mirror of
https://github.com/autc04/Retro68.git
synced 2024-12-14 10:32:23 +00:00
482 lines
16 KiB
HTML
482 lines
16 KiB
HTML
<?xml version="1.0" encoding="US-ASCII"?>
|
|
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
|
|
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
|
|
<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
|
|
<head>
|
|
<title>The GNU Implementation of java.awt.geom.FlatteningPathIterator</title>
|
|
<meta name="author" content="Sascha Brawer" />
|
|
<style type="text/css"><!--
|
|
td { white-space: nowrap; }
|
|
li { margin: 2mm 0; }
|
|
--></style>
|
|
</head>
|
|
<body>
|
|
|
|
<h1>The GNU Implementation of FlatteningPathIterator</h1>
|
|
|
|
<p><i><a href="http://www.dandelis.ch/people/brawer/">Sascha
|
|
Brawer</a>, November 2003</i></p>
|
|
|
|
<p>This document describes the GNU implementation of the class
|
|
<code>java.awt.geom.FlatteningPathIterator</code>. It does
|
|
<em>not</em> describe how a programmer should use this class; please
|
|
refer to the generated API documentation for this purpose. Instead, it
|
|
is intended for maintenance programmers who want to understand the
|
|
implementation, for example because they want to extend the class or
|
|
fix a bug.</p>
|
|
|
|
|
|
<h2>Data Structures</h2>
|
|
|
|
<p>The algorithm uses a stack. Its allocation is delayed to the time
|
|
when the source path iterator actually returns the first curved
|
|
segment (either <code>SEG_QUADTO</code> or <code>SEG_CUBICTO</code>).
|
|
If the input path does not contain any curved segments, the value of
|
|
the <code>stack</code> variable stays <code>null</code>. In this quite
|
|
common case, the memory consumption is minimal.</p>
|
|
|
|
<dl><dt><code>stack</code></dt><dd>The variable <code>stack</code> is
|
|
a <code>double</code> array that holds the start, control and end
|
|
points of individual sub-segments.</dd>
|
|
|
|
<dt><code>recLevel</code></dt><dd>The variable <code>recLevel</code>
|
|
holds how many recursive sub-divisions were needed to calculate a
|
|
segment. The original curve has recursion level 0. For each
|
|
sub-division, the corresponding recursion level is increased by
|
|
one.</dd>
|
|
|
|
<dt><code>stackSize</code></dt><dd>Finally, the variable
|
|
<code>stackSize</code> indicates how many sub-segments are stored on
|
|
the stack.</dd></dl>
|
|
|
|
<h2>Algorithm</h2>
|
|
|
|
<p>The implementation separately processes each segment that the
|
|
base iterator returns.</p>
|
|
|
|
<p>In the case of <code>SEG_CLOSE</code>,
|
|
<code>SEG_MOVETO</code> and <code>SEG_LINETO</code> segments, the
|
|
implementation simply hands the segment to the consumer, without actually
|
|
doing anything.</p>
|
|
|
|
<p>Any <code>SEG_QUADTO</code> and <code>SEG_CUBICTO</code> segments
|
|
need to be flattened. Flattening is performed with a fixed-sized
|
|
stack, holding the coordinates of subdivided segments. When the base
|
|
iterator returns a <code>SEG_QUADTO</code> and
|
|
<code>SEG_CUBICTO</code> segments, it is recursively flattened as
|
|
follows:</p>
|
|
|
|
<ol><li>Intialization: Allocate memory for the stack (unless a
|
|
sufficiently large stack has been allocated previously). Push the
|
|
original quadratic or cubic curve onto the stack. Mark that segment as
|
|
having a <code>recLevel</code> of zero.</li>
|
|
|
|
<li>If the stack is empty, flattening the segment is complete,
|
|
and the next segment is fetched from the base iterator.</li>
|
|
|
|
<li>If the stack is not empty, pop a curve segment from the
|
|
stack.
