Hi, are the delimiters \lgroup and \rgroup supported in MkIV? This \starttext $$\bigg\lgroup\dots\bigg\rgroup$$ \stoptext is ok in MkII (ConTeXt ver: 2011.01.06) but when compiled with MkIV (ConTeXt ver: 2010.07.30 11:35 and ver: 2010.07.30 11:35) the delimiters are not shown. (By the way, the same holds true for \rmoustache and \lmoustache.) Am I missing something? Searching the list archive and the contextgarden unfortunately didn't bring up in anything. Thanks in advance, Daniel
On Sat, 8 Jan 2011, Daniel Schopper wrote:
Am I missing something?
No. The math mappings of virtual fonts are incomplete. The symbols should work correctly with unicode math fonts like cambria, asana, and xits. But, it appears that lgroup and rgroup are missing from xits (Khaled?) and lmoustache and rmoustache are missing from asana! With Cambria there is a bug with scaling: \setupbodyfont[cambria] \enabletrackers[math.analyzing] \usemodule[fnt-25] \starttext \showmathfontcharacters[][0x23B0] \showmathfontcharacters[][0x3014] $$\bigg\lgroup\dots\bigg\rgroup$$ $$\left\lgroup\sum\right\rgroup$$ $$\lgroup\sum\rgroup$$ $$\bigg\lmoustache\dots\bigg\rmoustache$$ $$\left\lmoustache\sum\right\rmoustache$$ $$\lmoustache\sum\rmoustache$$ \stoptext @Hans, Taco, and Mojca: Do you know why the delimiters are not scaling properly? Do you know how to add them to the LM virtual fonts? I tried, but could not get a working solution. I can add the correct mappings to large-to-small table in math-vfu. But adding the mappings for lgroup and rgroup in tex-mr table, and adding lmoustache and rmoustache to tex-ex table does not appear to work. Apply the attached diff file and see what happens with the example above. I don't know how to proceed. BTW, where are the math virtual fonts defined!? I could not find any use of mathematics.makefont. Aditya
On Sat, Jan 08, 2011 at 12:52:38PM -0500, Aditya Mahajan wrote:
There are, but context is looking for the wring character; it is looking for U+3014 (which I think is a CJK bracket and does not look like \lgroup either) while it should be looking for U+27EE (whose Unicode annotation precisely says "lgroup"): \setupbodyfont[xits] \starttext $$\bigg\lgroup\dots\bigg\rgroup$$ \def\lgroup{\Udelimiter "4 "0 "27EE} \def\rgroup{\Udelimiter "4 "0 "27EF} $$\bigg\lgroup\dots\bigg\rgroup$$ \stoptext It was not extensible though, the version in git now grow, but starts with a relatively large size, check and tell me if it should be smaller. Regards, Khaled -- Khaled Hosny Arabic localiser and member of Arabeyes.org team Free font developer
On Sat, Jan 08, 2011 at 09:35:09PM +0200, Khaled Hosny wrote:
BTW, why luatex no longer prints the "missing character" messages in the log? I was using it to find what characters context is looking for, also now it is near impossible to know if the font is missing some needed character without reviewing every word by eye since luatex (context?) does not show a "missing character" glyph in such cases, or is there some replacement I've missed? Regards, Khaled -- Khaled Hosny Arabic localiser and member of Arabeyes.org team Free font developer
On Sat, 8 Jan 2011, Khaled Hosny wrote:
Hmm ... 27EE does not exist: http://unicode.org/charts/PDF/U32-27C0.pdf I believe that I used 3014 based on http://www-sop.inria.fr/apics/tralics/doc-l.html#cmd-lgroup What source do you use for the symbols? Aditya
Hmm ... 27EE does not exist: http://unicode.org/charts/PDF/U32-27C0.pdf
You're looking at the chart for Unicode 3.2 that is almost 10 years old... The latest version (now 6.0) is always available as http://www.unicode.org/charts/PDF/U27C0.pdf and has the character as Khaled says. Arthur
On Sat, Jan 08, 2011 at 02:59:35PM -0500, Aditya Mahajan wrote:
I opened Gnome's Character Map and searched for lgroup :) (but I've seen it before, so I know it exists, I just don't recall the code point). Now that you asked, I checked Unicode website and they have "Miscellaneous Mathematical Symbols-A": http://unicode.org/charts/PDF/U27C0.pdf Regards, Khaled -- Khaled Hosny Arabic localiser and member of Arabeyes.org team Free font developer
