-/*\r *******************************************************************************\r * Copyright (C) 2011, International Business Machines Corporation and *\r * others. All Rights Reserved. *\r *******************************************************************************\r */\rpackage com.ibm.icu.dev.test.translit;\r\rimport java.util.List;\r\rimport com.ibm.icu.dev.test.TestFmwk;\rimport com.ibm.icu.dev.test.util.ICUPropertyFactory;\rimport com.ibm.icu.dev.test.util.UnicodeProperty;\rimport com.ibm.icu.dev.test.util.UnicodeProperty.Factory;\rimport com.ibm.icu.dev.test.util.UnicodePropertySymbolTable;\rimport com.ibm.icu.text.UnicodeSet;\r\r/**\r * @author markdavis\r *\r */\rpublic class TestUnicodeProperty extends TestFmwk{\r public static void main(String[] args) {\r new TestUnicodeProperty().run(args);\r }\r static final UnicodeSet casedLetter = new UnicodeSet("[:gc=cased letter:]");\r static final UnicodeSet letter = new UnicodeSet("[:gc=L:]");\r\r\r public void TestBasic() {\r Factory factory = ICUPropertyFactory.make();\r UnicodeProperty property = factory.getProperty("gc");\r List values = property.getAvailableValues();\r assertTrue("Values contain GC values", values.contains("Unassigned"));\r final UnicodeSet lu = property.getSet("Lu");\r if (!assertTrue("Gc=L contains 'A'", lu.contains('A'))) {\r errln("Contents:\t" + lu.complement().complement().toPattern(false));\r }\r }\r\r public void TestSymbolTable() {\r Factory factory = ICUPropertyFactory.make();\r UnicodePropertySymbolTable upst = new UnicodePropertySymbolTable(factory);\r UnicodeSet.setDefaultXSymbolTable(upst);\r try {\r final UnicodeSet luSet = new UnicodeSet("[:gc=L:]");\r assertTrue("Gc=L contains 'A'", luSet.contains('A'));\r assertTrue("Gc=L contains 'Z'", luSet.contains('Z'));\r assertFalse("Gc=L contains 'a'", luSet.contains('1'));\r UnicodeSet casedLetter2 = new UnicodeSet("[:gc=cased letter:]");\r assertEquals("gc=lc are equal", casedLetter, casedLetter2);\r } finally {\r // restore the world\r UnicodeSet.setDefaultXSymbolTable(null);\r }\r }\r\r public void TestSymbolTable2() {\r Factory factory = new MyUnicodePropertyFactory();\r UnicodePropertySymbolTable upst = new UnicodePropertySymbolTable(factory);\r UnicodeSet.setDefaultXSymbolTable(upst);\r try {\r final UnicodeSet luSet = new UnicodeSet("[:gc=L:]");\r assertFalse("Gc=L contains 'A'", luSet.contains('A'));\r if (!assertTrue("Gc=L contains 'Z'", luSet.contains('Z'))) {\r errln("Contents:\t" + luSet.complement().complement().toPattern(false));\r }\r assertFalse("Gc=L contains 'a'", luSet.contains('1'));\r UnicodeSet casedLetter2 = new UnicodeSet("[:gc=cased letter:]");\r assertNotEquals("gc=lc should not be equal", casedLetter, casedLetter2);\r } finally {\r // restore the world\r UnicodeSet.setDefaultXSymbolTable(null);\r }\r }\r\r\r /**\r * For testing, override to set A-M to Cn.\r */\r static class MyUnicodeGCProperty extends UnicodeProperty.SimpleProperty {\r UnicodeProperty icuProperty = ICUPropertyFactory.make().getProperty("Gc");\r {\r setName(icuProperty.getName());\r setType(icuProperty.getType());\r }\r @Override\r protected String _getValue(int codepoint) {\r if (codepoint >= 'A' && codepoint <= 'M') {\r return "Unassigned";\r } else {\r return icuProperty.getValue(codepoint);\r }\r }\r @Override\r protected List _getValueAliases(String valueAlias, List result) {\r return icuProperty.getValueAliases(valueAlias, result);\r }\r @Override\r public List _getNameAliases(List result) {\r return icuProperty.getNameAliases();\r }\r }\r\r /**\r * For testing, override to set A-Z to Cn.\r */\r static class MyUnicodePropertyFactory extends ICUPropertyFactory {\r private MyUnicodePropertyFactory() {\r add(new MyUnicodeGCProperty());\r }\r }\r\r static class MyUnicodePropertySymbolTable extends UnicodePropertySymbolTable {\r public MyUnicodePropertySymbolTable(Factory factory) {\r super(factory);\r }\r }\r}\r
