Align corner brackets to the top and bottom when scaling #4200
Conversation
Could you provide some images comparing the old and the new behavior? Should make it easier to tell if that's a problem or not (e.g. if the previous behavior was unintended). |
|
I think there are images in #4188 but i wonder if it's the correct place to apply "fix". And i'm also surprised to see non of other tests fail |
Presumably because these are some of the few delimiters that aren't already vertically centered in most fonts. In fact, looking at http://mirrors.ctan.org/macros/unicodetex/latex/unicode-math/unimath-symbols.pdf they seem to be the only ones. I don't know the motivation behind centering delimiters, but I guess the safest option would be to only disable centering for these specific ones @bluebear94 ? |
|
This was done purposely, so the issue here is that these specific corner delimiters have to be handled in a different way. Perhaps check how LuaTeX/XeTeX deals with this? The same issue presumably appears there. |
|
Both LuaTeX and XeTeX have incorrect results with \usepackage{unicode-math}
\[\ulcorner c \urcorner \ulcorner \frac{a}{b} \urcorner\]
\[\left\ulcorner c \right\urcorner \left\ulcorner \frac{a}{b} \right\urcorner\]gives the following in LuaLaTeX:
and the following in XeLaTeX:
|
|
Okay, so I guess we'll have to come up with a solution ourselves. LuaTeX at least doesn't seem to center it even with \left and \right, even if they're not scaled at all @bluebear94. Unlike XeTeX, which seems to suffer from the same problem as typst. |
|
Can these specific delimiters be top and bottom aligned? |
bluebear94
changed the title
This broke scaling of the corner brackets, U+231C – U+231F. Hopefully this didn’t happen to be needed? Right?
bluebear94
force-pushed
the
mf/no-center-delims
branch
from
ba05ce5
to
d7931d6
Compare
| @@ -62,3 +62,7 @@ $ 1/(2 y (x) (2(3)) $ | |||
| // Test ignoring weak spacing immediately after the opening | |||
| // and immediately before the closing. | |||
| $ [#h(1em, weak: true)A(dif x, f(x) dif x)sum#h(1em, weak: true)] $ | |||
|
|
|||
| --- math-lr-corner-brackets --- | |||
| // Test positioning of U+231C to U+231F | |||
There was a problem hiding this comment.
could you add a comment here that they are supposed to look the same? the "vs" had me briefly thinking that they are supposed to look different.
could also rename it to issue-4188-lr-corner-brackets to link it to that issue
There was a problem hiding this comment.
I’ve changed the vs to an =.
See #4188
This broke scaling of the corner brackets, U+231C – U+231F. Hopefully this didn’t happen to be needed? Right?