/*
* Copyright (c) 2003, 2006, 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.
*/
package sun.font;
import java.awt.Font;
/* Remind: need to enhance to extend component list with a fallback
* list, which is not used in metrics or queries on the composite, but
* is used in drawing primitives and queries which supply an actual string.
* ie for a codepoint that is only in a fallback, font-wide queries such
* as FontMetrics.getHeight() will not take it into account.
* But getStringBounds(..) would take it into account.
* Its fuzzier for queries such as "canDisplay". If this does not include
* the fallback, then we probably want to add "canDisplayFallback()"
* But its probably OK to include it so long as only composites include
* fallbacks. If physicals do then it would be really confusing ..
*/
public final class CompositeFont extends Font2D {
private boolean[] deferredInitialisation;
String[] componentFileNames;
String[] componentNames;
/* because components can be lazily initialised the components field is
* private, to ensure all clients call getSlotFont()
*/
private PhysicalFont[] components;
int numSlots;
int numMetricsSlots;
int[] exclusionRanges;
int[] maxIndices;
int numGlyphs = 0;
int localeSlot = -1; // primary slot for this locale.
/* See isStdComposite() for when/how this is used */
boolean isStdComposite = true;
public CompositeFont(String name, String[] compFileNames,
String[] compNames, int metricsSlotCnt,
int[] exclRanges, int[] maxIndexes,
boolean defer, SunFontManager fm) {
handle = new Font2DHandle(this);
fullName = name;
componentFileNames = compFileNames;
componentNames = compNames;
if (compNames == null) {
numSlots = componentFileNames.length;
} else {
numSlots = componentNames.length;
}
/* Only the first "numMetricsSlots" slots are used for font metrics.
* the rest are considered "fallback" slots".
*/
numMetricsSlots = metricsSlotCnt;
exclusionRanges = exclRanges;
maxIndices = maxIndexes;
/*
* See if this is a windows locale which has a system EUDC font.
* If so add it as the final fallback component of the composite.
* The caller could be responsible for this, but for now it seems
* better that it is handled internally to the CompositeFont class.
*/
if (fm.getEUDCFont() != null) {
numSlots++;
if (componentNames != null) {
componentNames = new String[numSlots];
System.arraycopy(compNames, 0, componentNames, 0, numSlots-1);
componentNames[numSlots-1] =
fm.getEUDCFont().getFontName(null);
}
if (componentFileNames != null) {
componentFileNames = new String[numSlots];
System.arraycopy(compFileNames, 0,
componentFileNames, 0, numSlots-1);
}
components = new PhysicalFont[numSlots];
components[numSlots-1] = fm.getEUDCFont();
deferredInitialisation = new boolean[numSlots];
if (defer) {
for (int i=0; i<numSlots-1; i++) {
deferredInitialisation[i] = true;
}
}
} else {
components = new PhysicalFont[numSlots];
deferredInitialisation = new boolean[numSlots];
if (defer) {
for (int i=0; i<numSlots; i++) {
deferredInitialisation[i] = true;
}
}
}
fontRank = Font2D.FONT_CONFIG_RANK;
int index = fullName.indexOf('.');
if (index>0) {
familyName = fullName.substring(0, index);
/* composites don't call setStyle() as parsing the style
* takes place at the same time as parsing the family name.
* Do I really have to parse the style from the name?
* Need to look into having the caller provide this. */
if (index+1 < fullName.length()) {
String styleStr = fullName.substring(index+1);
if ("plain".equals(styleStr)) {
style = Font.PLAIN;
} else if ("bold".equals(styleStr)) {
style = Font.BOLD;
} else if ("italic".equals(styleStr)) {
style = Font.ITALIC;
} else if ("bolditalic".equals(styleStr)) {
style = Font.BOLD | Font.ITALIC;
}
}
} else {
familyName = fullName;
}
}
/* This method is currently intended to be called only from
* FontManager.getCompositeFontUIResource(Font)
* It creates a new CompositeFont with the contents of the Physical
* one pre-pended as slot 0.