\ No newline at end of file
+/* *******************************************************************************
+ * Copyright (C) 2011, International Business Machines Corporation and *
+ * others. All Rights Reserved. *
+ *******************************************************************************
+ */
+package com.ibm.icu.dev.test.translit;
+
+import java.util.List;
+
+import com.ibm.icu.dev.test.TestFmwk;
+import com.ibm.icu.dev.test.util.ICUPropertyFactory;
+import com.ibm.icu.dev.test.util.UnicodeProperty;
+import com.ibm.icu.dev.test.util.UnicodeProperty.Factory;
+import com.ibm.icu.dev.test.util.UnicodePropertySymbolTable;
+import com.ibm.icu.text.UnicodeSet;
+
+/**
+ * @author markdavis
+ *
+ */
+public class TestUnicodeProperty extends TestFmwk{
+ public static void main(String[] args) {
+ new TestUnicodeProperty().run(args);
+ }
+ static final UnicodeSet casedLetter = new UnicodeSet("[:gc=cased letter:]");
+ static final UnicodeSet letter = new UnicodeSet("[:gc=L:]");
+
+
+ public void TestBasic() {
+ Factory factory = ICUPropertyFactory.make();
+ UnicodeProperty property = factory.getProperty("gc");
+ List values = property.getAvailableValues();
+ assertTrue("Values contain GC values", values.contains("Unassigned"));
+ final UnicodeSet lu = property.getSet("Lu");
+ if (!assertTrue("Gc=L contains 'A'", lu.contains('A'))) {
+ errln("Contents:\t" + lu.complement().complement().toPattern(false));
+ }
+ }
+
+ public void TestSymbolTable() {
+ Factory factory = ICUPropertyFactory.make();
+ UnicodePropertySymbolTable upst = new UnicodePropertySymbolTable(factory);
+ UnicodeSet.setDefaultXSymbolTable(upst);
+ try {
+ final UnicodeSet luSet = new UnicodeSet("[:gc=L:]");
+ assertTrue("Gc=L contains 'A'", luSet.contains('A'));
+ assertTrue("Gc=L contains 'Z'", luSet.contains('Z'));
+ assertFalse("Gc=L contains 'a'", luSet.contains('1'));
+ UnicodeSet casedLetter2 = new UnicodeSet("[:gc=cased letter:]");
+ assertEquals("gc=lc are equal", casedLetter, casedLetter2);
+ } finally {
+ // restore the world
+ UnicodeSet.setDefaultXSymbolTable(null);
+ }
+ }
+
+ public void TestSymbolTable2() {
+ Factory factory = new MyUnicodePropertyFactory();
+ UnicodePropertySymbolTable upst = new UnicodePropertySymbolTable(factory);
+ UnicodeSet.setDefaultXSymbolTable(upst);
+ try {
+ final UnicodeSet luSet = new UnicodeSet("[:gc=L:]");
+ assertFalse("Gc=L contains 'A'", luSet.contains('A'));
+ if (!assertTrue("Gc=L contains 'Z'", luSet.contains('Z'))) {
+ errln("Contents:\t" + luSet.complement().complement().toPattern(false));
+ }
+ assertFalse("Gc=L contains 'a'", luSet.contains('1'));
+ UnicodeSet casedLetter2 = new UnicodeSet("[:gc=cased letter:]");
+ assertNotEquals("gc=lc should not be equal", casedLetter, casedLetter2);
+ } finally {
+ // restore the world
+ UnicodeSet.setDefaultXSymbolTable(null);
+ }
+ }
+
+
+ /**
+ * For testing, override to set A-M to Cn.