On Sat, 8 Jan 2011, Khaled Hosny wrote:
That was the same range that I had looked at, but somehow that pdf did not have 27EC-27EF. Neither does char-def.lua. So, I guess that these are newer additions. I will add them to char-def.lua and correct the mappings for lgroup and rgroup. Aditya
On Sun, Jan 09, 2011 at 01:13:25AM +0000, Arthur Reutenauer wrote:
That is why I asked Hans while ago how char-def.lua was generated, from what version of Unicode and whether it is updated/updatable or not; I was worried that it will soon be outdated, which seems to be the case now. Regards, Khaled -- Khaled Hosny Arabic localiser and member of Arabeyes.org team Free font developer
On Sun, 9 Jan 2011, Khaled Hosny wrote:
I don't know how it was generated. Mojca and I occasionally update the file by adding the names of math symbols. If the latest unicode symbol list is available in text format, it should not be too difficult to check if the table in char-def has any missing entries.
In this case, I just added these four missing characters by hand. Aditya
On Sun, Jan 09, 2011 at 01:59:20PM -0500, Aditya Mahajan wrote:
Unicode data are always available as plain text files: http://www.unicode.org/Public/UNIDATA/ -- Khaled Hosny Arabic localiser and member of Arabeyes.org team Free font developer
On Sun, Jan 9, 2011 at 19:59, Aditya Mahajan wrote:
It is definitely updatable, but not necessary regularly updated :) :) :) The code that generated the list cannot be used without modifications since there have been many manual changes on the old list, but it is definitely doable. Mojca
On 9-1-2011 8:58, Mojca Miklavec wrote:
I have a couple of scripts lying around and will have a look one of these days. Hans ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | voip: 087 875 68 74 | www.pragma-ade.com | www.pragma-pod.nl -----------------------------------------------------------------
On 9-1-2011 7:59, Aditya Mahajan wrote:
In this case, I just added these four missing characters by hand.
The first versions of the table were a combination of data present already in context and some unicode tables from the web (using scripts) but much in the math department was added manually as i found out that tables (like stix) were too flawed and had bugs so were useless for our purpose. We could do some occasional update by checking diffs from older to newer unicodes, not that I expect too many changes. Hans ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | voip: 087 875 68 74 | www.pragma-ade.com | www.pragma-pod.nl -----------------------------------------------------------------
There are, but context is looking for the wring character; it is looking for U+3014 (which I think is a CJK bracket
Indeed. It's one of the angle brackets that one finds in many CJK fonts.
while it should be looking for U+27EE (whose Unicode annotation precisely says "lgroup"):
Note that those annatations in the code charts are actually the Unicode 1.0 names of the characters, where they differ from the Unicode 1.1 names (when Unicode was unified with ISO 10646 and adopted some of its names). They should not, in general, be understood as the name of this character is some other standard, for example. But they are of course a good hint of what the character represents. Arthur
On Sat, Jan 08, 2011 at 08:29:17PM +0000, Arthur Reutenauer wrote:
I usually look for long description, it can sometimes be very helpful on what is the intended use of the characters (names are not very accurate, especially early characters which are often misnamed). Regards, Khaled -- Khaled Hosny Arabic localiser and member of Arabeyes.org team Free font developer
On Sat, 8 Jan 2011, Khaled Hosny wrote:
Looks OK to me, but I am not really an expert on the shapes of the symbols. Can you also make lmoustache and rmoustache extensible? Hans, I am attaching the patch for char-def.lua that corrects the location of lgroup and rgroup. I still do not know how to fix this for lm virtual math font. Aditya
participants (6)
-
Aditya Mahajan -
Arthur Reutenauer -
Daniel Schopper -
Hans Hagen -
Khaled Hosny -
Mojca Miklavec