Metrics (RFM), which calculates switch, edge, node, def, c-use, p-use and complexity of condition values. The RFM value used as the basis for detecting the long method smell using Refactoring Filtering Condition. There are five conditions in Refactoring Filtering Condition and each condition has a solution in the form of refactoring techniques that are different. This system has been tested by the Whitebox method and validation testing using the Blackbox method. This system can operate detection in less than 10 seconds and has a detection accuracy of 100 percent.
| Item Type: | Thesis (Sarjana) |
|---|---|
| Identification Number: | SKR/FILKOM/2019/308/051905626 |
| Uncontrolled Keywords: | - |
| Subjects: | 000 Computer science, information and general works > 004 Computer science > 004.2 System analysis and design, computer architecture, performance evaluation |
| Divisions: | Fakultas Ilmu Komputer > Teknik Informatika |
| Depositing User: | Endang Susworini |
| Date Deposited: | 24 Aug 2020 06:49 |
| Last Modified: | 24 Aug 2020 06:49 |
| URI: | http://repository.ub.ac.id/id/eprint/171223 |
Actions (login required)
 |
View Item |