|
|
|
|
<ul><li>If its <code>recLevel</code> exceeds the recursion limit,
|
|
hand the current segment to the consumer.</li>
|
|
|
|
<li>Calculate the squared flatness of the segment. If it smaller
|
|
than <code>flatnessSq</code>, hand the current segment to the
|
|
consumer.</li>
|
|
|
|
<li>Otherwise, split the segment in two halves. Push the right
|
|
half onto the stack. Then, push the left half onto the stack.
|
|
Continue with step two.</li></ul></li>
|
|
</ol>
|
|
|
|
<p>The implementation is slightly complicated by the fact that
|
|
consumers <em>pull</em> the flattened segments from the
|
|
<code>FlatteningPathIterator</code>. This means that we actually
|
|
cannot “hand the curent segment over to the consumer.”
|
|
But the algorithm is easier to understand if one assumes a
|
|
<em>push</em> paradigm.</p>
|
|
|
|
|
|
<h2>Example</h2>
|
|
|
|
<p>The following example shows how a
|
|
<code>FlatteningPathIterator</code> processes a
|
|
<code>SEG_QUADTO</code> segment. It is (arbitrarily) assumed that the
|
|
recursion limit was set to 2.</p>
|
|
|
|
<blockquote>
|
|
<table border="1" cellspacing="0" cellpadding="8">
|
|
<tr align="center" valign="baseline">
|
|
<th></th><th>A</th><th>B</th><th>C</th>
|
|
<th>D</th><th>E</th><th>F</th><th>G</th><th>H</th>
|
|
</tr>
|
|
<tr align="center" valign="baseline">
|
|
<th><code>stack[0]</code></th>
|
|
<td>—</td>
|
|
<td>—</td>
|
|
<td><i>S<sub>ll</sub>.x</i></td>
|
|
<td>—</td>
|
|
<td>—</td>
|
|
<td>—</td>
|
|
<td>—</td>
|
|
<td>—</td>
|
|
</tr>
|
|
<tr align="center" valign="baseline">
|
|
<th><code>stack[1]</code></th>
|
|
<td>—</td>
|
|
<td>—</td>
|
|
<td><i>S<sub>ll</sub>.y</i></td>
|
|
<td>—</td>
|
|
<td>—</td>
|
|
<td>—</td>
|
|
<td>—</td>
|
|
<td>—</td>
|
|
</tr>
|
|
<tr align="center" valign="baseline">
|
|
<th><code>stack[2]</code></th>
|
|
<td>—</td>
|
|
<td>—</td>
|
|
<td><i>C<sub>ll</sub>.x</i></td>
|
|
<td>—</td>
|
|
<td>—</td>
|
|
<td>—</td>
|
|
<td>—</td>
|
|
<td>—</td>
|
|
</tr>
|
|
<tr align="center" valign="baseline">
|
|
<th><code>stack[3]</code></th>
|
|
<td>—</td>
|
|
<td>—</td>
|
|
<td><i>C<sub>ll</sub>.y</i></td>
|
|
<td>—</td>
|
|
<td>—</td>
|
|
<td>—</td>
|
|
<td>—</td>
|
|
<td>—</td>
|
|
</tr>
|
|
<tr align="center" valign="baseline">
|
|
<th><code>stack[4]</code></th>
|
|
<td>—</td>
|
|
<td><i>S<sub>l</sub>.x</i></td>
|
|
<td><i>E<sub>ll</sub>.x</i>
|
|
= <i>S<sub>lr</sub>.x</i></td>
|
|
<td><i>S<sub>lr</sub>.x</i></td>
|
|
<td>—</td>
|
|
