Dear ConTeXt folks, using ConTeXt ver: 2011.05.18 22:26 MKIV fmt: 2011.5.27 int: english/english the following example produces different integral signs. --- minimal example --- \starttext Using the NEO keyboard layout several math symbols can be inserted directly like $∫$ or $∩$. But $\int$ and $∫$ look differently. \stoptext --- minimal example --- Please find all files attached. Thanks, Paul [1] http://neo-layout.org/
On Sun, Jun 12, 2011 at 15:06, Paul Menzel wrote:
Dear ConTeXt folks,
using
ConTeXt ver: 2011.05.18 22:26 MKIV fmt: 2011.5.27 int: english/english
the following example produces different integral signs.
--- minimal example --- \starttext Using the NEO keyboard layout several math symbols can be inserted directly like $∫$ or $∩$.
But $\int$ and $∫$ look differently. \stoptext --- minimal example ---
However the character ∫ looks the same as \intop and ∫\nolimits looks the same as \int. As to why \intop looks so weird, I have no idea. If you add \usetypescript[modern-base] \setupbodyfont[modern] it will work ok. Cambria Math also results in two different heights of the operator depending on whether limits are there or not. It might be a bug somewhere in virtual font setup of LM. In pdfTeX \int and \intop operators are positioned equally. Mojca
[My former message was not delivered and awaits moderator approval because it is bigger than 64 KB. I removed the PDF output using XITS.] Am Sonntag, den 12.06.2011, 23:52 +0200 schrieb Mojca Miklavec:
On Sun, Jun 12, 2011 at 15:06, Paul Menzel wrote:
ConTeXt ver: 2011.05.18 22:26 MKIV fmt: 2011.5.27 int: english/english
the following example produces different integral signs.
--- minimal example --- \starttext Using the NEO keyboard layout several math symbols can be inserted directly like $∫$ or $∩$.
But $\int$ and $∫$ look differently. \stoptext --- minimal example ---
I just wanted to bring this topic up again since it is still reproducible with ConTeXt MKIV 2011.11.25 21:29.
However the character ∫ looks the same as \intop and ∫\nolimits looks the same as \int.
Yes, it does.
As to why \intop looks so weird, I have no idea. If you add \usetypescript[modern-base] \setupbodyfont[modern] it will work ok. Cambria Math also results in two different heights of the operator depending on whether limits are there or not. It might be a bug somewhere in virtual font setup of LM. In pdfTeX \int and \intop operators are positioned equally.
Using XITS it does also work although in display math mode(?) ∫ and $\intop$ are not scaled. I use the following example. \setupbodyfont[xits] \setupinteraction[state=start] \useurl[context2008][http://wiki.contextgarden.net/Context_2008.04.10][][\ConTeXt\ version 4/10/2008] \starttext Using the NEO keyboard layout several math symbols can be inserted directly like $∫$ or $∩$. But $\int$ by \type{\int} and $∫$ by UTF-8 character look differently. As Mojca pointed out \type{\intop} is also not type set correctly: $\intop$. This suggests that it is a problem with the Latin Modern font. \type{\intop} was added in \from[context2008]. \blank • Using UTF-8 character: \startformula ∫_ℝ dx = ∞ \stopformula • Using UTF-8 character with \type{\nolimits}: \startformula ∫\nolimits_ℝ dx = ∞ \stopformula • Using \type{\intop}: \startformula \intop_ℝ dx = ∞ \stopformula • Using \type{\intop\nolimits} does not work. \startformula \intop\nolimits_ℝ dx = ∞ \stopformula • Using \type{\int}: \startformula \int_ℝ dx = ∞ \stopformula • Using \type{\int_\limits_a}: \startformula \int\limits_a dx = ∞ \stopformula \blank \CONTEXT\ \contextmark\ \contextversion \stoptext Is there a chance to fix `\intop` in Latin Modern (and XITS). Would it be a good idea to map ∫ to $\int$ instead of `\intop`? Do people use `\intop` a lot? Thanks, Paul PS: Although this seems font related, some more information just in case. ∫ has the Unicode number(?) #222B [1] and `xev` returns the following. KeyPress event, serial 33, synthetic NO, window 0x4c00001, root 0x102, subw 0x0, time 340829925, (244,21), root:(256,134), state 0x4000, keycode 39 (keysym 0x8bf, integral), same_screen YES, XLookupString gives 3 bytes: (e2 88 ab) "∫" XmbLookupString gives 3 bytes: (e2 88 ab) "∫" XFilterEvent returns: False KeyRelease event, serial 33, synthetic NO, window 0x4c00001, root 0x102, subw 0x0, time 340829989, (244,21), root:(256,134), state 0x4000, keycode 39 (keysym 0x8bf, integral), same_screen YES, XLookupString gives 3 bytes: (e2 88 ab) "∫" XFilterEvent returns: False [1] http://unicode.org/charts/nameslist/n_2200.html
