ConTeXt, XeTeX, special effects in Indic font for Kannada
On my Mac (Mountain Lion) I would like to use ConTeXt with a complex OpenType font named Kedage for the Indian language Kannada, and I thought that I could do this using XeTeX (because LuaTeX seems unable to handle Indic fonts at present). This freely available font (found on Linux systems, for instance) comes in four files: Kedage-n.TTF, Kedage-b.TTF, Kedage-i.TTF, Kedage-t.TTF (normal, bold, italic and bolditalic, respectively). I have been partly successful with the simple command \definefontsynonym[kan][name:Kedage] and then using \definedfont[kan] where I want to have the font in the text. Also \font\kannada="Kedage" (and \kannada later in the text) seems to work well without any further indications of language, script etc. However, I would also like to use the bold, italic and bolditalic forms of the font. Simply using \bold and \em does not work. Here I see two possibilities: (1) use the XeTeX parameters FakeSlant and FakeBold, which I have done in LaTeX. But I do not know how this is done in ConTeXt. (2) use the other font files (Kedage-b.TTF etc.), and again I do not know how to do this. Can anybody help me out, either way? (I have attached a sample.) Robert
On 8/8/2013 1:00 PM, Robert Zydenbos wrote:
On my Mac (Mountain Lion) I would like to use ConTeXt with a complex OpenType font named Kedage for the Indian language Kannada, and I thought that I could do this using XeTeX (because LuaTeX seems unable to handle Indic fonts at present).
forget about xetex, first check what context mkiv van do
This freely available font (found on Linux systems, for instance) comes in four files: Kedage-n.TTF, Kedage-b.TTF, Kedage-i.TTF, Kedage-t.TTF (normal, bold, italic and bolditalic, respectively).
I have been partly successful with the simple command
\definefontsynonym[kan][name:Kedage]
and then using
\definedfont[kan]
where I want to have the font in the text. Also \font\kannada="Kedage" (and \kannada later in the text) seems to work well without any further indications of language, script etc.
it all depends on how much processing is needed in context, there is experimental support for devanagari \definedfont[file:kedage-n*devanagari-one] but it all depends on what the font provides
However, I would also like to use the bold, italic and bolditalic forms of the font. Simply using \bold and \em does not work.
Here I see two possibilities:
(1) use the XeTeX parameters FakeSlant and FakeBold, which I have done in LaTeX. But I do not know how this is done in ConTeXt.
\starttext \definefontfeature[slanted] [slant=.2] \definefontfeature[boldened][extend=1.2] \definedfont[file:kedage-n*default] ಇಡೀ ಪ್ರಪಂಚದ ಜನರು ಒಂದೇ ಭಾಷೆಯನ್ನು ಮಾತಾಡುತ್ತಿ ದ್ದರು \definedfont[file:kedage-n*default,slanted] ಇಡೀ ಪ್ರಪಂಚದ ಜನರು ಒಂದೇ ಭಾಷೆಯನ್ನು ಮಾತಾಡುತ್ತಿ ದ್ದರು \definedfont[file:kedage-n*default,boldened] ಇಡೀ ಪ್ರಪಂಚದ ಜನರು ಒಂದೇ ಭಾಷೆಯನ್ನು ಮಾತಾಡುತ್ತಿ ದ್ದರು \definedfont[Serif*default] test \definedfont[Serif*default,boldened] test \stoptext
(2) use the other font files (Kedage-b.TTF etc.), and again I do not know how to do this.
see type-imp-* files (or give simplefonts a try) \definefontfeature[slanted] [slant=.2] \definefontfeature[boldened][extend=1.2] \starttypescript [serif] [kedage] [name] \definefontsynonym[Serif] [file:kedage-n] \definefontsynonym[SerifBold] [file:kedage-n][features={default,boldened}] \definefontsynonym[SerifItalic] [file:kedage-n][features={default,slanted}] \definefontsynonym[SerifBoldItalic][file:kedage-n][features={default,boldened,slanted}] \stoptypescript \starttypescript [kedage] \definetypeface[kedage][rm][serif][kedage][default] \definetypeface[kedage][mm][math] [modern][default] \stoptypescript \setupbodyfont[kedage] \starttext ಇದು ಕನ್ನಡದಲ್ಲಿದೆ. ಬಹಳ ರೋಚಕವಾದುದು, ಹೀಗೆ ಬರೆಯುವುದು ಎಂದು ಭಾವಿಸುತ್ತೇನೆ. {\bf ಈಗ ದಪ್ಪಗೆ} , ಮತ್ತು {\it ಕೊಂಕು} . ಇದು ಕನ್ನಡದಲ್ಲಿದೆ. ಬಹಳ ರೋಚಕವಾದುದು, ಹೀಗೆ ಬರೆಯುವುದು ಎಂದು ಭಾವಿಸುತ್ತೇನೆ.\par {\bf ಈಗ ದಪ್ಪಗೆ}\par , ಮತ್ತು\par {\it ಕೊಂಕು}\par .\par \stoptext Beware: there is bug in the luatex backend with respect to switching slanted/boldened that sometimes shows up. Hans ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | voip: 087 875 68 74 | www.pragma-ade.com | www.pragma-pod.nl -----------------------------------------------------------------
Hans,
Just now I tried your suggestions:
On Aug 8, 2013, at 13:51 , Hans Hagen
Here I see two possibilities:
(1) use the XeTeX parameters FakeSlant and FakeBold, which I have done in LaTeX. But I do not know how this is done in ConTeXt.
