Tag Info

Hot answers tagged

10

With kind help from @cfr, you've established that: The font does contain ligature glyphs. fontspec can't make use of AAT ligature information even if it's available. The font omits OpenType feature data necessary for automatic ligature support. So what can you do? You can either add the missing mapping data to the font, or create an external OpenType ...


5

The source of the problem The problem is caused by the fact the Junicode font has true proportional old style numbers, whereas some other fonts have, both, fixed width old style numbers and proportional old style numbers. The linguex package sets an internal length that is a multiple of the smallest width digit in the current font. The smallest width ...


5

Using || for producing a double em-dash is abusing the system. The fact that | produces an em-dash, when the output font encoding is OT1, is purely incidental and should never be relied upon. The “right” way in TeX/LaTeX for obtaining an em-dash is by typing --- and this works also with fontspec, provided the Ligatures=TeX option, that's activated by ...


5

After a few false starts, it became clear that || is the result of a ConTeXt convention, and for reasons more interesting than mere caprice. From ConTeXt's lang-mis.mkiv file: One of TEX's strong points in building paragraphs is the way hyphenations are handled. Although for real good hyphenation of non-english languages some extensions to the ...


5

It seems a bug somewhere. A temporary workaround: \documentclass{article} \usepackage{fontspec} %\setmainfont{Linux Libertine O} \newcommand{\hdigit}[1]{% \accent\string"02C6 #1% } \begin{document} \^5 causes a collision! And 5^^^^0302 too. But \hdigit{5} doesn't. \end{document}


4

This is a known problem. As Joseph Wright wrote on the miktex mailing list: Will Robertson knows about this and it will be fixed soon. For the present \usepackage{expl3} \ExplSyntaxOn \cs_new_eq:NN \str_case:nnn \str_case:nnF \cs_new_eq:NN \str_case_x:nnn \str_case_x:nnF \cs_new_eq:NN \int_case:nnn \int_case:nnF \ExplSyntaxOff ...


4

You can assess glyphs by name: \documentclass{article} \usepackage{fontspec} \setmainfont{Linux Libertine O} %for comparision: \setsansfont[RawFeature=+aalt]{Linux Libertine O} \begin{document} Lessingstraße \XeTeXglyph\the\XeTeXglyphindex "germandbls.sc" \relax \sffamily Lessingstraße \end{document} The main problem is to find out the correct name ...


4

This works in both xelatex and lualatex and sets the document in Pagella for me: \documentclass[11 pt, a4paper]{article} \usepackage[T1]{fontenc} \usepackage{fontspec} \ExplSyntaxOn \NewDocumentCommand{\onlywarniffontnotfound}{} { \font_suppress_not_found_error: \msg_redirect_name:nnn {fontspec}{font-not-found}{warning} } \ExplSyntaxOff ...


3

Here package polyglossia change the definition of \arabic to get Indian form of numbers ٣ ٢ ١ ... You can modify the format of footnote numbering by redefining the command \thefootnote \documentclass{memoir} \usepackage{polyglossia} \setmainlanguage{arabic} \setotherlanguage{english} ...


2

Define your colour shade of black first and then add that to the font specification using the Color (or Colour) key-value: \documentclass{article} \usepackage{fontspec,lipsum,xcolor} \begin{document} \fontspec[Scale=0.9]{Trebuchet MS} \lipsum[1] \definecolor{myblack}{rgb}{.7,.7,.7}% 30% black \addfontfeature{Color=myblack}% \lipsum[1] \end{document} ...


2

There's another way to do this: just declare a new font family for the tight spaced text. For some reason the simple version works with the latin script but fails with the Arabic script. I don't know why. To solve the Arabic script problem I've used the trick described here: Scope of \addfontfeature. With this solution, you don't need a \nottight command ...


2

I get wrong output in both cases. At least in the english text they are more font family involved so you need a bit more sophisticated font changing. Try this instead: \documentclass[11pt,a4paper]{article} \usepackage{fontspec} \usepackage{polyglossia} \newfontfamily\khmerfont[Script=Khmer,Scale=0.9]{Khmer OS} \setdefaultlanguage[numerals=khmer]{khmer} ...


2

It is rather unclear what you actually want. But here two suggestions: You could still use SourceSerifPro as main font and fake a slanted font: \documentclass[]{scrartcl} \usepackage{fontspec} \defaultfontfeatures{Ligatures=TeX} \setmainfont{SourceSerifPro}[ ItalicFont = SourceSerifPro, ItalicFeatures={FakeSlant=0.2}] \begin{document} This is a ...


1

Rather than WordSpace, you should use \spaceskip, which respects grouping. The problems is that WordSpace=... sets \fontdimen parameters and these assignments are global (affecting the current font only, though). \documentclass{scrartcl} \usepackage{fontspec} \setmainfont{Arial Unicode MS} \newfontfamily\arabicfont[Script=Arabic,Numbers=Arabic]{Arial ...


1

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % Two Column One Page Curriculum Vitae % LaTeX Template % Version 1.1 (24/1/13) % % This template has been downloaded from: % http://www.LaTeXTemplates.com % % Original author: % Alessandro (The CV Inn) % % IMPORTANT: THIS TEMPLATE NEEDS TO BE COMPILED WITH XeLaTeX % % This template uses several fonts not included ...


1

I think it might be better to wrap do and undo the \addfontfeature in the same command. The problem is figuring out what you needs to be undone. If you set WordSpace to -2.0, then reversing it is achieved by subsequently "setting" it to -0.5 (my rudimentary math skills suggest: -2 * -0.5 = 1...). Of course, the true values are not whole numbers, but you get ...


1

Imho it is a problem with the endash in your font: \documentclass{article} \usepackage{fontspec} \newfontfamily\greekfont[Script=Greek]{GFS Porson} \begin{document} \greekfont Οὑτωσ \fbox{—}τοίνυν, Οὑτωσ \fbox{–}τοίνυν, \end{document}



Only top voted, non community-wiki answers of a minimum length are eligible