cgal/Packages/Arrangement/doc_tex/basic/Arrangement_2_ref/intro.tex

299 lines
14 KiB
TeX

% +------------------------------------------------------------------------+
% | Reference manual chapter: intro.tex (Arrangement)
% +------------------------------------------------------------------------+
% |
% | Package: arr (Arrangement_2)
% |
% +------------------------------------------------------------------------+
%+---------------------------------------------------------------------------80
%| update log
%|
%| 21 Jun 2000 - Shai Hirsch
%| Separated from arr.tex (See previous changes in change log there).
%|
%+---------------------------------------------------------------------------80
% +========================================================================+
% Introduction
% +========================================================================+
%\clearpage
%\section{Reference Pages for 2D Arrangements}
\chapter{2D Arrangements}
\subsection*{Introduction}
\ccRefLabel{arr_ref_intro}
The arrangement class holds a planar map and a hierarchy tree.
Vertices, halfedges and faces of the arrangement derive from
those of the planar map (with the additional functionality of
the arrangement),
in the same way that the vertices, halfedges and faces of the topological map
class derive from those of the Dcel class
%in a way similar to what is done between the
%topological map and the Dcel class
(see Chapter\ref{I1_ChapterTopologicalMap}).
The hierarchy tree is implemented using the \ccc{In_place_list}
class (see the chapter on STL Extensions in the Support Library Manual). Every level of a curve
hierarchy is a list of tree nodes.The \ccc{Curve_node} and
\ccc{Edge_node} of
the hierarchy derive from \ccc{Subcurve_node}. This enables the
polymorphic structure of the tree. The \ccc{Subcurve_node} is
derived from the \ccc{Base_node} which is a template parameter
of the arrangement. This enables the addition of attributes to the nodes
of the hierarchy tree by adding them inside the \ccc{Base_node}.
%Since the arrangement holds a planar map its traits requirements
%are a superset of the requirements from the planar map traits.
\subsection*{Arrangement Dcel}
The arrangement is parametrized with the interface class \ccc{Dcel}.
The \ccc{Dcel} is a container class that defines the underlying
combinatorial data structure used by the planar map. We define the
concept \ccc{ArrangementDcel_2} \lcTex{(\ccRefPage{ArrangementDcel_2})}
where the requirements for a \ccc{Dcel} class are defined.
As part of the \ccc{Dcel} we define requirements for its vertex, halfedge and
face constituents. If we consider these constituents as (informally)
subconcepts of the \ccc{Dcel} concept then we have the following models for
its constituents.
%-----------------------------------------------------
The \ccStyle{Arr_2_vertex_base<Point>} is a model for the
Dcel vertex subconcept,
the \mbox{\ccc{Arr_2_halfedge_base<Base_node>}} is a
model for the Dcel halfedge subconcept
and the \mbox{\ccc{Arr_2_face_base}} is a
model for the Dcel face subconcept.
\ccInclude{CGAL/Arr_2_bases.h}
The \ccStyle{Arr_2_default_dcel<Traits>} is a model of the Dcel concept
described above. Its template parameter is the traits class.
It is a wrapper for \ccStyle{Pm_dcel} instantiated
with \mbox{\ccc{Arr_2_vertex_base<Traits::Point>}},\\
\mbox{\ccc{Arr_2_halfedge_base<Arr_base_node<Traits::X_curve> >}} and
\mbox{\ccStyle{Arr_2_face_base}}.
\ccInclude{CGAL/Arr_2_default_dcel.h}
\subsection*{Models for an Arrangement Traits Class}
%-----------------------------------------------
We supply a traits class for segments that use the types
and predicates of the CGAL kernel, a traits class for
circle arcs as well as a traits class for polyline curves.
Since the requirements of the arrangement traits
are a superset of the requirements of the planar map traits,
any traits class that works with arrangements can work with
planar maps as well.
The different traits classes are described at the end of this chapter.
\begin{ccAdvanced}
\subsection*{I/O functions}
The \ccc{Arrangement} package supports I/O functions, which include reading
an arrangement representation from
the standard input or writing it to the standard output,
and also sending an arrangement to a graphic stream.
As already mentioned at chapter \ccRefPage{Pm_Ref_intro}, the motivation for using
I/O functions is not only to be able to draw the arrangement
to a window for instance, but also to be capable to save an arrangement
in a text file by writing it and reloading it from a text file by reading it.
Reading an arrangement from the standard input or printing it to the
standard output may be done simply with the \ccc{Extractor} (\ccc{ >>
}) and \ccc{Insertor} (\ccc{ << }) operators defined for
\ccc{Arrangement_2}, respectively.