<td><i>S<sub>rl</sub>.x</i></td>
|
|
<td>—</td>
|
|
<td>—</td>
|
|
</tr>
|
|
<tr align="center" valign="baseline">
|
|
<th><code>stack[5]</code></th>
|
|
<td>—</td>
|
|
<td><i>S<sub>l</sub>.y</i></td>
|
|
<td><i>E<sub>ll</sub>.x</i>
|
|
= <i>S<sub>lr</sub>.y</i></td>
|
|
<td><i>S<sub>lr</sub>.y</i></td>
|
|
<td>—</td>
|
|
<td><i>S<sub>rl</sub>.y</i></td>
|
|
<td>—</td>
|
|
<td>—</td>
|
|
</tr>
|
|
<tr align="center" valign="baseline">
|
|
<th><code>stack[6]</code></th>
|
|
<td>—</td>
|
|
<td><i>C<sub>l</sub>.x</i></td>
|
|
<td><i>C<sub>lr</sub>.x</i></td>
|
|
<td><i>C<sub>lr</sub>.x</i></td>
|
|
<td>—</td>
|
|
<td><i>C<sub>rl</sub>.x</i></td>
|
|
<td>—</td>
|
|
<td>—</td>
|
|
</tr>
|
|
<tr align="center" valign="baseline">
|
|
<th><code>stack[7]</code></th>
|
|
<td>—</td>
|
|
<td><i>C<sub>l</sub>.y</i></td>
|
|
<td><i>C<sub>lr</sub>.y</i></td>
|
|
<td><i>C<sub>lr</sub>.y</i></td>
|
|
<td>—</td>
|
|
<td><i>C<sub>rl</sub>.y</i></td>
|
|
<td>—</td>
|
|
<td>—</td>
|
|
</tr>
|
|
<tr align="center" valign="baseline">
|
|
<th><code>stack[8]</code></th>
|
|
<td><i>S.x</i></td>
|
|
<td><i>E<sub>l</sub>.x</i>
|
|
= <i>S<sub>r</sub>.x</i></td>
|
|
<td><i>E<sub>lr</sub>.x</i>
|
|
= <i>S<sub>r</sub>.x</i></td>
|
|
<td><i>E<sub>lr</sub>.x</i>
|
|
= <i>S<sub>r</sub>.x</i></td>
|
|
<td><i>S<sub>r</sub>.x</i></td>
|
|
<td><i>E<sub>rl</sub>.x</i>
|
|
= <i>S<sub>rr</sub>.x</i></td>
|
|
<td><i>S<sub>rr</sub>.x</i></td>
|
|
<td>—</td>
|
|
</tr>
|
|
<tr align="center" valign="baseline">
|
|
<th><code>stack[9]</code></th>
|
|
<td><i>S.y</i></td>
|
|
<td><i>E<sub>l</sub>.y</i>
|
|
= <i>S<sub>r</sub>.y</i></td>
|
|
<td><i>E<sub>lr</sub>.y</i>
|
|
= <i>S<sub>r</sub>.y</i></td>
|
|
<td><i>E<sub>lr</sub>.y</i>
|
|
= <i>S<sub>r</sub>.y</i></td>
|
|
<td><i>S<sub>r</sub>.y</i></td>
|
|
<td><i>E<sub>rl</sub>.y</i>
|
|
= <i>S<sub>rr</sub>.y</i></td>
|
|
<td><i>S<sub>rr</sub>.y</i></td>
|
|
<td>—</td>
|
|
</tr>
|
|
<tr align="center" valign="baseline">
|
|
<th><code>stack[10]</code></th>
|
|
<td><i>C.x</i></td>
|
|
<td><i>C<sub>r</sub>.x</i></td>
|
|
<td><i>C<sub>r</sub>.x</i></td>
|
|
<td><i>C<sub>r</sub>.x</i></td>
|
|
<td><i>C<sub>r</sub>.x</i></td>
|
|
<td><i>C<sub>rr</sub>.x</i></td>
|
|
<td><i>C<sub>rr</sub>.x</i></td>
|
|
<td>—</td>
|
|
</tr>
|
|
<tr align="center" valign="baseline">
|
|
<th><code>stack[11]</code></th>
|
|
<td><i>C.y</i></td>
|
|
<td><i>C<sub>r</sub>.y</i></td>
|
|
<td><i>C<sub>r</sub>.y</i></td>
|
|
<td><i>C<sub>r</sub>.y</i></td>
|
|
<td><i>C<sub>r</sub>.y</i></td>
