![](https://secure.gravatar.com/avatar/d0e94fb08d4620cfbd2f8f59a7868637.jpg?s=120&d=mm&r=g)
On 11/3/22 13:27, Henning Hraban Ramm via ntg-context wrote:
If it works for you, then it’s a problem of my installation, otherwise a regression in current LMTX...
No, I get the same error message, this looks promising but doesn't work. And at the risk of sounding grumpy, but all of this is too much manual intervention; I would like to have a completely automated script. So I'll show what I have and add a few comments: local Feature_Table = { } local Parsed_Features = { } local My_Font = environment.arguments["font"] local Font_Features = os.resultof("otfinfo -f " .. My_Font) for Feature_Sigla, Feature_Name in string.gmatch(Font_Features, "(%S-)\t(.-)\n") do Feature_Table[Feature_Sigla] = Feature_Name end COMMENTS Expectation: this should return a list of all features and their descriptions as a lua table. If you run the file as context test.cld --font="/path/to/font" this works, and you get a table Feature_Table={ ["aalt"]="Access All Alternates", ["c2sc"]="Small Capitals From Capitals"} etc. CODE for name, _ in pairs(Feature_Table) do Context_Name = "f:" .. name Parsed_Features[name] = Context_Name context.definefontfeature( { Context_Name }, { name="yes" } ) end COMMENTS This should produce a lua table with ConTeXt fontfeatures and define those features. The table appears to be correct: Parsed_Features={ ["aalt"]="f:aalt", ["c2sc"]="f:c2sc"} etc. So I expected that this should be equivalent to \definefontfeature [f:aalt] [aalt=yes] and return a list of defined features. CODE local Test_String = [[ ABCDEFGHIJKLMNOPQRSTUVWXYZ abcdefghijklmnopqrstuvwxyz 1234567890 1/4 2/8 3/9 ffi ft fi fl ffl ]] local Font_Name = os.resultof("otfinfo -a " .. My_Font) context.definefontfamily( { "Test" }, { "serif" }, { Font_Name }, {} ) context.setupbodyfont{ "Test, 12pt" } COMMENTS Again, this works; the font is set as bodyfont. CODE context.starttext() for name, feature in pairs(Feature_Table) do context.start () context.addfeature { Parsed_Features[name] } context.midaligned (feature) context.bold (Parsed_Features[name]) context.par () context.blank{ "medium" } context(Test_String) context.par () context.stop () end context.stoptext() COMMENTS And this doesn't work: the names are picked up, but the font features are not defined; the sample text does not reflect any features. So this is what I've got... maybe one of the wizards can help. Thomas