[XeTeX] Ligature unavailable in Estrangelo Antioch

[Sargon Hasso]

On Thu, May 14, 2009 at 11:40 PM, Sargon Hasso <sargon.hasso at gmail.com>wrote:

> Gareth,
> I don't have the Estrangelo Antioch font (I will check them after I find
> them), but recall I did mention that I worked on the meltho fonts. It
> appears that the ligature rules are not defined for lamad+alap or taw+alap.
> In my East Syriac Fonts, I do have them defined (see attached).
> You can see clearly the lamad+alap ligature (ignoring your rendered
> namedglyph: mine is mapped to a different glyph).
>
> For taw+alap, in this version of the font, you might think it is not a
> ligature, in fact it is. This was a font that someone did not want to have a
> different taw+alap ligature. In East Syriac fonts, we have 3 different
> taw+alap ligatures. There are no rules which one should be active.
> Originally, we had a rule that always picked a ligature, but some people did
> not like that rule so we relaxed it. For those people who wanted the
> taw+alap rule imposed, they wanted to have an option to select among the 3
> ligatures. In TTF specs, you could have have more than one alternative, but,
> unfortunately, Microsoft did not implement this ligature selection
> alternative in their rendering engine (uniscribe). So, we cheated, and
> created 3 different versions of the same font with obligatory taw+alap rule
> each with different ligature to cater to different audience.
>
> The sade+nun: although you mentioned that it renders correctly, it is in
> fact wrong. This ligature should only renders if it falls at the end of the
> word. In my font, sade+nun at the beginning of a word is not ligated.
> In Syriac, we really don't have mandatory ligature rules (with the
> exception, I believe, of the lamad+alap, but this has other complications).
> Some people like the ligatures, others don't. Unfortunately, like I
> mentioned above, we don't have that option in the TTF implementation of
> Microsoft. If XeLaTeX can utilize that feature, I can certainly add those
> additional ligatures in the Syriac fonts and have the user control the
> selection. I will dig up my TTF specs and refresh my memory.
> In East Syriac: we have two other ligatures: the he+yud, and taw+yud.
> Again, these are optional ligatures and not all people like them. In West
> Syriac fonts, there some additional ligatures like: lamad+lamad+alap (it
> looks better with a ligature), alap+lamad, and I think few more that make
> the character combination aesthetically more pleasing in ligated form than
> their isolated form.
>
> Look at the last row of this glyph table, and you will see several
> ligatures for taw+alap (forgive the print screen capture quality).
>
>
> Sargon
>
>
> Gareth Hughes wrote:
>
> Estrangelo Antioch is one of the freely available Meltho fonts for
> Syriac. I can see that the font has three ligature glyphs, but XeTeX is
> only forming one of them, to access the others I need to make a special
> glyph call. The minimal input and output are attached. When I use
> Fontforge to read the lookups, the two ligatures are mentioned in
> ligature substitution subtable 9 in joining and isolated forms. However,
> I don't know how to get this table to be used.
>
> Gareth.
>
>
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://tug.org/pipermail/xetex/attachments/20090515/df918b4b/attachment-0001.html 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/jpeg
Size: 104171 bytes
Desc: not available
Url : http://tug.org/pipermail/xetex/attachments/20090515/df918b4b/attachment-0001.jpe