/* * Copyright (c) 2003, 2020, Oracle and/or its affiliates. All rights reserved. * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER. * * This code is free software; you can redistribute it and/or modify it * under the terms of the GNU General Public License version 2 only, as * published by the Free Software Foundation. Oracle designates this * particular file as subject to the "Classpath" exception as provided * by Oracle in the LICENSE file that accompanied this code. * * This code is distributed in the hope that it will be useful, but WITHOUT * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License * version 2 for more details (a copy is included in the LICENSE file that * accompanied this code). * * You should have received a copy of the GNU General Public License version * 2 along with this work; if not, write to the Free Software Foundation, * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. * * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA * or visit www.oracle.com if you need additional information or have any * questions. */ /* * * (C) Copyright IBM Corp. 1999-2003 - All Rights Reserved * * The original version of this source code and documentation is * copyrighted and owned by IBM. These materials are provided * under terms of a License Agreement between IBM and Sun. * This technology is protected by multiple US and International * patents. This notice and attribution to IBM may not be removed. */ /* * GlyphLayout is used to process a run of text into a run of run of * glyphs, optionally with position and char mapping info. * * The text has already been processed for numeric shaping and bidi. * The run of text that layout works on has a single bidi level. It * also has a single font/style. Some operations need context to work * on (shaping, script resolution) so context for the text run text is * provided. It is assumed that the text array contains sufficient * context, and the offset and count delimit the portion of the text * that needs to actually be processed. * * The font might be a composite font. Layout generally requires * tables from a single physical font to operate, and so it must * resolve the 'single' font run into runs of physical fonts. * * Some characters are supported by several fonts of a composite, and * in order to properly emulate the glyph substitution behavior of a * single physical font, these characters might need to be mapped to * different physical fonts. The script code that is assigned * characters normally considered 'common script' can be used to * resolve which physical font to use for these characters. The input * to the char to glyph mapper (which assigns physical fonts as it * processes the glyphs) should include the script code, and the * mapper should operate on runs of a single script. * * To perform layout, call get() to get a new (or reuse an old) * GlyphLayout, call layout on it, then call done(GlyphLayout) when * finished. There's no particular problem if you don't call done, * but it assists in reuse of the GlyphLayout. */ package sun.font; import java.lang.ref.SoftReference; import java.awt.Font; import java.awt.font.FontRenderContext; import java.awt.font.GlyphVector; import java.awt.geom.AffineTransform; import java.awt.geom.NoninvertibleTransformException; import java.awt.geom.Point2D; import java.util.ArrayList; import java.util.concurrent.ConcurrentHashMap; import static java.lang.Character.*; public final class GlyphLayout { // data for glyph vector private GVData _gvdata; // cached glyph layout data for reuse private static volatile GlyphLayout cache; // reusable private LayoutEngineFactory _lef; // set when get is called, unset when done is called private TextRecord _textRecord; // the text we're working on, used by iterators private ScriptRun _scriptRuns; // iterator over script runs private FontRunIterator _fontRuns; // iterator over physical fonts in a composite private int _ercount; private ArrayList _erecords; private Point2D.Float _pt; private FontStrikeDesc _sd; private float[] _mat; private float ptSize; private int _typo_flags; private int _offset; public static final class LayoutEngineKey { private Font2D font; private int script; private int lang; LayoutEngineKey() { } LayoutEngineKey(Font2D font, int script, int lang) { init(font, script, lang); } void init(Font2D font, int script, int lang) { this.font = font; this.script = script; this.lang = lang; } LayoutEngineKey copy() { return new LayoutEngineKey(font, script, lang); } Font2D