thdltools/Jskad
1
0
Fork 0
Code Issues Pull requests Projects Releases Wiki Activity

TM->TMW and TMW->TM conversion in RTF is now supported. I've

Browse source 
noticed that formatting is mostly OK but sometimes gets bungled slightly.
I tried everything I could think of, and now I'm passing the buck to Java's
RTF support.

TMW_RTF_TO_THDL_WYLIE (now misnamed) support TMW->TM
conversion (but not TM->TMW).  There is an automated test case for a
TMW->TM conversion.

I have full confidence in this conversion.  Even the smallest glitch in the core
functionality (not formatting) would surprise me.

Note that the JUnit test TMW_RTF_TO_THDL_WYLIETest sometimes fails
due to one- or two-line diffs between the actual and expected outputs.  This
is because Java's RTF support is not deterministic, I'm guessing, and is not
a real failure.  I'm too lazy to make a more elaborate sed/diff mechanism
that works on all platforms, and that would complicate the build anyway.
This commit is contained in:
dchandler 2003-05-31 23:21:29 +00:00
parent 56d5ac7210
commit 0235263ddf
8 changed files with 416 additions and 103 deletions
Download patch file Download diff file Expand all files Collapse all files

42
source/org/thdl/tib/text/DuffData.java
View file

@ -19,9 +19,13 @@ Contributor(s): ______________________________________.
package org.thdl.tib.text;
/**
* A wrapper object for a stretch of TibetanMachineWeb data that shares the same font.
* A piece of DuffData consists of a font number and a string.
* The font number is a number from one to ten, corresponding
* A wrapper object for a stretch of TibetanMachineWeb (TMW) or
* TibetanMachine (TM) data that shares the same font. A piece of
* DuffData consists of a font number and a string. The fact that this
* stretch is TMW vs. TM is not stored in this object; the client must
* remember that itself.
*
* For TMW, the font number is a number from one to ten, corresponding
* to the ten TibetanMachineWeb fonts, as follows:
* <p>
* 1 - TibetanMachineWeb<br>
@ -29,9 +33,18 @@ package org.thdl.tib.text;
* ...<br>
* 10 - TibetanMachineWeb9<br>
* <p>
* The string represents a contiguous stretch of data in that
* font, i.e. a stretch of TibetanMachineWeb that doesn't require a font change.
*/
* For TM, the font number is a number from one to five, corresponding
* to the five TibetanMachineWeb fonts, as follows:
* <p>
* 1 - TibetanMachine<br>
* 2 - TibetanMachineSkt1<br>
* 3 - TibetanMachineSkt2<br>
* 4 - TibetanMachineSkt3<br>
* 5 - TibetanMachineSkt4<br>
* <p>
* The string represents a contiguous stretch of data in that font,
* i.e. a stretch of TibetanMachineWeb or TibetanMachine that doesn't
* require a font change. */
public class DuffData {
/**
* a string of text
@ -43,11 +56,24 @@ public class DuffData {
public int font;
/**
* @param s a string of TibetanMachineWeb text
* @param i a TibetanMachineWeb font number
* @param s a string of TibetanMachineWeb or TibetanMachine text
* @param i a TibetanMachineWeb or TibetanMachine font number
*/
public DuffData(String s, int i) {
text = s;
font = i;
}
/** Default constructor. The DuffData is invalid after this until
you call setData or manipulate the public fields. */
public DuffData() { }
/** Changes the text and font this DuffData represents.
* @param c a character of TibetanMachineWeb or TibetanMachine text
* @param i a TibetanMachineWeb or TibetanMachine font number
*/
public void setData(char c, int i) {
text = new String(new char[] { c });
font = i;
}
}