\ccInclude{CGAL/IO/Arr_iostream.h}
The ability of sending the \ccc{Arrangement}
into a graphic stream as \ccStyle{leda_window}, Postscript file or
Geomview is also provided, users simply have to apply the Insertor
operator on the graphic stream and their arrangement instance.
Users of I/O functions for the arrangement package are required to define I/O
operators for the curves defined in their \ccc{Traits} classes.
When using \ccc{Traits} classes in which this operators are already defined
(as \ccc{Segment Traits} or \ccc{Circle Traits} ) the operator definition is not obligated,
however using \ccc{Traits} classes as the \ccc{Polyline Traits} will force the user to define
I/O operators on his polyline curve.
The \ccc{Arrangement} class data consists of the induced planar map and the
obtained hierarchy tree. Hence, the data sent to the output stream or
read from an input stream should contain both parts.
The format of the planar map part is as specified in the Planar Map
refernce pages (\ccRefPage{Pm_Ref_intro}).
The format of the hierarchy tree is specified below.
The format of the output file is defined in a way the reading function
will construct the arrangement efficiently.
The induced planar map is constructed efficiently as specified in the Planar Map refernce pages (\ccRefPage{Pm_Ref_intro}), and the hierarchy tree is also
constructed efficiently by directly accessing its parts and updating them.
When constructing an arrangement from an input stream, no use of its
insertion functions is performed.
Consequently, the reading function constructs the arrangement very
efficiently, and hence users who would like to save their arrangement and reload it have to construct their
arrangement by the insertion functions only once.
After saving the arrangement to a text file it can be reloaded very efficiently when needed.
When users would like to read their arrangement from the standard input or print it into
the standard output, they may simply use the \ccc{Extractor} (\ccc{ >> }) and \ccc{Insertor} (\ccc{ << })
operators defined for \ccc{Arrangement} respectively.
If users add attributes to their arrangement components, reading (resp. writing) arrangement would
be done by inheriting the class \ccStyle{Arr_file_scanner} (resp. \ccStyle{Arr_file_writer} )
and overriding all the relevant function for scanning (resp. writing) the arrangement components.
After the definition of the inherited class users have to call the function
\ccStyle{read} of \ccc{Arrangement} (resp. the global function \ccStyle{write_arr} ) with
the inherited class as a parameter.
The ability of sending the \ccc{Arrangement} into a graphic stream as \ccStyle{leda_window},
Postscript file or Geomview is also provided,
users simply have to use the Insertor operator operated on the graphic stream and their arrangement.
When users would like to add attributes to their arrangement components
and send their arrangement to a graphic stream,
they have to inherit the class \ccStyle{Pm_drawer} and then call the global function \ccStyle{draw_pm}
with this class and their arrangement as parameters.
The function \ccStyle{Pm_drawer} is used both for \ccc{Planar map} and \ccc{Arrangement}, since drawing
an arrangement is defined as drawing only its planar map part.
\paragraph{Format}
\ccRefLabel{Arr_IO_format}
As mentioned above, the format of the planar map part is as specified
in the Planar Map refernce pages (\ccRefPage{Pm_Ref_intro}).
Here, we are representing the format of the hierarchy tree.
Generally, the format represents the curve nodes list of the \ccc{arrangement}.
Each component of one curve node is compound from all the subcurves and edge nodes
this curve node contains. This data is associative with geometric information and some
topological information in order to be able to update the hierarchy tree efficiently.
The format is detailed below:
\begin{enumerate}
\item The data begins with a line of one integer specifying the number
of curve nodes the hierarchy tree has.
\item The list of curve nodes:
each curve node has the following format:
\begin{enumerate}
\item Its associative curve.
\item An integer specifying the number of levels the curve node has.
\item The list of all subcurves levels. Each described level goes as follows:
\begin{enumerate}
\item An integer specifying the number of subcurve nodes in the current level.
\item List of subcurve nodes belong to the current level.
Each subcurve consist of the following:
\begin{enumerate}
\item Pointers to subcurves (edge nodes, if the next level is the last)
of lower level which are given as the begin and past end indices
of children subcurve nodes (edge nodes).
\item The curve associative with the subcurve node.
\end{enumerate}
\end{enumerate}
\item An integer specifying the number of edge nodes.
\item List of edge nodes. Each described edge node goes as follows:
\begin{enumerate}
\item A pointer to an overlapping edge node.
This pointer is represented by two indices, the first is an index to a curve node
and the second is an index to an edge node defined in that curve node.
\item An index to the associated halfedge.
\item The associated curve.
\end{enumerate}
\end{enumerate}
\end{enumerate}
Other rules concerning the format are detailed in the Planar Map
refernce pages (\ccRefPage{Pm_Ref_intro}).