|
|
<td><i>C<sub>rr</sub>.y</i></td>
|
|
<td><i>C<sub>rr</sub>.y</i></td>
|
|
<td>—</td>
|
|
</tr>
|
|
<tr align="center" valign="baseline">
|
|
<th><code>stack[12]</code></th>
|
|
<td><i>E.x</i></td>
|
|
<td><i>E<sub>r</sub>.x</i></td>
|
|
<td><i>E<sub>r</sub>.x</i></td>
|
|
<td><i>E<sub>r</sub>.x</i></td>
|
|
<td><i>E<sub>r</sub>.x</i></td>
|
|
<td><i>E<sub>rr</sub>.x</i></td>
|
|
<td><i>E<sub>rr</sub>.x</i></td>
|
|
<td>—</td>
|
|
</tr>
|
|
<tr align="center" valign="baseline">
|
|
<th><code>stack[13]</code></th>
|
|
<td><i>E.y</i></td>
|
|
<td><i>E<sub>r</sub>.y</i></td>
|
|
<td><i>E<sub>r</sub>.y</i></td>
|
|
<td><i>E<sub>r</sub>.y</i></td>
|
|
<td><i>E<sub>r</sub>.y</i></td>
|
|
<td><i>E<sub>rr</sub>.y</i></td>
|
|
<td><i>E<sub>rr</sub>.x</i></td>
|
|
<td>—</td>
|
|
</tr>
|
|
<tr align="center" valign="baseline">
|
|
<th><code>stackSize</code></th>
|
|
<td>1</td>
|
|
<td>2</td>
|
|
<td>3</td>
|
|
<td>2</td>
|
|
<td>1</td>
|
|
<td>2</td>
|
|
<td>1</td>
|
|
<td>0</td>
|
|
</tr>
|
|
<tr align="center" valign="baseline">
|
|
<th><code>recLevel[2]</code></th>
|
|
<td>—</td>
|
|
<td>—</td>
|
|
<td>2</td>
|
|
<td>—</td>
|
|
<td>—</td>
|
|
<td>—</td>
|
|
<td>—</td>
|
|
<td>—</td>
|
|
</tr>
|
|
<tr align="center" valign="baseline">
|
|
<th><code>recLevel[1]</code></th>
|
|
<td>—</td>
|
|
<td>1</td>
|
|
<td>2</td>
|
|
<td>2</td>
|
|
<td>—</td>
|
|
<td>2</td>
|
|
<td>—</td>
|
|
<td>—</td>
|
|
</tr>
|
|
<tr align="center" valign="baseline">
|
|
<th><code>recLevel[0]</code></th>
|
|
<td>0</td>
|
|
<td>1</td>
|
|
<td>1</td>
|
|
<td>1</td>
|
|
<td>1</td>
|
|
<td>2</td>
|
|
<td>2</td>
|
|
<td>—</td>
|
|
</tr>
|
|
</table>
|
|
</blockquote>
|
|
|
|
<ol>
|
|
|
|
<li>The data structures are initialized as follows.
|
|
|
|
<ul><li>The segment’s end point <i>E</i>, control point
|
|
<i>C</i>, and start point <i>S</i> are pushed onto the stack.</li>
|
|
|
|
<li>Currently, the curve in the stack would be approximated by one
|
|
single straight line segment (<i>S</i> – <i>E</i>).
|
|
Therefore, <code>stackSize</code> is set to 1.</li>
|
|
|
|
<li>This single straight line segment is approximating the original
|
|
curve, which can be seen as the result of zero recursive
|
|
splits. Therefore, <code>recLevel[0]</code> is set to
|
|
zero.</li></ul>
|
|
|
|
Column A shows the state after the initialization step.</li>
|
|
|
|
<li>The algorithm proceeds by taking the topmost curve segment
|
|
(<i>S</i> – <i>C</i> – <i>E</i>) from the stack.
