11. Tagging and tracing Program Integrated Information
Makoto Arisawa | Faculty of Graduate School of Media and Governance, Keio University
There are two main types of translation involving computer programs. One involves manuals and the other involves ProgramIntegrated Information (PII). This chapter focuses on PII translation. PII translation is substantially different from ordinary text translation. PII is separated out of the programs into externalised text resource files to allow for translation outside the program development laboratory. The contexts of the operations have been discarded. The translators have to translate phrases and words without context in these text resource files. The Translation Verification Test (TVT), which is done with the normal operations of the program, compensates for the lack of context during translation. If the TVT tester finds an inappropriate translation in the GUI (Graphical User Interface), the file it came from and which line in the file is unknown. We have developed a utility program to make it easy to find a source location. The utility adds a short group of ID characters in front of every PII string. We used this systematic approach for CATIA®1 and found many advantages, such as locating the hard-coded strings that are the biggest problems in programinternationalisation. This ID can be inserted independently of program development. We also developed a utility program that helps TVT testers refer to both the original and target PII strings as pairs. This chapter describes the approach in detail. In addition, this chapter presents statistics about PII files. This important statistical information has not been considered in the program internationalisation communities.