Bug 118737 - (displaystyle) Add support for displaystyle
 (displaystyle) Summary: Add support for displaystyle
 Product: Classification: Status: RESOLVED FIXED WebKit Unclassified MathML Version: 528+ (Nightly build) Platform: All All Importance: P2 Normal Assigned To: Nobody Keywords: 99620 133845 155542 155638 155639 155714 3251 mathml-in-mathjax 124907 Show dependency tree / graph

 See Also: Reported: 2013-07-16 09:49 PDT by Frédéric Wang (:fredw) Modified: 2016-07-11 11:57 PDT (History) CC List: 3 users (show) alex cfleizach mrobinson

Attachments
testcase (1.00 KB, text/html)
2013-07-16 09:49 PDT, Frédéric Wang (:fredw)
no flags Details

 Note You need to log in before you can comment on or make changes to this bug.
 Frédéric Wang (:fredw) 2013-07-16 09:49:06 PDT Created attachment 206790 [details] testcase and currently display the formula as a block or inline box. However, they should also set the initial value of displaystyle to true and false respectively. The displaystyle value can also be set via an explicit displaystyle attribute and is reset to false by some elements (like mfrac, scripts, roots etc). When displaystyle=true, the operators marked largeop (either via an explicit attribute or as computed from the operator dictionary) should be drawn larger. When displaystyle=false, the operators marked movablelimits (either via an explicit attribute or as computed from the operator dictionary) should have its under/over limits drawn as sub/sup. See the testcase, where a sum symbol is used (largeop and movablelimits per the operator dictionary). See also bucket 3 of the MathML Acid3 tests: http://fred-wang.github.io/AcidTestsMathML/acid3/d http://fred-wang.github.io/AcidTestsMathML/acid3/description.html See chapter 3 of the MathML spec for details: http://www.w3.org/TR/2010/REC-MathML3-20101021/chapter3.html Frédéric Wang (:fredw) 2016-07-11 11:57:21 PDT Displaystyle maybe be involved in other cases (e.g. in scriptlevel incrementation, see bug 49309) but I think we now support the most important cases (largeop, movablelimits and some choice of MATH constants). So I guess we can close this bug.