|
|
|
|
<ul><li>The recursion level of this segment (stored in
|
|
<code>recLevel[0]</code>) is zero, which is smaller than
|
|
the limit 2.</li>
|
|
|
|
<li>The method <code>java.awt.geom.QuadCurve2D.getFlatnessSq</code>
|
|
is called to calculate the squared flatness.</li>
|
|
|
|
<li>For the sake of argument, we assume that the squared flatness is
|
|
exceeding the threshold stored in <code>flatnessSq</code>. Thus, the
|
|
curve segment <i>S</i> – <i>C</i> – <i>E</i> gets
|
|
subdivided into a left and a right half, namely
|
|
<i>S<sub>l</sub></i> – <i>C<sub>l</sub></i> –
|
|
<i>E<sub>l</sub></i> and <i>S<sub>r</sub></i> –
|
|
<i>C<sub>r</sub></i> – <i>E<sub>r</sub></i>. Both halves are
|
|
pushed onto the stack, so the left half is now on top.
|
|
|
|
<br /> <br />The left half starts at the same point
|
|
as the original curve, so <i>S<sub>l</sub></i> has the same
|
|
coordinates as <i>S</i>. Similarly, the end point of the right
|
|
half and of the original curve are identical
|
|
(<i>E<sub>r</sub></i> = <i>E</i>). More interestingly, the left
|
|
half ends where the right half starts. Because
|
|
<i>E<sub>l</sub></i> = <i>S<sub>r</sub></i>, their coordinates need
|
|
to be stored only once, which amounts to saving 16 bytes (two
|
|
<code>double</code> values) for each iteration.</li></ul>
|
|
|
|
Column B shows the state after the first iteration.</li>
|
|
|
|
<li>Again, the topmost curve segment (<i>S<sub>l</sub></i>
|
|
– <i>C<sub>l</sub></i> – <i>E<sub>l</sub></i>) is
|
|
taken from the stack.
|
|
|
|
<ul><li>The recursion level of this segment (stored in
|
|
<code>recLevel[1]</code>) is 1, which is smaller than
|
|
the limit 2.</li>
|
|
|
|
<li>The method <code>java.awt.geom.QuadCurve2D.getFlatnessSq</code>
|
|
is called to calculate the squared flatness.</li>
|
|
|
|
<li>Assuming that the segment is still not considered
|
|
flat enough, it gets subdivided into a left
|
|
(<i>S<sub>ll</sub></i> – <i>C<sub>ll</sub></i> –
|
|
<i>E<sub>ll</sub></i>) and a right (<i>S<sub>lr</sub></i>
|
|
– <i>C<sub>lr</sub></i> – <i>E<sub>lr</sub></i>)
|
|
half.</li></ul>
|
|
|
|
Column C shows the state after the second iteration.</li>
|
|
|
|
<li>The topmost curve segment (<i>S<sub>ll</sub></i> –
|
|
<i>C<sub>ll</sub></i> – <i>E<sub>ll</sub></i>) is popped from
|
|
the stack.
|
|
|
|
<ul><li>The recursion level of this segment (stored in
|
|
<code>recLevel[2]</code>) is 2, which is <em>not</em> smaller than
|
|
the limit 2. Therefore, a <code>SEG_LINETO</code> (from
|
|
<i>S<sub>ll</sub></i> to <i>E<sub>ll</sub></i>) is passed to the
|
|
consumer.</li></ul>
|
|
|
|
The new state is shown in column D.</li>
|
|
|
|
|
|
<li>The topmost curve segment (<i>S<sub>lr</sub></i> –
|
|
<i>C<sub>lr</sub></i> – <i>E<sub>lr</sub></i>) is popped from
|
|
the stack.
