list-n-m-n.txt

N:M / M:N List Entry Handling in Fonts
======================================

A list may include more than one Unicode character under a single
entry--this is an N:M correspondence.

Conversely, a single Unicode character may occur in more than one
entry--this is an M:N correspondence.

A general mechanism for these cases is the addition of OpenType
feature codes after the list number; these are most often SALT codes,
but may be style-set (SS) or character-variation (CV) codes as well.

When features are given with @list and are used for a list page, the
first listed feature is used as the reference form for the list
entry. To use the unmarked glyph (the one with no feature extensions)
as the reference form, give '.0' as the first feature.

N:M
===

Where a single Unicode character corresponds to multiple entries the
font should merge all of the sign forms as SALT alternates of the
Unicode character.  The SALT forms of different list entries can be
noted in the @list entry in the abbreviated form consisting of
period+digit(s):

 @list RSP006 .0 .1 .2 .3 .4
 @list RSP008 .5 .6

One of the glyphs should be selected as the base form for the Unicode
character--this can be from any of the list entries and does not need
to be the lead character of any of the list entries.


M:N
===

Where a single list entry contains more than one character, the entry
should be given sub-entry indexes using the carat, ^, character.  The
list display will merge the sub-entries and preserve the association
of glyphs and sub-entries automatically:

 @sign DIŠ
 @list RSP005^a
 ...

 @sign DIŠ@t
 @list RSP005^b
 ...


@list/@lref Extended Syntax
===========================

The simplest way to gloss @list and @lref entries is to follow them
with @note.  In addition, an Extended Syntax (ES) can be used to set
the fields that are used in list pages.

The fields are: sign-name; unicode-or-image; values.  Feature lists
may follow the list ES and the usual @note can also be used for
additional unstructured commentary on the @list/@lref.


    @list/@lref LIST  SIGN_NAME ( <image-filename | utokens ) atftokens OTFEAT

An abbreviated form can be used when the @list/@lref only has an image:

   @lref LIST <image-filename

If there is no sign-name, but other fields are needed, an uppercase
'O' should be used for the sign-name.  If there are no values but
other fields are needed, a single hyphen, '-', should be used.  In
both cases the list output for these fields in the .tsv is an empty
field, i.e., the 'O' or '-' are deleted in the output.