Hi Hans, everybody there’s a bug report concerning one of the Noto fonts: https://github.com/lualatex/luaotfload/issues/292 Fonts are freely available: https://www.google.com/get/noto/ Quick link to the example translated to Context: https://bitbucket.org/phg/lua-la-tex-tests/src/tip/context/cnt-urdu.tex Those features are what’s applied by Luaotfload by default for Arabic. Best, Philipp
On 11/27/2015 10:56 PM, Philipp Gesang wrote:
Hi Hans, everybody
there’s a bug report concerning one of the Noto fonts:
https://github.com/lualatex/luaotfload/issues/292
Fonts are freely available: https://www.google.com/get/noto/ Quick link to the example translated to Context:
https://bitbucket.org/phg/lua-la-tex-tests/src/tip/context/cnt-urdu.tex
Those features are what’s applied by Luaotfload by default for Arabic.
something cursive but anyhow, there us no urdu language (dflt has to be used) ----------------------------------------------------------------- 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 Sat, Nov 28, 2015 at 3:56 PM, Hans Hagen
something cursive but anyhow, there us no urdu language (dflt has to be used)
I reported the bug originally. I changed the fontspec directive in my test file to not specify the language. I got a better output, but still pretty bad. Details w/ screenshot: https://github.com/lualatex/luaotfload/issues/292#issuecomment-160283333 Deepak PS: Btw fontspec does not complain about the language. In XeTeX, my experience has been that if a language is not present, it logs a message to stdout. I am not sure if fontspec on LuaLaTeX has the same behavior, but I did not see a message complaining about language. Also, here is the otfinfo output: $> otfinfo -s NotoNastaliqUrdu-Regular.ttf arab Arabic arab.ARA Arabic/Arabic arab.FAR Arabic/Farsi arab.KSH Arabic/Kashmiri arab.SND Arabic/Sindhi arab.URD Arabic/Urdu dflt <unknown script> latn Latin
On 11/28/2015 12:49 PM, Deepak Jois wrote:
On Sat, Nov 28, 2015 at 3:56 PM, Hans Hagen
wrote: something cursive but anyhow, there us no urdu language (dflt has to be used)
I reported the bug originally. I changed the fontspec directive in my test file to not specify the language. I got a better output, but still pretty bad.
Details w/ screenshot: https://github.com/lualatex/luaotfload/issues/292#issuecomment-160283333
Deepak
PS: Btw fontspec does not complain about the language. In XeTeX, my experience has been that if a language is not present, it logs a message to stdout. I am not sure if fontspec on LuaLaTeX has the same behavior, but I did not see a message complaining about language.
Also, here is the otfinfo output:
$> otfinfo -s NotoNastaliqUrdu-Regular.ttf arab Arabic arab.ARA Arabic/Arabic arab.FAR Arabic/Farsi arab.KSH Arabic/Kashmiri arab.SND Arabic/Sindhi arab.URD Arabic/Urdu dflt <unknown script> latn Latin
well, only for gsub, not for gpos where only arab/dflt is defined ... okay we could have another heuristic for that but of course it then becomes harder to *not* use some feature ----------------------------------------------------------------- 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 Sun, Nov 29, 2015 at 7:06 AM, Deepak Jois
On Sun, Nov 29, 2015 at 12:32 AM, Hans Hagen
wrote: well, only for gsub, not for gpos where only arab/dflt is defined
Are you saying that is a bug with the font? If so, I could report it, for what it’s worth.
Tried to look into the spec regarding GPOS lookups: https://www.microsoft.com/typography/otspec/gpos.htm <quote> … To access GPOS information, clients should use the following procedure: 1. Locate the current script in the GPOS ScriptList table. 2. If the language system is known, search the script for the correct LangSys table; otherwise, use the script's default language system (DefaultLangSys table). … </quote>
On 11/29/2015 4:38 AM, Deepak Jois wrote:
On Sun, Nov 29, 2015 at 7:06 AM, Deepak Jois
wrote: On Sun, Nov 29, 2015 at 12:32 AM, Hans Hagen
wrote: well, only for gsub, not for gpos where only arab/dflt is defined
Are you saying that is a bug with the font? If so, I could report it, for what it’s worth.
Tried to look into the spec regarding GPOS lookups: https://www.microsoft.com/typography/otspec/gpos.htm
<quote> …
To access GPOS information, clients should use the following procedure:
1. Locate the current script in the GPOS ScriptList table.
2. If the language system is known, search the script for the correct LangSys table; otherwise, use the script's default language system (DefaultLangSys table).
… </quote>
yes but for testing and control i don't like hard coded defaults so that has to go via an option; a next version will have a 'default' option that when set to true will fallback on dflt .. and tex is about control (btw, features and what should be on/off are the messiest part of opentype ... and often very inconsistent so ..) 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 11/29/2015 2:36 AM, Deepak Jois wrote:
On Sun, Nov 29, 2015 at 12:32 AM, Hans Hagen
wrote: well, only for gsub, not for gpos where only arab/dflt is defined
Are you saying that is a bug with the font? If so, I could report it, for what it’s worth.
when it comes to features and its implementation all bugs are also features (so if someone puts something positioning related in a gsub that is quite possible and i've seen it) in general i think that features / script / languages are often defined a bit sloppy so i always look at what a font really supports (same for names in fonts ... i only trust filenames) 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 11/27/2015 10:56 PM, Philipp Gesang wrote:
Hi Hans, everybody
there’s a bug report concerning one of the Noto fonts:
https://github.com/lualatex/luaotfload/issues/292
Fonts are freely available: https://www.google.com/get/noto/ Quick link to the example translated to Context:
https://bitbucket.org/phg/lua-la-tex-tests/src/tip/context/cnt-urdu.tex
Those features are what’s applied by Luaotfload by default for Arabic.
a next version provides autoscript=position|substitution|yes and autolanguage=position|substitution|yes (no uploads now, probably when luatex 0.86 releases) 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 -----------------------------------------------------------------
participants (3)
-
Deepak Jois
-
Hans Hagen
-
Philipp Gesang