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

TMW->EWTS:

Browse source 
Fixed part of bug 998476 and part of an undocumented bug.  Discovered a
new bug, "aM" should be generated but only "M" is.

The undocumented bug was that laMA was generated when lAM should have been.

The part of bug 998476 that was fixed: laM, laH, etc. are now generated.

This does nothing about paN etc.

Some refactoring here; this is not a minimal diff.

Added tests of TMW->EWTS that use ACIP to get the TMW in place
because EWTS->TMW is a faulty keyboard at present.
This commit is contained in:
dchandler 2005-02-07 03:17:40 +00:00
parent a82afad92c
commit 8dcb623382
10 changed files with 1033 additions and 699 deletions
Download patch file Download diff file Expand all files Collapse all files

76
source/org/thdl/tib/text/TGCPair.java
View file

@ -18,6 +18,8 @@ Contributor(s): ______________________________________.
package org.thdl.tib.text;
import org.thdl.util.ThdlDebug;
/** An ordered pair consisting of a Tibetan grapheme cluster's (see
{@link org.thdl.tib.text.tshegbar.UnicodeGraphemeCluster} for a
@ -105,8 +107,13 @@ public class TGCPair implements THDLWylieConstants {
b.append(ch);
}
}
if (vowelWylie != null)
if (vowelWylie != null) {
if (consonantWylie != null // we need a stack to put an 'a' onto
&& !TibetanMachineWeb.startsWithWylieVowelSequence(vowelWylie)) {
b.append("a"); // we want laM, not lM -- see bug 998476
}
b.append(vowelWylie);
}
return b.toString();
}
public String getACIP() {
@ -182,8 +189,9 @@ public class TGCPair implements THDLWylieConstants {
int realClassification = -37;
if (vowelWylie == null && classification == TYPE_TIBETAN)
realClassification = CONSONANTAL_WITHOUT_VOWEL;
if (vowelWylie != null && classification == TYPE_TIBETAN)
if (vowelWylie != null && classification == TYPE_TIBETAN) {
realClassification = CONSONANTAL_WITH_VOWEL;
}
if (vowelWylie == null && classification == TYPE_SANSKRIT)
realClassification = SANSKRIT_WITHOUT_VOWEL;
if (vowelWylie != null && classification == TYPE_SANSKRIT)
@ -204,10 +212,72 @@ public class TGCPair implements THDLWylieConstants {
if (vowelWylie.equals("uA") || vowelWylie.equals("Au"))
vowelWylie = "U";
}
this.vowelWylie = vowelWylie;
// Normalize vowelWylie such that any real vowel (i.e., a
// vowel for which TibetanMachineWeb.isWylieVowel(..) returns
// true) comes before any non-vowel but vowelish combining
// character like 'M', '?', '~M', '~M`', or 'H':
this.vowelWylie = normalizedVowel(vowelWylie);
this.classification = realClassification;
}
private static final int MAX_CHARACTERS_IN_VOWELISH = 3; // ~M` has 3 characters
/** Returns v such that all the normal vowels in v come first and
the combining marks (Sanskrit-ish) come last. Relative order
in each of the two groups is preserved. */
private static String normalizedVowel(String v) {
if (null == v || "".equals(v)) return null;
StringBuffer vowel_sb = new StringBuffer();
StringBuffer vowelish_sb = new StringBuffer();
int mark = 0;
int maxVowelishLength
= Math.max(TibetanMachineWeb.getMaxEwtsVowelLength(),
TGCPair.MAX_CHARACTERS_IN_VOWELISH);
for (int i = 0; i < v.length(); i++) {
// Grab ~M` if it's there because you don't want to grab
// ~M and then have ` left over, which is not vowelish.
for (int j = maxVowelishLength; j > 0; j--) {
if (i + j <= v.length()) {
String x = v.substring(mark, i + j);
if (TibetanMachineWeb.isWylieVowel(x)) {
mark = i + j;
i += j - 1;
vowel_sb.append(x);
} else if (isWylieVowelishButNotVowel(x)) {
mark = i + j;
i += j - 1;
vowelish_sb.append(x);
}
}
}
}
if (mark < v.length()) {
vowelish_sb.append(v.substring(mark));
ThdlDebug.noteIffyCode();
// FIXME(dchandler): what should I do here? I doubt v is
// valid.
}
if (vowelish_sb.length() > 0) { // just an optimization
vowel_sb.append(vowelish_sb);
return vowel_sb.toString();
} else {
ThdlDebug.verify(vowel_sb.toString().equals(v));
return v;
}
}
/** Returns true if v is in the set { "M", "H", "~M", "~M`", "?" }. */
private static boolean isWylieVowelishButNotVowel(String v) {
boolean ans = (v.equals("M")
|| v.equals("H")
|| v.equals("~M`")
|| v.equals("~M")
|| v.equals("?"));
ThdlDebug.verify(!ans || TGCPair.MAX_CHARACTERS_IN_VOWELISH >= v.length());
return ans;
}
public String toString() {
return "<TGCPair wylie=" + getWylie() + " classification="
+ classification + "/>";