\starttext
\definefontfeature[slanted] [slant=.2] \definefontfeature[boldened][extend=1.2]
\definedfont[file:kedage-n*default] ...
The slanting and stretched effects come out nicely -- but the complex combinations of the signs are not supported as they are in XeTeX. XeTeX supports the combinations, but not the effects (see attachments hagen-kan-luatex.pdf and hagen-kan-xetex.pdf).
(2) use the other font files (Kedage-b.TTF etc.), and again I do not know how to do this.
\definefontfeature[slanted] [slant=.2] \definefontfeature[boldened][extend=1.2]
\starttypescript [serif] [kedage] [name] ...
No, unfortunately that one doesn't work so well (see attachment hagen-kan1-luatex.pdf).
in context, there is experimental support for devanagari \definedfont[file:kedage-n*devanagari-one]
but it all depends on what the font provides
Is there any way to find out the font's possibilities, if one doesn't have any technical documentation about it? Robert
On 8/8/2013 2:26 PM, Robert Zydenbos wrote:
Is there any way to find out the font's possibilities, if one doesn't have any technical documentation about it?
looking at tables in the font ... in fact, this font is somewhat messed up as it uses latin glyph names for kannaga ligatures concerning indic scripts ... this is an ongoing project: - kai eigner made the first setup, for devangari only but extending is a matter of adding more info - i cleaned up the code and optimized it (runs reasonable fast now) and and probably in the process broke some things anyhow, I added a few more entries in tables and get different output but also the impression that some glyphs get lost) so we're not there yet \definefontfeature[slanted] [slant=.2] \definefontfeature[boldened][extend=1.2] \definefontfeature[kannada][devanagari-one][script=knda] \starttypescript [serif] [kedage] [name] \definefontsynonym[Serif] [file:kedage-n][features={kannada}] \definefontsynonym[SerifBold] [file:kedage-n][features={kannada,default,boldened}] \definefontsynonym[SerifItalic] [file:kedage-n][features={kannada,default,slanted}] \definefontsynonym[SerifBoldItalic][file:kedage-n][features={kannada,default,boldened,slanted}] \stoptypescript \starttypescript [kedage] \definetypeface[kedage][rm][serif][kedage][default] \definetypeface[kedage][mm][math] [modern][default] \stoptypescript \setupbodyfont[kedage] \enabletrackers[otf.analyzing] \starttext ಇದು ಕನ್ನಡದಲ್ಲಿದೆ. ಬಹಳ ರೋಚಕವಾದುದು, ಹೀಗೆ ಬರೆಯುವುದು ಎಂದು ಭಾವಿಸುತ್ತೇನೆ. {\definedfont[file:kedage-n]ಇದು ಕನ್ನಡದಲ್ಲಿದೆ. ಬಹಳ ರೋಚಕವಾದುದು, ಹೀಗೆ ಬರೆಯುವುದು ಎಂದು ಭಾವಿಸುತ್ತೇನೆ.} % ಇದು ಕನ್ನಡದಲ್ಲಿದೆ. ಬಹಳ ರೋಚಕವಾದುದು, ಹೀಗೆ ಬರೆಯುವುದು ಎಂದು ಭಾವಿಸುತ್ತೇನೆ.\par % {\bf ಈಗ ದಪ್ಪಗೆ}\par % , ಮತ್ತು\par % {\it ಕೊಂಕು}\par % .\par \stoptext we can probably best move this discussion off list .. for instance we need to figure out what -- 0C82 anusvara -- 0C83 visarga -- 0CBD avagraha -- 0CD5 length mark -- 0CD6 ai length mark -- 0CE0 letter ll -- 0CE1 letter rr -- 0CE2 vowel sign l -- 0CE2 vowel sign ll -- 0CF1 sign -- 0CF2 sign -- OCE6 - OCEF digits does (maybe some has to be added to some tables). It's probably no big deal to get some more scripts working given what's already there ... more a matter of testing (and no, i can't read those scripts). Hans ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | voip: 087 875 68 74 | www.pragma-ade.com | www.pragma-pod.nl -----------------------------------------------------------------
On Thu, Aug 8, 2013 at 1:51 PM, Hans Hagen
forget about xetex, first check what context mkiv van do
Sorry to hijack this thread, but... Hans, it would be great to see Indic languages support in context/mkiv. At the moment I am typesetting, for example, in Malayalam using mkii+xetex. So I tested context (from TeXLive 2013 distribution) but found that the rendering is not correct. There are examples and screenshots here in an old thread: http://tex.stackexchange.com/questions/89118/context-xetex-texlive-2012-does... I am interested in helping, at least by testing, especially Malayalam support. Harfbuzz has excellent opentype support for Malayalam now, which could be used for comparing the rendering. -- Cheers, Rajeesh http://rajeeshknambiar.wordpress.com
participants (3)
-
Hans Hagen -
Rajeesh K Nambiar -
Robert Zydenbos