*/
CompositeFont(PhysicalFont physFont, CompositeFont compFont) {
isStdComposite = false;
handle = new Font2DHandle(this);
fullName = physFont.fullName;
familyName = physFont.familyName;
style = physFont.style;
numMetricsSlots = 1; /* Only the physical Font */
numSlots = compFont.numSlots+1;
/* Ugly though it is, we synchronize here on the FontManager class
* because it is the lock used to do deferred initialisation.
* We need to ensure that the arrays have consistent information.
* But it may be possible to dispense with the synchronisation if
* it is harmless that we do not know a slot is already initialised
* and just need to discover that and mark it so.
*/
synchronized (FontManagerFactory.getInstance()) {
components = new PhysicalFont[numSlots];
components[0] = physFont;
System.arraycopy(compFont.components, 0,
components, 1, compFont.numSlots);
if (compFont.componentNames != null) {
componentNames = new String[numSlots];
componentNames[0] = physFont.fullName;
System.arraycopy(compFont.componentNames, 0,
componentNames, 1, compFont.numSlots);
}
if (compFont.componentFileNames != null) {
componentFileNames = new String[numSlots];
componentFileNames[0] = null;
System.arraycopy(compFont.componentFileNames, 0,
componentFileNames, 1, compFont.numSlots);
}
deferredInitialisation = new boolean[numSlots];
deferredInitialisation[0] = false;
System.arraycopy(compFont.deferredInitialisation, 0,
deferredInitialisation, 1, compFont.numSlots);
}
}
/* This is used for deferred initialisation, so that the components of
* a logical font are initialised only when the font is used.
* This can have a positive impact on start-up of most UI applications.
* Note that this technique cannot be used with a TTC font as it
* doesn't know which font in the collection is needed. The solution to
* this is that the initialisation checks if the returned font is
* really the one it wants by comparing the name against the name that
* was passed in (if none was passed in then you aren't using a TTC
* as you would have to specify the name in such a case).
* Assuming there's only two or three fonts in a collection then it
* may be sufficient to verify the returned name is the expected one.
* But half the time it won't be. However since initialisation of the
* TTC will initialise all its components then just do a findFont2D call
* to locate the right one.
* This code allows for initialisation of each slot on demand.
* There are two issues with this.
* 1) All metrics slots probably may be initialised anyway as many
* apps will query the overall font metrics. However this is not an
* absolute requirement
* 2) Some font configuration files on Solaris reference two versions
* of a TT font: a Latin-1 version, then a Pan-European version.
* One from /usr/openwin/lib/X11/fonts/TrueType, the other from
* a euro_fonts directory which is symlinked from numerous locations.
* This is difficult to avoid because the two do not share XLFDs so
* both will be consequently mapped by separate XLFDs needed by AWT.
* The difficulty this presents for lazy initialisation is that if
* all the components are not mapped at once, the smaller version may
* have been used only to be replaced later, and what is the consequence
* for a client that displayed the contents of this font already.
* After some thought I think this will not be a problem because when
* client tries to display a glyph only in the Euro font, the composite
* will ask all components of this font for that glyph and will get
* the euro one. Subsequent uses will all come from the 100% compatible
* euro one.
*/
private void doDeferredInitialisation(int slot) {
if (deferredInitialisation[slot] == false) {
return;
}
/* Synchronize on FontManager so that is the global lock
* to update its static set of deferred fonts.
* This global lock is rarely likely to be an issue as there
* are only going to be a few calls into this code.
*/
SunFontManager fm = SunFontManager.getInstance();
synchronized (fm) {
if (componentNames == null) {
componentNames = new String[numSlots];
}
if (components[slot] == null) {
/* Warning: it is possible that the returned component is
* not derived from the file name argument, this can happen if:
* - the file can't be found
* - the file has a bad font
* - the font in the file is superseded by a more complete one
* This should not be a problem for composite font as it will
* make no further use of this file, but code debuggers/
* maintainers need to be conscious of this possibility.
