# Table columns are clipped to column width in HTML-CSS but not SVG #807

Closed
opened this issue May 7, 2014 · 9 comments

Projects
None yet
2 participants
Member

### pkra commented May 7, 2014

 http://www.w3.org/Math/testsuite/build/main/Presentation/TablesAndMatrices/mtable/mtable2-full.xhtml renders quite differently for me. On Chrome 34, the bottom-center cell overlaps with bottom-left On FF 29, the bottom-center cell "underlaps", i.e., disappears at the edge of the cell.

Member

### dpvc commented Aug 19, 2014

 I don't see any overlapping with the bottom left in SVG output. Can you post a screen shot? The "underlapping" is actually clipping to the column width. I suppose that could be allowed to overlap the next column to the right. The real issue is that line breaking isn't being done within the columns, but that's really a separate problem.
Member Author

### pkra commented Aug 19, 2014

 I don't see any overlapping with the bottom left in SVG output. Can you post a screen shot? Here's a screenshot for overlap -- happens on FF 31, Chrome 36, ubuntu 14.04. The real issue is that line breaking isn't being done within the columns, but that's really a separate problem. Right. Should I split this into another issue?
Member

### dpvc commented Aug 19, 2014

 Sorry, I would have called that the bottom-center overlapping the bottom-right not the bottom-left. So the difference is that the HTML-CSS is clipped and the SVG isn't. I can probably get the HTML-CSS not to clip, which (I suspect) is the better behavior. Should I split this into another issue? Line breaking in the columns isn't likely to be done any time soon. You can start an issue if you want just to keep it in mind.
Member Author

### pkra commented Aug 19, 2014

 Ah, sorry for being confused. So the difference is that the HTML-CSS is clipped and the SVG isn't. Right. ("clipping" must have eluded me at time of writing...) I can probably get the HTML-CSS not to clip, which (I suspect) is the better behavior. To be honest, I dislike both but admittedly without clipping, people don't miss the content. Is overflow: scroll an option?
Member

### dpvc commented Aug 19, 2014

 Is overflow: scroll an option? Not without some substantial rewriting at the moment. The clipping is actually being done at the column level, not the cell, currently, and I don't think you want the whole column to scroll. Also, scrolling requires block-level elements, and I don't think the cells are stored in block-level elements currently. Finally, I'm not sure I can do that in SVG anyway (at least not without implementing it myself).
Member Author

### pkra commented Aug 19, 2014

 Alright -- ouch on all accounts ;-) no-clipping it is.

Open

### dpvc pushed a commit to dpvc/MathJax that referenced this issue Sep 14, 2014

 Don't clip left or right-hand sides, only top and bottom (which are w… 
…hat need clipping to hide exctra height/depth, and prevent the positioning spans from overlaying the material above). Resolves issue mathjax#807, and may hekp issue mathjax#646.
 0a04a6c 

Closed

Member

### dpvc commented Sep 14, 2014

 The issue807 branch of my fork of MathJax should resolve this.

### dpvc pushed a commit to dpvc/MathJax that referenced this issue Nov 2, 2014

 Merge branch 'issue807' into develop. Issue mathjax#807. 
 771fc39 
Member

### dpvc commented Nov 13, 2014

 ==> Merged (Nov. 2)

### dpvc pushed a commit to mathjax/MathJax-test that referenced this issue Jan 17, 2015

 Add test for clipping of table columns in HTML-CSS (mathjax/MathJax#807) 
 fe7b89d 

Member

### dpvc commented Jan 17, 2015

 ==> In Testsuite MathMLToDisplay/Presentation/TablesAndMatrices/issue807.html