script-metadata

http://unicode.org/cldr/trac/ticket/3871

Here is a proposed structure for supplemental data

<script>

<scriptData type="Latn" lines="top-to-bottom" characters="left-to-right" spaces="true" shaping="minimal" usage="recommended" originRegion="150" sample="A">

lines/characters =

• • as in the locale orientation element.

spaces =

• • true if the script normally uses spaces to separate words

shaping = (in normal usage)

• • none if no shaping is normally needed

  • minimal if only minor shaping is normally needed, such as accent placement

  • major if glyphs normally need to rearrange, or change shape depending on context

usage = value in UAX31

originRegion = continent or subcontinent where script originated. Following the Unicode book / charts (see the spreadsheet).

sample = character with distinctive glyph that can be used to represent the script in icons (eg missing glyph)

The orientation field in a locale would only be used for overriding.