/*
 * Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
 * Use of this file is governed by the BSD 3-clause license that
 * can be found in the LICENSE.txt file in the project root.
 */

package org.antlr.v4.gui;

Font metrics. The only way to generate accurate images in any format that contain text is to know the font metrics. Specifically, we need to know the width of every character and the maximum height (since we want all characters to fit within same line height). I used ttf2tfm to dump the font metrics from Mac TrueType fonts and then converted that to a Java class for use in a PostScript generator for trees. Commands:
 $ ttf2tfm /Library/Fonts/Arial\ Black.ttf > metrics
Then run metrics into python code after stripping header/footer:
 #
 # Process lines from ttf2tfm that look like this:
 # Glyph  Code   Glyph Name                Width  llx    lly      urx    ury
 # ------------------------------------------------------------------------
 #     3  00020  space                       333  0,     0 --     0,     0
 #
 lines = open("metrics").read().split('\n')
 print "public class FontName {"
 print "    {"
 maxh = 0;
 for line in lines[4:]: # skip header 0..3
		 all = line.split(' ')
		 words = [x for x in all if len(x)>0]
		 ascii = int(words[1], 16)
		 height = int(words[8])
		 if height>maxh: maxh = height
		 if ascii>=128: break
		 print "        widths[%d] = %s; // %s" % (ascii, words[3], words[2])
 print "        maxCharHeight = "+str(maxh)+";"
 print "    }"
 print "}"
Units are 1000th of an 'em'.
/** Font metrics. The only way to generate accurate images * in any format that contain text is to know the font metrics. * Specifically, we need to know the width of every character and the * maximum height (since we want all characters to fit within same line height). * I used ttf2tfm to dump the font metrics from Mac TrueType fonts and * then converted that to a Java class for use in a PostScript generator * for trees. Commands: * * <pre> * $ ttf2tfm /Library/Fonts/Arial\ Black.ttf &gt; metrics * </pre> * * Then run metrics into python code after stripping header/footer: * * <pre> * # * # Process lines from ttf2tfm that look like this: * # Glyph Code Glyph Name Width llx lly urx ury * # ------------------------------------------------------------------------ * # 3 00020 space 333 0, 0 -- 0, 0 * # * lines = open("metrics").read().split('\n') * print "public class FontName {" * print " {" * maxh = 0; * for line in lines[4:]: # skip header 0..3 * all = line.split(' ') * words = [x for x in all if len(x)&gt;0] * ascii = int(words[1], 16) * height = int(words[8]) * if height&gt;maxh: maxh = height * if ascii&gt;=128: break * print " widths[%d] = %s; // %s" % (ascii, words[3], words[2]) * * print " maxCharHeight = "+str(maxh)+";" * print " }" * print "}" * </pre> * * Units are 1000th of an 'em'. */
public abstract class BasicFontMetrics { public static final int MAX_CHAR = '\u00FF'; protected int maxCharHeight; protected int[] widths = new int[MAX_CHAR+1]; public double getWidth(String s, int fontSize) { double w = 0; for (char c : s.toCharArray()) { w += getWidth(c, fontSize); } return w; } public double getWidth(char c, int fontSize) { if ( c > MAX_CHAR || widths[c]==0 ) return widths['m']/1000.0; // return width('m') return widths[c]/1000.0 * fontSize; } public double getLineHeight(int fontSize) { return maxCharHeight / 1000.0 * fontSize; } }