|
|
|
|
<ul><li>The recursion level of this segment (stored in
|
|
<code>recLevel[1]</code>) is 2, which is <em>not</em> smaller than
|
|
the limit 2. Therefore, a <code>SEG_LINETO</code> (from
|
|
<i>S<sub>lr</sub></i> to <i>E<sub>lr</sub></i>) is passed to the
|
|
consumer.</li></ul>
|
|
|
|
The new state is shown in column E.</li>
|
|
|
|
<li>The algorithm proceeds by taking the topmost curve segment
|
|
(<i>S<sub>r</sub></i> – <i>C<sub>r</sub></i> –
|
|
<i>E<sub>r</sub></i>) from the stack.
|
|
|
|
<ul><li>The recursion level of this segment (stored in
|
|
<code>recLevel[0]</code>) is 1, which is smaller than
|
|
the limit 2.</li>
|
|
|
|
<li>The method <code>java.awt.geom.QuadCurve2D.getFlatnessSq</code>
|
|
is called to calculate the squared flatness.</li>
|
|
|
|
<li>For the sake of argument, we again assume that the squared
|
|
flatness is exceeding the threshold stored in
|
|
<code>flatnessSq</code>. Thus, the curve segment
|
|
(<i>S<sub>r</sub></i> – <i>C<sub>r</sub></i> –
|
|
<i>E<sub>r</sub></i>) is subdivided into a left and a right half,
|
|
namely
|
|
<i>S<sub>rl</sub></i> – <i>C<sub>rl</sub></i> –
|
|
<i>E<sub>rl</sub></i> and <i>S<sub>rr</sub></i> –
|
|
<i>C<sub>rr</sub></i> – <i>E<sub>rr</sub></i>. Both halves
|
|
are pushed onto the stack.</li></ul>
|
|
|
|
The new state is shown in column F.</li>
|
|
|
|
<li>The topmost curve segment (<i>S<sub>rl</sub></i> –
|
|
<i>C<sub>rl</sub></i> – <i>E<sub>rl</sub></i>) is popped from
|
|
the stack.
|
|
|
|
<ul><li>The recursion level of this segment (stored in
|
|
<code>recLevel[2]</code>) is 2, which is <em>not</em> smaller than
|
|
the limit 2. Therefore, a <code>SEG_LINETO</code> (from
|
|
<i>S<sub>rl</sub></i> to <i>E<sub>rl</sub></i>) is passed to the
|
|
consumer.</li></ul>
|
|
|
|
The new state is shown in column G.</li>
|
|
|
|
<li>The topmost curve segment (<i>S<sub>rr</sub></i> –
|
|
<i>C<sub>rr</sub></i> – <i>E<sub>rr</sub></i>) is popped from
|
|
the stack.
|
|
|
|
<ul><li>The recursion level of this segment (stored in
|
|
<code>recLevel[2]</code>) is 2, which is <em>not</em> smaller than
|
|
the limit 2. Therefore, a <code>SEG_LINETO</code> (from
|
|
<i>S<sub>rr</sub></i> to <i>E<sub>rr</sub></i>) is passed to the
|
|
consumer.</li></ul>
|
|
|
|
The new state is shown in column H.</li>
|
|
|
|
<li>The stack is now empty. The FlatteningPathIterator will fetch the
|
|
next segment from the base iterator, and process it.</li>
|
|
|
|
</ol>
|
|
|
|
<p>In order to split the most recently pushed segment, the
|
|
<code>subdivideQuadratic()</code> method passes <code>stack</code>
|
|
directly to
|
|
<code>QuadCurve2D.subdivide(double[],int,double[],int,double[],int)</code>.
|
|
Because the stack grows towards the beginning of the array, no data
|
|
needs to be copied around: <code>subdivide</code> will directly store
|
|
the result into the stack, which will have the contents shown to the
|
|
right.</p>
|
|
|
|
</body>
|
|
</html>
|