LocalizedForms.md

Localized Forms

If you write software that renders text, for example a word processor, a web browser, or an operating system library, we hope that you’ll find TestFont-LocalizedForms.otf useful for testing.

What are localized letter forms?

In Asian languages such as Chinese, Japanese, and Korean, many characters look very similar. Therefore, Unicode has assigned them a single codepoint; this is called “Han Unification.” However, while these characters have the same root and the same meaning, their typographic rendering style is recognizably different across languages. To render text in acceptable quality, a graphics library needs to know the text language. If the font supports this, typographic adjustments will tweak the appearance of the text.

Localized forms also exist in languages outside of Asia. An example is Polish, where the “kreska” mark on the letters ćńóśź should be slightly less angled than in other languages. Another example is Wolof versus Sami, which both use the eng letter Ŋŋ but in slightly different style.

Why should I care?

If you build a web browser, typesetting system, or similar software application, you should make sure that localized letter forms work correctly; you system will work better for text in certain languages. Of course, there’s always things beyond your control that can go wrong. For example, if you receive input tagged with the wrong language, or if your operating system comes with buggy fonts, there is not much you can do. But if your software passes this test, you can be reasonably sure that your system has done its part of the overall picture.

How to identify languages?

Modern web browsers (and similar software) use a tagging system called IETF BCP47 for identifying languages. For example, the BCP47 code for the Welsh language is cy. But not all BCP47 codes are just two letters: for example, de-LI-fonipa means German as used in Liechtenstein, written with the symbols of the International Phonetic Alphabet. Should you build a build a new software library for handling text, its interface should definitely take BCP47 codes.

However, since OpenType is much older than BCP47, its fonts have differnt language tags. For example, the OpenType code for Welsh is WEL. Your text rendering software needs to bridge this gap, and this test font can help making sure that there are no bugs.

What does the font do?

The test font will always rewrite letter sequences like OT:WEL to a face symbol. If the font’s locl feature for Welsh has been activated, this face will be a smiley; otherwise, it will have a frowning look. The frowning version also uses much more ink, so it is easy to spot among a sea of smileys.

The font has a locl feature for every language system tag defined in OpenType 1.7.

How can I use this font for testing?

For testing web browsers, simply go to our browser test page. For testing word processors and other desktop applications, install TestFont-LocalizedForms.otf in your operating system. Then, type the characters OT:PLK into a new document, mark them as Polish using whatever widget your application may offer for this, and select “TestFont-LocalizedForms” as the font. If your application supports Polish letterforms, you’ll see a happy face; otherwise you’ve found a bug worth reporting. Make sure to test every language in the OpenType language registry.