\ccExample {of usage in I/O functions}
\label{ARR_sec:example11}
The input of the program is a text file presenting the \ccc{Arrangement}:
\ccIncludeExampleCode{Arrangement_2/example11.cin}
The example above presents an arrangement containing two segments intersect in their interior.
The segments coordinates are {\em (0,0)}, {\em (1,1)} and {\em (0,1)}, {\em (1,0)}.
The \ccc{Arrangement} instance that was used to produce this example was templated
with the \ccStyle{Arr_segment_exact_traits} class, which was templated with the representation
class \ccStyle{Cartesian<Quotient<int> >}.
The first part of the input represents the planar map our arrangement contains, and hence its format
is identically the same as \ccRefPage{Pm_IO_format}. It can be seen that the planar map described here
represents all the vertcies and halfedges obtained by the disjoint subcurves of the arrangement.
The next part of the input file presents the hierarchy tree of our arrangement.
This presentation begins with the number 2, indicating there are two
curve nodes in our arrangement. The list of curve nodes follows.
The first curve node begins withh its associated curve {\em (0,0)}, {\em (1,1)}, it has 0 levels since
it does not contain any subcurve nodes (only curve nodes and edge nodes).
The number of edge nodes the curve {\em (0,0)}, {\em (1,1)} induces is two,
and the curves these two edge nodes are associated with are
{\em (0,0)}, {\em (1/2,1/2)} and \em{(1/2,1/2)}, \em{(1,1)} correspondingly.
The indices of each edge node to an overlapping edge node are pointing to the edge node
itself since there are no overlapping in the input file.
Finally, the indices of the halfedge indicates the halfedge our edge node is associated with.
For example the first edge node induced by the curve {\em (0,0)}, {\em (1,1)} is associated
with the first halfedge presented in the halfedges list. The reader may verify that this halfedge
holds the same curve the edge node holds.
The current format may not be comfortable for a user to read, because the usage of indices.
There is a possibility to define a verbose format which contains instead of
indices, the components themselves, and hence the user has the option to have a format
which is easy for human to read.
This format cannot be scanned by the reading functions of \ccc {Arrangement}.
\ccIncludeExampleCode{Arrangement_2/example11.C}
The output is the \ccc{Arrangement} written in both formats, non verbose and verbose.
\ccIncludeExampleCode{Arrangement_2/example11.cout}
For more examples see chapter \ccRefPage{Pm_Ref_intro}.
More details are given in sections
\ccc{Arr_file_scanner<Arrangement>}\lcTex{
(\ccRefPage{CGAL::Arr_file_scanner<Arrangement>})},
\ccc{Arr_file_writer<Arrangement>}\lcTex{
(\ccRefPage{CGAL::Arr_file_writer<Arrangement>})}
\end{ccAdvanced}
\begin{ccTexOnly}
\subsection*{Concepts}
\ccRefConceptPage{ArrangementBaseNode_2}\\
\ccRefConceptPage{ArrangementDcel_2}\\
\ccRefConceptPage{ArrangementDcelHalfedge_2}\\
\ccRefConceptPage{ArrangementTraits_2}\\
\subsection*{Classes}
\ccRefIdfierPage{CGAL::Arrangement_2<Dcel,Traits,Base_node>}\\
\ccRefIdfierPage{CGAL::Arr_base_node<Curve>}\\
\ccRefIdfierPage{CGAL::Arrangement_2<Dcel,Traits,Base_node>::Subcurve_node}\\
\ccRefIdfierPage{CGAL::Arrangement_2<Dcel,Traits,Base_node>::Curve_node}\\
\ccRefIdfierPage{CGAL::Arrangement_2<Dcel,Traits,Base_node>::Edge_node}\\
\ccRefIdfierPage{CGAL::Arrangement_2<Dcel,Traits,Base_node>::Halfedge}\\
\ccRefIdfierPage{CGAL::Arr_segment_exact_traits<R>}\\
\ccRefIdfierPage{CGAL::Arr_leda_segment_exact_traits}\\
\ccRefIdfierPage{CGAL::Arr_circles_real_traits<NT>}\\
\ccRefIdfierPage{CGAL::Arr_segment_circle_traits<NT>}\\
\ccRefIdfierPage{CGAL::Segment_circle_2<NT>} \\
\ccRefIdfierPage{CGAL::Arr_polyline_traits<R, Container>}\\
\ccRefIdfierPage{CGAL::Arr_leda_polyline_traits<Container>}\\
\ccRefIdfierPage{CGAL::Arr_file_scanner<Arrangement>}\\
\ccRefIdfierPage{CGAL::Arr_file_writer<Arrangement>}\\
\subsection*{Functions}
\ccRefIdfierPage{CGAL::write_arr}\\
\end{ccTexOnly}