111
source/org/thdl/tib/text/TibetanDocument.java
View file

@ -150,16 +150,18 @@ public class TibetanDocument extends DefaultStyledDocument {
* @param pos the position at which you want to insert text
*/
public int insertDuff(int pos, DuffData[] glyphs) {
return insertDuff(tibetanFontSize, pos, glyphs);
return insertDuff(tibetanFontSize, pos, glyphs, true);
}
private int insertDuff(int fontSize, int pos, DuffData[] glyphs) {
private int insertDuff(int fontSize, int pos, DuffData[] glyphs, boolean asTMW) {
if (glyphs == null)
return pos;
MutableAttributeSet mas;
for (int i=0; i<glyphs.length; i++) {
mas = TibetanMachineWeb.getAttributeSet(glyphs[i].font);
mas = ((asTMW)
? TibetanMachineWeb.getAttributeSet(glyphs[i].font)
: TibetanMachineWeb.getAttributeSetTM(glyphs[i].font));
appendDuff(fontSize, pos, glyphs[i].text, mas);
pos += glyphs[i].text.length();
}
@ -350,7 +352,8 @@ public class TibetanDocument extends DefaultStyledDocument {
TibetanMachineWeb. Works within the range [start, end).
Using a negative number for end means that this will run to
the end of the document. Be sure to set the size for Tibetan
as you like it before using this. SPEED_FIXME: might be
as you like it before using this (well, it usually gets it
right on its own, but just in case). SPEED_FIXME: might be
faster to run over the elements, if they are one per font. */
public void replaceTahomaCurlyBracesAndBackslashes(int begin, int end) {
if (end < 0)
@ -382,7 +385,7 @@ public class TibetanDocument extends DefaultStyledDocument {
} catch (Exception e) {
// leave it as tibetanFontSize
}
insertDuff(fontSize, i, toReplaceWith);
insertDuff(fontSize, i, toReplaceWith, true);
remove(i+1, 1);
}
}
@ -393,4 +396,102 @@ public class TibetanDocument extends DefaultStyledDocument {
ThdlDebug.noteIffyCode();
}
}
/** Converts all TibetanMachineWeb glyphs in the document to
TibetanMachine. Works within the range [start, end). Using a
negative number for end means that this will run to the end of
the document. Be sure to set the size for Tibetan as you like
it before using this (well, it usually gets it right on its
own, but just in case). SPEED_FIXME: might be faster to run
over the elements, if they are one per font. */
public void convertToTM(int begin, int end) {
convertTMW_TM(begin, end, true);
}
/** Converts all TibetanMachine glyphs in the document to
TibetanMachineWeb. Works within the range [start, end).
Using a negative number for end means that this will run to
the end of the document. Be sure to set the size for Tibetan
as you like it before using this (well, it usually gets it
right on its own, but just in case). SPEED_FIXME: might be
faster to run over the elements, if they are one per font. */
public void convertToTMW(int begin, int end) {
convertTMW_TM(begin, end, false);
}
/** Helper function.
@see convertToTMW(int,int)
@see convertToTM(int,int) */
private void convertTMW_TM(int begin, int end, boolean toTM) {
if (end < 0)
end = getLength();
if (begin >= end)
return;
int i = begin;
try {
DuffData[] equivalent = new DuffData[1];
equivalent[0] = new DuffData();
while (i < end) {
AttributeSet attr = getCharacterElement(i).getAttributes();
String fontName = StyleConstants.getFontFamily(attr);
int fontNum
= (toTM
? TibetanMachineWeb.getTMWFontNumber(fontName)
: TibetanMachineWeb.getTMFontNumber(fontName));
if (0 != fontNum) {
DuffCode dc = null;
try {
if (toTM) {
dc = TibetanMachineWeb.mapTMWtoTM(fontNum - 1,
getText(i,1).charAt(0));
} else {
dc = TibetanMachineWeb.mapTMtoTMW(fontNum - 1,
getText(i,1).charAt(0));
}
if (null != dc) {
equivalent[0].setData(dc.getCharacter(),
dc.getFontNum());
}
} catch (ArrayIndexOutOfBoundsException e) {
// we handle this below...
System.out.println("FIXME: "
+ (toTM ? "TMW->TM" : "TM->TMW")
+ " conversion is in trouble");
System.out.println("font is " + (fontNum - 1)
+ ", char is "
+ (int)getText(i,1).charAt(0)
+ "; pos is " + i);
ThdlDebug.noteIffyCode();
}
if (null != dc) {
int fontSize = tibetanFontSize;
try {
fontSize = ((Integer)getCharacterElement(i).getAttributes().getAttribute(StyleConstants.FontSize)).intValue();
} catch (Exception e) {
// leave it as tibetanFontSize
}
insertDuff(fontSize, i, equivalent, !toTM);
remove(i+1, 1);
} else {
// DLC FIXME: insert into document a string
// saying "there's no TM equivalent for this."
// (For now, I'm inserting the alphabet and
// all the numbers in a big font in TMW to try
// and get some attention. And I've
// *documented* this on the website.)
String trickyTMW
= "!-\"-#-,-%-&-'-(-)-*-+-,-.-/-0-1-2-3-4-5-6-7-8-9-:-;-<-=->-?-0-1-2-3-4-5-6-7-8-9-";
equivalent[0] = new DuffData(trickyTMW, 1);
insertDuff(72, i, equivalent, true);
i += trickyTMW.length() + 1;
}
}
i++;
}
} catch (BadLocationException ble) {
ble.printStackTrace();
ThdlDebug.noteIffyCode();
}
}
}