*/
if (componentFileNames != null &&
componentFileNames[slot] != null) {
components[slot] =
fm.initialiseDeferredFont(componentFileNames[slot]);
}
if (components[slot] == null) {
components[slot] = fm.getDefaultPhysicalFont();
}
String name = components[slot].getFontName(null);
if (componentNames[slot] == null) {
componentNames[slot] = name;
} else if (!componentNames[slot].equalsIgnoreCase(name)) {
components[slot] =
(PhysicalFont) fm.findFont2D(componentNames[slot],
style,
FontManager.PHYSICAL_FALLBACK);
}
}
deferredInitialisation[slot] = false;
}
}
/* To called only by FontManager.replaceFont */
void replaceComponentFont(PhysicalFont oldFont, PhysicalFont newFont) {
if (components == null) {
return;
}
for (int slot=0; slot<numSlots; slot++) {
if (components[slot] == oldFont) {
components[slot] = newFont;
if (componentNames != null) {
componentNames[slot] = newFont.getFontName(null);
}
}
}
}
public boolean isExcludedChar(int slot, int charcode) {
if (exclusionRanges == null || maxIndices == null ||
slot >= numMetricsSlots) {
return false;
}
int minIndex = 0;
int maxIndex = maxIndices[slot];
if (slot > 0) {
minIndex = maxIndices[slot - 1];
}
int curIndex = minIndex;
while (maxIndex > curIndex) {
if ((charcode >= exclusionRanges[curIndex])
&& (charcode <= exclusionRanges[curIndex+1])) {
return true; // excluded
}
curIndex += 2;
}
return false;
}
public void getStyleMetrics(float pointSize, float[] metrics, int offset) {
PhysicalFont font = getSlotFont(0);
if (font == null) { // possible?
super.getStyleMetrics(pointSize, metrics, offset);
} else {
font.getStyleMetrics(pointSize, metrics, offset);
}
}
public int getNumSlots() {
return numSlots;
}
public PhysicalFont getSlotFont(int slot) {
/* This is essentially the runtime overhead for deferred font
* initialisation: a boolean test on obtaining a slot font,
* which will happen per slot, on initialisation of a strike
* (as that is the only frequent call site of this method.
*/
if (deferredInitialisation[slot]) {
doDeferredInitialisation(slot);
}
SunFontManager fm = SunFontManager.getInstance();
try {
PhysicalFont font = components[slot];
if (font == null) {
try {
font = (PhysicalFont) fm.
findFont2D(componentNames[slot], style,
FontManager.PHYSICAL_FALLBACK);
components[slot] = font;
} catch (ClassCastException cce) {
font = fm.getDefaultPhysicalFont();
}
}
return font;
} catch (Exception e) {
return fm.getDefaultPhysicalFont();
}
}
FontStrike createStrike(FontStrikeDesc desc) {
return new CompositeStrike(this, desc);
}
/* This is set false when the composite is created using a specified
* physical font as the first slot and called by code which
* selects composites by locale preferences to know that this
* isn't a font which should be adjusted.
*/
public boolean isStdComposite() {
return isStdComposite;
}
/* This isn't very efficient but its infrequently used.
* StandardGlyphVector uses it when the client assigns the glyph codes.
* These may not be valid. This validates them substituting the missing
* glyph elsewhere.
*/
protected int getValidatedGlyphCode(int glyphCode) {
int slot = glyphCode >>> 24;
if (slot >= numSlots) {
return getMapper().getMissingGlyphCode();
}
int slotglyphCode = glyphCode & CompositeStrike.SLOTMASK;
PhysicalFont slotFont = getSlotFont(slot);
if (slotFont.getValidatedGlyphCode(slotglyphCode) ==
slotFont.getMissingGlyphCode()) {
return getMapper().getMissingGlyphCode();
} else {
return glyphCode;
}
}
public CharToGlyphMapper getMapper() {
if (mapper == null) {
mapper = new CompositeGlyphMapper(this);
}
return mapper;
}
public boolean hasSupplementaryChars() {
for (int i=0; i<numSlots; i++) {
if (getSlotFont(i).hasSupplementaryChars()) {
return true;
}
}
return false;
}
public int getNumGlyphs() {
if (numGlyphs == 0) {
numGlyphs = getMapper().getNumGlyphs();
}
return numGlyphs;
}
/**代码未完, 请加载全部代码(NowJava.com).**/