[Xy-pic] Spacing problems with triangles

[Ross Moore]

Hi Jurgen,

On 30/03/2004, at 11:03 PM, Jürgen Koslowski wrote:

> Hello,
>
> I encountered an annoying spacing problem when displaying triangular
> diagrams.  The top corner is much too close to the preceeding text, in

Yes, your are right.
There is a problem whereby the full size of the diagram is lost
when the \twocell (and other 2-cell macros) are used.

> case of a short line even protrudes into that line.  So far I have
> used invisible entries to get around this problem, is there a better
> way?  Enclosing the diagram into \vcenter{...} does not help.
>


I've devised a fix.
Try including the follow code in the document-preamble,
after the Xy-pic package has been loaded.


\makeatletter
\gdef\xymerge at MinMax{}
\xydef@\twocell{\hbox\bgroup\xysave at MinMax\@twocell}
\xydef@\uppertwocell{\hbox\bgroup\xysave at MinMax\@uppertwocell}
\xydef@\lowertwocell{\hbox\bgroup\xysave at MinMax\@lowertwocell}
\xydef@\compositemap{\hbox\bgroup\xysave at MinMax\@compositemap}
\xydef@\twocelll#1#{\hbox\bgroup\xysave at MinMax\xy@\save\save@\@twocelll{ 
#1}}

\xydef@\xysave at MinMax{\xdef\xymerge at MinMax{%
   \noexpand\ifdim\X at max<\the\X at max \X at max=\the\X at max\noexpand\fi
   \noexpand\ifdim\X at min>\the\X at min \X at min=\the\X at min\noexpand\fi
   \noexpand\ifdim\Y at max<\the\Y at max \Y at max=\the\Y at max\noexpand\fi
   \noexpand\ifdim\Y at min>\the\Y at min \Y at min=\the\Y at min\noexpand\fi
  }}
\xydef@\drop at Twocell{\boxz@ \xymerge at MinMax}


\xydef@\twocell at DONE{%
  \edef\tmp@{\egroup
   \X at min=\the\X at min \X at max=\the\X at max
   \Y at min=\the\Y at min \Y at max=\the\Y at max}\tmp@
  \L at c=\X at c \advance\L at c-\X at min \R at c=\X at max \advance\R at c-\X at c
  \D at c=\Y at c \advance\D at c-\Y at min \U at c=\Y at max \advance\U at c-\Y at c
  \ht\z@=\U at c \dp\z@=\D at c \dimen@=\L at c \advance\dimen@\R at c \wdz@=\dimen@
  \computeLeftUpness@
  \setboxz at h{\kern-\X at p \raise-\Y at c\boxz@ }%
  \dimen@=\L at c \advance\dimen@\R at c \wdz@=\dimen@ \ht\z@=\U at c \dp\z@=\D at c
  \Edge at c={\rectangleEdge}\Invisible at false \Hidden at false
  \edef\Drop@@{\noexpand\drop at Twocell
   \noexpand\def\noexpand\Leftness@{\Leftness@}%
   \noexpand\def\noexpand\Upness@{\Upness@}}%
  \edef\Connect@@{\noexpand\connect at Twocell
   \noexpand\ifdim\X at max<\the\X at max \X at max=\the\X at max\noexpand\fi
   \noexpand\ifdim\X at min>\the\X at min \X at min=\the\X at min\noexpand\fi
   \noexpand\ifdim\Y at max<\the\Y at max \Y at max=\the\Y at max\noexpand\fi
   \noexpand\ifdim\Y at min>\the\Y at min \Y at min=\the\Y at min\noexpand\fi }%
  \xymerge at MinMax
}%
\makeatother


> Here is an example:
>
> \documentclass[11pt]{article}
> \usepackage[arrow,2cell,line,dvips,ps]{xy}
>
> %\CompileMatrices
> \UseAllTwocells
>
> \setlength{\textwidth}{16 cm}
> \begin{document}
>
>   If for a 2-cell
> \[
> \xy
>   ;<50 pt,0 pt>:
>   (0.0,0)              *+{A}="A",
>   (0.5,\halfrootthree) *+{B}="B",
>   (1.0,0)              *+{C}="C",
>   "A";"C",{\twocell<\omit>{<-3>{\ \phi}}},
>   \ar_{t}"A";"C"
>   \ar^{r}"A";"B"
>   \ar^{s}"B";"C"
> \endxy
> \]

This example should work properly now.

Please try the above fix, and report any problems that you notice.


Cheers

	Ross


>   pasting at $s$ is bijective, we call $\langle s,\phi\rangle$ a right
>   extension of $t$ along $r$.
>
>   Even if the line preceeding the diagram is longer, the placement
>   is better but we still have a problem.  There seems to be too little
>   space above the diagram:
> \[
> \xy
>   ;<50 pt,0 pt>:
>   (0.0,0)              *+{A}="A",
>   (0.5,\halfrootthree) *+{B}="B",
>   (1.0,0)              *+{C}="C",
>   "A";"C",{\twocell<\omit>{<-3>{\ \phi}}},
>   \ar_{t}"A";"C"
>   \ar^{r}"A";"B"
>   \ar^{s}"B";"C"
> \endxy
> \]
>   I don't want to put in extra invisible elements to get the spacing  
> right.
>
> \end{document}
>
> -- Juergen
>
> -- 
> Juergen Koslowski               If I don't see you no more on this  
> world
> ITI, TU Braunschweig               I'll meet you on the next one
> koslowj at iti.cs.tu-bs.de               and don't be late!
> http://www.iti.cs.tu-bs.de/~koslowj      Jimi Hendrix (Voodoo Child,  
> SR)
> _______________________________________________
> xy-pic mailing list
> http://tug.org/mailman/listinfo/xy-pic
>
------------------------------------------------------------------------
Ross Moore                                         ross at maths.mq.edu.au
Mathematics Department                             office: E7A-419
Macquarie University                               tel: +61 +2 9850 8955
Sydney, Australia                                  fax: +61 +2 9850 8114
------------------------------------------------------------------------