font() { return font; } int script() { return script; } int lang() { return lang; } public boolean equals(Object rhs) { if (this == rhs) return true; if (rhs == null) return false; try { LayoutEngineKey that = (LayoutEngineKey)rhs; return this.script == that.script && this.lang == that.lang && this.font.equals(that.font); } catch (ClassCastException e) { return false; } } public int hashCode() { return script ^ lang ^ font.hashCode(); } } public static interface LayoutEngineFactory { /** * Given a font, script, and language, determine a layout engine to use. */ public LayoutEngine getEngine(Font2D font, int script, int lang); /** * Given a key, determine a layout engine to use. */ public LayoutEngine getEngine(LayoutEngineKey key); } public static interface LayoutEngine { /** * Given a strike descriptor, text, rtl flag, and starting point, append information about * glyphs, positions, and character indices to the glyphvector data, and advance the point. * * If the GVData does not have room for the glyphs, throws an IndexOutOfBoundsException and * leave pt and the gvdata unchanged. */ public void layout(FontStrikeDesc sd, float[] mat, float ptSize, int gmask, int baseIndex, TextRecord text, int typo_flags, Point2D.Float pt, GVData data); } /** * Return a new instance of GlyphLayout, using the provided layout engine factory. * If null, the system layout engine factory will be used. */ public static GlyphLayout get(LayoutEngineFactory lef) { if (lef == null) { lef = SunLayoutEngine.instance(); } GlyphLayout result = null; synchronized(GlyphLayout.class) { if (cache != null) { result = cache; cache = null; } } if (result == null) { result = new GlyphLayout(); } result._lef = lef; return result; } /** * Return the old instance of GlyphLayout when you are done. This enables reuse * of GlyphLayout objects. */ public static void done(GlyphLayout gl) { gl._lef = null; cache = gl; // object reference assignment is thread safe, it says here... } private static final class SDCache { public Font key_font; public FontRenderContext key_frc; public AffineTransform dtx; public AffineTransform gtx; public Point2D.Float delta; public FontStrikeDesc sd; private SDCache(Font font, FontRenderContext frc) { key_font = font; key_frc = frc; // !!! add getVectorTransform and hasVectorTransform to frc? then // we could just skip this work... dtx = frc.getTransform(); dtx.setTransform(dtx.getScaleX(), dtx.getShearY(), dtx.getShearX(), dtx.getScaleY(), 0, 0); float ptSize = font.getSize2D(); if (font.isTransformed()) { gtx = font.getTransform(); gtx.scale(ptSize, ptSize); delta = new Point2D.Float((float)gtx.getTranslateX(), (float)gtx.getTranslateY()); gtx.setTransform(gtx.getScaleX(), gtx.getShearY(), gtx.getShearX(), gtx.getScaleY(), 0, 0); gtx.preConcatenate(dtx); } else { delta = ZERO_DELTA; gtx = new AffineTransform(dtx); gtx.scale(ptSize, ptSize); } /* Similar logic to that used in SunGraphics2D.checkFontInfo(). * Whether a grey (AA) strike is needed is size dependent if * AA mode is 'gasp'. */ int aa = FontStrikeDesc.getAAHintIntVal(frc.getAntiAliasingHint(), FontUtilities.getFont2D(font), (int)Math.abs(ptSize)); int fm = FontStrikeDesc.getFMHintIntVal (frc.getFractionalMetricsHint()); sd = new FontStrikeDesc(dtx, gtx, font.getStyle(), aa, fm); } private static final Point2D.Float ZERO_DELTA = new Point2D.Float(); private static SoftReference> cacheRef; private static final class SDKey { private final Font font; private final FontRenderContext frc; private final int hash; SDKey(Font font, FontRenderContext frc) { this.font = font; this.frc = frc; this.hash = font.hashCode() ^ frc.hashCode(); } public int hashCode() { return hash; } public boolean equals(Object o) { try { SDKey rhs = (SDKey)o; return hash == rhs.hash && font.equals(rhs.font) && frc.equals(rhs.frc); } catch (ClassCastException e) { } return false; } } public static SDCache get(Font font, FontRenderContext frc) { // It is possible a translation component will be in the FRC. // It doesn't affect us except adversely as we would consider // FRC's which are really the same to be different. If we // detect a translation component, then we need to exclude it // by creating a new transform which excludes the translation. if (frc.isTransformed()) { AffineTransform transform = frc.getTransform(); if (transform.getTranslateX() != 0 || transform.getTranslateY() != 0) { transform = new AffineTransform(transform.getScaleX(), transform.getShearY(), transform.getShearX(), transform.getScaleY(), 0, 0); frc = new FontRenderContext(transform, frc.getAntiAliasingHint(), frc.getFractionalMetricsHint() ); } } SDKey key = new SDKey(font, frc); // garbage, yuck... ConcurrentHashMap cache = null; SDCache res = null; if (cacheRef != null) { cache = cacheRef.get(); if (cache != null) { res = cache.get(key); } } if (res == null) { res = new SDCache(font, frc); if (cache == null) { cache = new ConcurrentHashMap(10); cacheRef = new SoftReference>(cache); } else if (cache.size() >= 512) { cache.clear(); } cache.put(key, res); } return res; } } /** * Create a glyph vector. * @param font the font to use * @param frc the font render context * @param text the text, including optional context before start and after start + count * @param offset the start of the text to lay out * @param count the length of the text to lay out * @param flags bidi and context flags {@see #java.awt.Font} * @param result a StandardGlyphVector to modify, can be null * @return the layed out glyphvector, if result was passed in, it is returned */ public StandardGlyphVector layout(Font font, FontRenderContext frc, char[] text, int offset, int count, int flags, StandardGlyphVector result) { if (text == null || offset < 0 || count < 0 || (count > text.length - offset)) { throw new IllegalArgumentException(); } init(count); // need to set after init // go through the back door for this if (font.hasLayoutAttributes()) { AttributeValues values = ((AttributeMap)font.getAttributes()).getValues(); if (values.getKerning() != 0) _typo_flags |= 0x1; if (values.getLigatures() != 0) _typo_flags |= 0x2; } _offset = offset; // use cache now - can we use the strike cache for this? SDCache txinfo = SDCache.get(font, frc); _mat[0] = (float)txinfo.gtx.getScaleX(); _mat[1] = (float)txinfo.gtx.getShearY(); _mat[2] = (float)txinfo.gtx.getShearX(); _mat[3] = (float)txinfo.gtx.getScaleY(); _pt.setLocation(txinfo.delta); ptSize = font.getSize2D(); int lim = offset + count; int min = 0; int max = text.length; if (flags != 0) { if ((flags & Font.LAYOUT_RIGHT_TO_LEFT) != 0) { _typo_flags |= 0x80000000; // RTL } if ((flags & Font.LAYOUT_NO_START_CONTEXT) != 0) { min = offset; } if ((flags & Font.LAYOUT_NO_LIMIT_CONTEXT) != 0) { max = lim; } } int lang = -1; // default for now Font2D font2D = FontUtilities.getFont2D(font); if (font2D instanceof FontSubstitution) { font2D = ((FontSubstitution)font2D).getCompositeFont2D(); } _textRecord.init(text, offset, lim, min, max); int start = offset; if (font2D instanceof CompositeFont) { _scriptRuns.init(text, offset, count); // ??? how to handle 'common' chars _fontRuns.init((CompositeFont)font2D, text, offset, lim); while (_scriptRuns.next()) { int limit = _scriptRuns.getScriptLimit(); int script = _scriptRuns.getScriptCode(); while (_fontRuns.next(script, limit)) { Font2D pfont = _fontRuns.getFont(); /* layout can't deal with NativeFont instances. The * native font is assumed to know of a suitable non-native * substitute font. This currently works because * its consistent with the way NativeFonts delegate * in other cases too. */ if (pfont instanceof NativeFont) { pfont = ((NativeFont)pfont).getDelegateFont(); } int gmask = _fontRuns.getGlyphMask(); int pos = _fontRuns.getPos(); nextEngineRecord(start, pos, script, lang, pfont, gmask); start = pos; } } } else { _scriptRuns.init(text, offset, count); // ??? don't worry about 'common' chars while (_scriptRuns.next()) { int limit = _scriptRuns.getScriptLimit(); int script = _scriptRuns.getScriptCode(); nextEngineRecord(start, limit, script, lang, font2D, 0); start = limit; } } int ix = 0; int stop = _ercount; int dir = 1; if (_typo_flags < 0) { // RTL ix = stop - 1; stop = -1; dir = -1; } // _sd.init(dtx, gtx, font.getStyle(), frc.isAntiAliased(), frc.usesFractionalMetrics()); _sd = txinfo.sd; for (;ix != stop; ix += dir) { EngineRecord er = _erecords.get(ix); for (;;) { try { er.layout(); break; } catch (IndexOutOfBoundsException e) { if (_gvdata._count >=0) { _gvdata.grow(); } } } // Break out of the outer for loop if layout fails. if (_gvdata._count < 0) { break; } } // If layout fails (negative glyph count) create an un-laid out GV instead. // ie default positions. This will be a lot better than the alternative of // a complete blank layout. StandardGlyphVector gv; if (_gvdata._count < 0) { gv = new StandardGlyphVector(font, text, offset, count, frc); if (FontUtilities.debugFonts()) { FontUtilities.logWarning("OpenType layout failed on font: " + font); } } else { gv = _gvdata.createGlyphVector(font, frc, result); } // System.err.println("Layout returns: " + gv); return gv; } // // private methods // private GlyphLayout() { this._gvdata = new GVData(); this._textRecord = new TextRecord(); this._scriptRuns = new ScriptRun(); this._fontRuns = new FontRunIterator(); this._erecords = new ArrayList<>(10); this._pt = new Point2D.Float(); this._sd = new FontStrikeDesc(); this._mat = new float[4]; } private void init(int capacity) { this._typo_flags = 0; this._ercount = 0; this._gvdata.init(capacity); } private void nextEngineRecord(int start, int limit, int script, int lang, Font2D font, int gmask) { EngineRecord er = null; if (_ercount == _erecords.size()) { er = new EngineRecord(); _erecords.add(er); } else { er = _erecords.get(_ercount); } er.init(start, limit, font, script, lang, gmask); ++_ercount; } /** * Storage for layout to build glyph vector data, then generate a real GlyphVector */ public static final class GVData { public int _count; // number of glyphs, >= number of chars public int _flags; public int[] _glyphs; public float[] _positions; public int[] _indices; private static final int UNINITIALIZED_FLAGS = -1; public void init(int size) { _count = 0; _flags = UNINITIALIZED_FLAGS; if (_glyphs == null || _glyphs.length < size) { if (size < 20) { size = 20; } _glyphs = new int[size]; _positions = new float[size * 2 + 2]; _indices = new int[size]; } } public void grow() { grow(_glyphs.length / 4); // always grows because min length is 20 } public void grow(int delta) { int size = _glyphs.length + delta; int[] nglyphs = new int[size]; System.arraycopy(_glyphs, 0, nglyphs, 0, _count); _glyphs = nglyphs; float[] npositions = new float[size * 2 + 2]; System.arraycopy(_positions, 0, npositions, 0, _count * 2 + 2); _positions = npositions; int[] nindices = new int[size]; System.arraycopy(_indices, 0, nindices, 0, _count); _indices = nindices; } public StandardGlyphVector createGlyphVector(Font font, FontRenderContext frc, StandardGlyphVector result) { // !!! default initialization until we let layout engines do it if (_flags == UNINITIALIZED_FLAGS) { _flags = 0; if (_count > 1) { // if only 1 glyph assume LTR boolean ltr = true; boolean rtl = true; int rtlix = _count; // rtl index for (int i = 0; i < _count && (ltr || rtl); ++i) { int cx = _indices[i]; ltr = ltr && (cx == i); rtl = rtl && (cx == --rtlix); } if (rtl) _flags |= GlyphVector.FLAG_RUN_RTL; if (!rtl && !ltr) _flags |= GlyphVector.FLAG_COMPLEX_GLYPHS; } // !!! layout engines need to tell us whether they performed // position adjustments. currently they don't tell us, so // we must assume they did _flags |= GlyphVector.FLAG_HAS_POSITION_ADJUSTMENTS; } int[] glyphs = new int[_count]; System.arraycopy(_glyphs, 0, glyphs, 0, _count); float[] positions = null; if ((_flags & GlyphVector.FLAG_HAS_POSITION_ADJUSTMENTS) != 0) { positions = new float[_count * 2 + 2]; System.arraycopy(_positions, 0, positions, 0, positions.length); } int[] indices = null; if ((_flags & GlyphVector.FLAG_COMPLEX_GLYPHS) != 0) { indices = new int[_count]; System.arraycopy(_indices, 0, indices, 0, _count); } if (result == null) { result = new StandardGlyphVector(font, frc, glyphs, positions, indices, _flags); } else { result.initGlyphVector(font, frc, glyphs, positions, indices, _flags); } return result; } } /** * Utility class to keep track of script runs, which may have to be reordered rtl when we're * finished. */ private final class EngineRecord { private int start; private int limit; private int gmask; private int eflags; private LayoutEngineKey key; private LayoutEngine engine; EngineRecord() { key = new LayoutEngineKey(); } void init(int start, int limit, Font2D font, int script, int lang, int gmask) { this.start = start; this.limit = limit; this.gmask = gmask; this.key.init(font, script, lang); this.eflags = 0; // only request canonical substitution if we have combining marks for (int i = start; i < limit; ++i) { int ch = _textRecord.text[i]; if (isHighSurrogate((char)ch) && i < limit - 1 && isLowSurrogate(_textRecord.text[i+1])) { // rare case ch = toCodePoint((char)ch,_textRecord.text[++i]); // inc } int gc = getType(ch); if (gc == NON_SPACING_MARK || gc == ENCLOSING_MARK || gc == COMBINING_SPACING_MARK) { // could do range test also this.eflags = 0x4; break; } } this.engine = _lef.getEngine(key); // flags? } void layout() { _textRecord.start = start; _textRecord.limit = limit; engine.layout(_sd, _mat, ptSize, gmask, start - _offset, _textRecord, _typo_flags | eflags, _pt, _gvdata); } } }