10

The Unicode characters ⸤ (U+2E24) and ⸥ (U+2E25) are sometimes used to groups words together in syntactic analysis. For example:

⸤The daughter ⸤of my brother⸥ ⸥ sees ⸤an orange tree⸥ ⸤with her binoculars⸥.

I am using LuaLaTeX and would like to keep using the Computer Modern font, but it doesn't seem to include these codepoints (they render as spaces). I can see several workarounds, but I don't know how to implement any of them:

  • Rotate tipa's \textcorner;
  • Truncate $\lfloor$ and $\rfloor$;
  • Steal those specific characters from a font that does support them.

Any of these works for me. Good to note: I'm using csquotes with \MakeOuterQuote{"}, so \char"2E24 as suggested here won't work.


For what it's worth, a very minimal example:

\documentclass{article}
\usepackage{csquotes}
\MakeOuterQuote{"}
\begin{document}
"⸤The daughter ⸤of my brother⸥ ⸥ sees ⸤an orange tree⸥ ⸤with her binoculars⸥."
\end{document}
3
  • 1
    I don't know why this got down voted. Ugly suggestion: you can use \char"231E" and \char"231F in math mode, if this result looks good enough.
    – mickep
    Mar 13, 2023 at 13:42
  • 1
    @mickep I didn't downvote but it could be that the lack of MWE caused the downvote.
    – Marijn
    Mar 13, 2023 at 13:45
  • 2
    The amssymb collection has four "corners": \ulcorner, \urcorner, \llcorner, and \lrcorner, of which the latter two are what you are looking for. They must be used in math mode. Mar 13, 2023 at 15:05

3 Answers 3

6

Borrow them.

\documentclass{article}
\usepackage{fontspec}
\usepackage{newunicodechar}

\newfontface{\lowbrackets}{Noto Sans}
\NewDocumentCommand{\blhb}{}{{\lowbrackets\symbol{"2E24}}}
\NewDocumentCommand{\brhb}{}{{\lowbrackets\symbol{"2E25}}}

\newunicodechar{⸤}{\blhb}
\newunicodechar{⸥}{\brhb}

\begin{document}

⸤The daughter ⸤of my brother⸥⸥ sees ⸤an orange tree⸥ ⸤with her binoculars⸥.

\end{document}

enter image description here

10
  • 4
    unicode-math already contains these symbols, called \llcorner and \lrcorner. Why rename them? Would be more reasonable to just make them available also as text. Mar 13, 2023 at 15:14
  • @barbarabeeton This answer is independent of unicode-math
    – egreg
    Mar 13, 2023 at 17:09
  • @barbarabeeton Besides, \llcorner is U+231E and \lrcorner is U+231F; in STIX2 they have a very different shape.
    – egreg
    Mar 13, 2023 at 17:17
  • 1
    @egreg -- if the shape in STIX2 is different, then STIX2 is wrong. I compiled the collection submitted to Unicode from which those characters were added. The model was the shapes in the font supported by amssymb, and that is what this question asks for. I'm happy to report an error in STIX2, if that is appropriate. Mar 13, 2023 at 20:22
  • 1
    @egreg -- Okay, thanks. Earlier comments mentioned U+231E and U+231F, and I haven't figured out how to determine from just an image what Unicode is meant. Knowing that U+2E24 and U+2E25 are what's actually meant makes a big difference. Sorry for the noise. Mar 13, 2023 at 21:46
4

Here an approach using lowered pmboxdraw characters. Still a bit too big maybe but an easy solution.

I added a bit of "poor man's kerning" to put the characters closer to the letters.

\documentclass{article}
\usepackage{newunicodechar}
\usepackage{graphicx}
\usepackage{pmboxdraw}
\usepackage{csquotes}
\MakeOuterQuote{"}
\newunicodechar{⸤}{\smash{\raisebox{-4pt}{\textSFii}\hskip-1pt}}
\newunicodechar{⸥}{\hskip-1pt\smash{\raisebox{-4pt}{\textSFiv}}}
\begin{document}
\noindent "⸤The daughter ⸤of my brother⸥ ⸥ sees ⸤an orange tree⸥ ⸤with her binoculars⸥."

\noindent "Furthermore the moon is made of cheese, that's why astronauts bounce on it."
\end{document}

enter image description here

Version without the fake kerning:

enter image description here

With clipping:

\usepackage{trimclip}
\newunicodechar{⸤}{\smash{\clipbox{0pt 0pt 0pt 2.5pt}{\raisebox{-4pt}{\textSFii}}\hskip-1pt}}
\newunicodechar{⸥}{\hskip-1pt\smash{\clipbox{0pt 0pt 0pt 2.5pt}{\raisebox{-4pt}{\textSFiv}}}}

enter image description here

2
  • They're definitely too tall; is there any way you could truncate them?
    – Mew
    Mar 13, 2023 at 18:19
  • @Mew you can cut the characters off from the top using \clipbox from the trimclip package, see edit.
    – Marijn
    Mar 14, 2023 at 9:07
3

Since the other answers are using newunicodechar, I figured I'd give it a shot myself by combining that with Barbara Beeton's suggestion. The output is below. Pro: the line thickness is the same on the vertical and the horizontal stroke. Con: the horizontal looks longer than the vertical, which I don't like.

Edit: the latter turns out to be an optical illusion. They are the same length zoomed-in. This doesn't really matter though; I still want the bracket to look taller than it is wide.

\documentclass{article}
\usepackage{csquotes}
\MakeOuterQuote{"}

\usepackage{newunicodechar}
\usepackage{unicode-math}
\newunicodechar{⸤}{\hskip-1.25pt$\llcorner$\hskip-1.25pt}
\newunicodechar{⸥}{\hskip-1.25pt$\lrcorner$}

\begin{document}
⸤The daughter ⸤of my brother⸥ ⸥ sees ⸤an orange tree⸥ ⸤with her binoculars⸥.
\end{document}

brackets

2
  • Actually, the horizontal and vertical arms of the corner glyphs are the same length; that's part of their definition. Any apparent difference is an optical illusion. But I have been corrected; what's wanted is apparently something different. Mar 13, 2023 at 21:49
  • @barbarabeeton Wow! You're right! That's quite the unexpected optical illusion. I stand corrected too.
    – Mew
    Mar 13, 2023 at 23:10

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .