Programinės įrangos testavimas internacionalizavimo požiūriu
Software Testing According to Internationalization Bugs

Summary/Abstract: Bugs, found in localized software, may be classified into two categories: localization bugs and internationalization bugs. The localization bugs are made by the localizer and may be fixed by the localizer. Internationalization bugs are made by the author of software source code and may be fixed by the author. The paper discusses how the two kinds of bugs can be separated from each other in the localized program, or how to find internationalization bugs in original program before its localization. Early bug reports enable product's author to fix bugs before starting the localization process, or before the public release of the product. In the worst case, early identification of internationalization bugs enables the localizer to estimate the influence of not fixed bugs on localization quality. The tools and methods to identify the internationalization bugs are discussed: analysis of localization resources, pseudotranslation (pseudolocalization), local experimental localization, experimentation with the program during its run-time. The investigation was based on testing of ten popular programs. Some experience on search for internalization bugs is presented as well. The questionnaire with the list of the most frequent internationalisation bugs is presented. The bugs are classified into three categories: alphabetic symbols, formats, and texts. The questionnaire may be considered as a guide for testers enabling to find bugs faster and in deeper layers of the program.

• Details
• Contents