89
source/org/thdl/tib/text/TibetanMachineWeb.java
View file

@ -74,7 +74,12 @@ public class TibetanMachineWeb implements THDLWylieConstants {
private static String fileName = "tibwn.ini";
private static final String DELIMITER = "~";
private static Set top_vowels;
/** a way of encoding the choice of TibetanMachineWeb font from
that family of 10 fonts: */
private static SimpleAttributeSet[] webFontAttributeSet = new SimpleAttributeSet[11];
/** a way of encoding the choice of TibetanMachine font from
that family of 5 fonts: */
private static SimpleAttributeSet[] normFontAttributeSet = new SimpleAttributeSet[6];
private static boolean hasDisambiguatingKey; //to disambiguate gy and g.y=
private static char disambiguating_key;
private static boolean hasSanskritStackingKey; //for stacking Sanskrit
@ -244,6 +249,12 @@ public class TibetanMachineWeb implements THDLWylieConstants {
StyleConstants.setFontFamily(webFontAttributeSet[i],tmwFontNames[i]);
}
normFontAttributeSet[0] = null;
for (int i=1; i<normFontAttributeSet.length; i++) {
normFontAttributeSet[i] = new SimpleAttributeSet();
StyleConstants.setFontFamily(normFontAttributeSet[i],tmFontNames[i]);
}
StringTokenizer sTok;
topSet = new HashSet();
leftSet = new HashSet();
@ -560,6 +571,23 @@ public static SimpleAttributeSet getAttributeSet(int font) {
return null;
}
/**
* Gets the AttributeSet for the given TibetanMachine font.
* This information is required in order to be able to put styled
* text into {@link TibetanDocument TibetanDocument}.
* @param font the number of the TibetanMachineWeb font for which
* you want the SimpleAttributeSet: TibetanMachine = 1,
* TibetanMachineSkt1 = 2, etc. up to 5
* @return a SimpleAttributeSet for the given font - that is,
* a way of encoding the font itself
*/
public static SimpleAttributeSet getAttributeSetTM(int font) {
if (font > -1 && font < normFontAttributeSet.length)
return normFontAttributeSet[font];
else
return null;
}
/**
* Says whether or not the character is formatting.
* @param c the character to be checked
@ -851,6 +879,10 @@ public static DuffCode getHalfHeightGlyph(String hashKey) {
return dc[REDUCED_C];
}
private static final DuffCode TMW_cr = new DuffCode(1, '\r');
private static final DuffCode TMW_lf = new DuffCode(1, '\n');
private static final DuffCode TMW_tab = new DuffCode(1, '\t');
/** Returns the DuffCode for the TibetanMachineWeb glyph corresponding
to the given TibetanMachine font
(0=norm,1=Skt1,2=Skt2,3=Skt3,4=Skt4) and character(32-254).
@ -861,14 +893,36 @@ public static DuffCode getHalfHeightGlyph(String hashKey) {
an existing TibetanMachine glyph, null is returned. In general,
though, this method may raise a runtime exception if you pass in a
(font, ord) that doesn't correspond to an existing TibetanMachine
glyph. */
public static DuffCode mapTMtoTMW(int font, int ordinal) {
glyph.
Only a few control characters are supported: '\r' (carriage
return), '\n' (line feed), and '\t' (tab).
*/
public static DuffCode mapTMtoTMW(int font, int ordinal)
throws ArrayIndexOutOfBoundsException {
if (ordinal < 32) {
if (ordinal == (int)'\r')
return TMW_cr;
else if (ordinal == (int)'\n')
return TMW_lf;
else if (ordinal == (int)'\t')
return TMW_tab;
else {
// for robustness, just return font 1, char ordinal.
ThdlDebug.noteIffyCode();
return null;
}
}
DuffCode ans = TMtoTMW[font][ordinal-32];
// comment this out to test via main(..):
ThdlDebug.verify(null != ans);
return ans;
}
private static final DuffCode TM_cr = new DuffCode(1, '\r');
private static final DuffCode TM_lf = new DuffCode(1, '\n');
private static final DuffCode TM_tab = new DuffCode(1, '\t');
/** Returns the DuffCode for the TibetanMachine glyph corresponding to
the given TibetanMachineWeb font
(0=TibetanMachineWeb,1=TibetanMachineWeb1,...) and character(32-127).
@ -880,8 +934,27 @@ public static DuffCode mapTMtoTMW(int font, int ordinal) {
existing TibetanMachineWeb glyph, null is returned. In general,
though, this method may raise a runtime exception if you pass in a
(font, ord) that doesn't correspond to an existing
TibetanMachineWeb glyph. */
public static DuffCode mapTMWtoTM(int font, int ordinal) {
TibetanMachineWeb glyph.
Only a few control characters are supported: '\r' (carriage
return), '\n' (line feed), and '\t' (tab).
*/
public static DuffCode mapTMWtoTM(int font, int ordinal)
throws ArrayIndexOutOfBoundsException {
if (ordinal < 32) {
if (ordinal == (int)'\r')
return TM_cr;
else if (ordinal == (int)'\n')
return TM_lf;
else if (ordinal == (int)'\t')
return TM_tab;
else {
// for robustness, just return font 1, char ordinal.
ThdlDebug.noteIffyCode();
return null;
}
}
DuffCode ans = TMWtoTM[font][ordinal-32];
// comment this out to test via main(..):
ThdlDebug.verify(null != ans || (font == 7 && ordinal == 91));
@ -975,7 +1048,13 @@ private static DuffCode getTMtoTMW(int font, int code) {
return TMtoTMW[font][code];
}
private static int getTMFontNumber(String name) {
/**
* Gets the TibetanMachine font number for this font name.
* @param name a font name
* @return between 1 and 5 if the font is one
* of the TibetanMachine fonts, otherwise 0
*/
public static int getTMFontNumber(String name) {
String internedName = name.intern();
for (int i=1; i<tmFontNames.length; i++) {
if (internedName == tmFontNames[i])