apart from this IMHO very useful change (mapping to \int), maybe we should create a unicode-math module which replaces some commands with symbols. i find myself always using this for chemical and mathematical documents: \catcode`→=\active \def→{\xrightarrow} \catcode`↔=\active \def↔{\xleftrightarrow} \catcode`√=\active \def√{\sqrt} PS: i recently posted two questions to this list which i got no replies to:“simplefonts: math in headers” and“headers: numberconversion=something or \definestructureconversionset” do you see them or did they get lost in transmission?
Am Sonntag, den 11.12.2011, 13:28 +0100 schrieb Philipp A.:
apart from this IMHO very useful change (mapping to \int), maybe we should create a unicode-math module which replaces some commands with symbols.
Do you mean symbols by commands?
i find myself always using this for chemical and mathematical documents:
\catcode`→=\active \def→{\xrightarrow} \catcode`↔=\active \def↔{\xleftrightarrow} \catcode`√=\active \def√{\sqrt}
Sorry for being a noob. I thought for the first three that is somehow done already. If not, such a mechanism would indeed be useful.
PS: i recently posted two questions to this list which i got no replies to:“simplefonts: math in headers” and“headers: numberconversion=something or \definestructureconversionset”
do you see them or did they get lost in transmission?
The message were delivered to the subscribers. You can also check that in the archive [1]. We just have to be patient. I hope Wolfgang is not sick/ill and is doing fine by being on vacation or something similar. PPS: Do you also use the Neo layout [4]? [1] http://www.ntg.nl/pipermail/ntg-context/2011/thread.html [2] http://www.ntg.nl/pipermail/ntg-context/2011/063723.html [3] http://www.ntg.nl/pipermail/ntg-context/2011/063739.html [4] http://neo-layout.org/
On Sun, 11 Dec 2011, Philipp A. wrote:
apart from this IMHO very useful change (mapping to \int), maybe we should create a unicode-math module which replaces some commands with symbols. i find myself always using this for chemical and mathematical documents:
\catcode`→=\active \def→{\xrightarrow} \catcode`↔=\active \def↔{\xleftrightarrow} \catcode`√=\active \def√{\sqrt}
Have a look at char-def.lua. All the symbol mappings are defined there. Please feel free to complete the table (it is a boring and time consuming process, and I have only managed to cover about half of the table). Regarding your mappings, IIRC -> etc are mapped to \rightarrow etc, (not \xrightarrow). We can talk about including style variants for the mappings. Aditya
wow, there is really much magic going on in context. one reason why i love it. i once wrote a script that showed some nice little gui for a similar file (i.e. xkb/symbols/??) that displayed the characters on a grid similar to the keyboard layout and allowed displaying of the different layers. i used it to customize the standard ubuntu keyboard layout to have some much-used symbols, but haven’t really got around to try switching to another layout than quertz (as it would tie me even stronger to my own pc and would take significant time to master) maybe it would be nice to have a similar program that displays the mappings in the char-def.lua since i can’t search for the symbols directly, have no overview and editing would involve looking up the codes or names in a char map, finding it in the file, and so on. xrightarrow and friends are nice, because they adjust their length to the over-and underscripts and work without parameters, too. so $A → B →{woo} C$ would work flawlessly and look nice both in code and the document.
participants (4)
-
Aditya Mahajan -
Mojca Miklavec -
Paul Menzel -
Philipp A.