Using Pseudointernationalisation to test internationalised sites
Pseudointernationalisation is the process of creating a pseudo-translation set for your content and resource files. Not only does this help you to locate any UI elements that have been hard coded into the UI, it will show how different string lengths affect the layout of the site. Additionally it will test your site with extended character sets. The process itself is really simple.
- Pick a language that your site will most likely not utilise.
- Create a set of obviously fake set of translations for that language.
- Switch your site to use the language (by whichever language switching mechanisms are available to your site)
- Observe the new layout and go from there.
Your overall strategy will change depending on the target languages, but of course, the closer you can simulate these languages the fewer anomalies you will have on your production site. There are general rules of thumb, and statistical data that you can use to help you get close. For example, a short word in English will translate to something 4 times as long in some Latin based languages.
To help generate these translations, you can use Psuedoizer. Psuedoizer is a really simple tool that takes a few of these rules into account, generating the translated content for you. It’s a console application that will convert your RESX files into a pseudo-translation. The pseudo-translation is essentially English substituted character for character from the extended character set. For example, an English ‘B’ will become a Greek Beta. It will simulate extended length by padding the translations. The benefit of using pseudo-English is that the site is still readable. And you get full access to the source, so you can easily incorporate translations into your data seeding mechanism for your test environment.
Being a console application, it can easily be included as part of a build script. You can get the Psuedoizer tool here: https://github.com/shanselman/Psuedoizer
