Greek Fonts (variant character forms)

From The Digital Classicist Wiki
Jump to navigation Jump to search

How to handle variant characters forms in Greek

There are several issues involved here.

  1. How you should encode (e.g.) a sigma in the web page or database so that you can find 3-bar, 4-bar, lunate etc. sigmas using a single search, say for ΣΩΖ-. The answer this is emphatically that the Unicode character for sigma should be used, rather than (say), something that looks like the 3-bar sigma such as the lightning-bolt koppa at U+03DF.
  2. How should you display a lightning-bolt shaped character on a screen or printed page? Ideally the answer would be to use a font with variant form sigmas (see 3. below), but while waiting for this to be implemented, one might decide to use a similarly shaped character from a different codepoint for display only.
  3. How to indicate via markup that a particular sigma is 3-bar or 4-bar or lunate, vel sim., in such a way that a web browser/document printer etc can extract and deal with this information thus printing it correctly on the scree/page?

It is possible to encode an OpenType font with multiple glyphs for any given character codepoint (so multiple variant sigmas, epsilons, alphas, etc.). Dumbarton Oaks has an example font, Athena Ruby.

David Perry writes (November 2006):

1. The version of Cardo that is available on my web page (v. .98) has a very large number of Greek variant letterforms. These are located in the supplementary private use area, beginning at U+F0000, so you won't see them if you look at the font with Windows Character Map or Word's Insert/Symbol (these don't handle anything beyond the BMP). You can see them with BabelMap or look at the user's manual beginning on p. 34 for a complete list.

2. In addition to the PUA codepoints, these variants are all accessible through the OpenType stylistic alternates feature if you have a program that supports it (Mellel and InDesign, as far as I know at the moment). It's frustrating that a mechanism exists that allows standard Unicode values for storage and alternate forms for display, but it is implemented in so few applications. That may change with the release of Windows Vista, which supports many OpenType features at the system level. Or it may not . . .

See also