+ */
+ static class MyUnicodeGCProperty extends UnicodeProperty.SimpleProperty {
+ UnicodeProperty icuProperty = ICUPropertyFactory.make().getProperty("Gc");
+ {
+ setName(icuProperty.getName());
+ setType(icuProperty.getType());
+ }
+ @Override
+ protected String _getValue(int codepoint) {
+ if (codepoint >= 'A' && codepoint <= 'M') {
+ return "Unassigned";
+ } else {
+ return icuProperty.getValue(codepoint);
+ }
+ }
+ @Override
+ protected List _getValueAliases(String valueAlias, List result) {
+ return icuProperty.getValueAliases(valueAlias, result);
+ }
+ @Override
+ public List _getNameAliases(List result) {
+ return icuProperty.getNameAliases();
+ }
+ }
+
+ /**
+ * For testing, override to set A-Z to Cn.
+ */
+ static class MyUnicodePropertyFactory extends ICUPropertyFactory {
+ private MyUnicodePropertyFactory() {
+ add(new MyUnicodeGCProperty());
+ }
+ }
+
+ static class MyUnicodePropertySymbolTable extends UnicodePropertySymbolTable {
+ public MyUnicodePropertySymbolTable(Factory factory) {
+ super(factory);
+ }
+ }
+}
-/*
\r *******************************************************************************
\r * Copyright (C) 1996-2011, Google, International Business Machines Corporation and *
\r * others. All Rights Reserved. *
\r *******************************************************************************
\r */
\rpackage com.ibm.icu.dev.test.util;
\r\rimport java.util.Comparator;
\rimport java.util.HashMap;
\rimport java.util.List;
\rimport java.util.Locale;
\rimport java.util.Set;
\r\rimport com.ibm.icu.dev.test.util.UnicodeProperty.PatternMatcher;
\rimport com.ibm.icu.impl.UnicodeRegex;
\rimport com.ibm.icu.text.UTF16;
\rimport com.ibm.icu.text.UnicodeSet;
\r\r/**
\r * Allows for overriding the parsing of UnicodeSet property patterns.
\r * <p>
\r * WARNING: If this UnicodePropertySymbolTable is used with {@code UnicodeSet.setDefaultXSymbolTable}, and the
\r * Unassigned characters (gc=Cn) are different than in ICU other than in ICU, you MUST call
\r * {@code UnicodeProperty.ResetCacheProperties} afterwards. If you then call {@code UnicodeSet.setDefaultXSymbolTable}
\r * with null to clear the value, you MUST also call {@code UnicodeProperty.ResetCacheProperties}.
\r *
\r * @author markdavis
\r */
\rpublic class UnicodePropertySymbolTable extends UnicodeSet.XSymbolTable {
\r UnicodeRegex unicodeRegex;
\r final UnicodeProperty.Factory factory;
\r\r public UnicodePropertySymbolTable(UnicodeProperty.Factory factory) {
\r unicodeRegex = new UnicodeRegex().setSymbolTable(this);
\r this.factory = factory;
\r }
\r\r\r // public boolean applyPropertyAlias0(String propertyName,
\r // String propertyValue, UnicodeSet result) {
\r // if (!propertyName.contains("*")) {
\r // return applyPropertyAlias(propertyName, propertyValue, result);
\r // }
\r // String[] propertyNames = propertyName.split("[*]");
\r // for (int i = propertyNames.length - 1; i >= 0; ++i) {
\r // String pname = propertyNames[i];
\r //
\r // }
\r // return null;
\r // }
\r\r public boolean applyPropertyAlias(String propertyName,
\r String propertyValue, UnicodeSet result) {
\r boolean status = false;
\r boolean invert = false;
\r int posNotEqual = propertyName.indexOf('\u2260');
\r int posColon = propertyName.indexOf(':');
\r if (posNotEqual >= 0 || posColon >= 0) {
\r if (posNotEqual < 0) posNotEqual = propertyName.length();
\r if (posColon < 0) posColon = propertyName.length();
\r int opPos = posNotEqual < posColon ? posNotEqual : posColon;
\r propertyValue = propertyValue.length() == 0 ? propertyName.substring(opPos+1)
\r : propertyName.substring(opPos+1) + "=" + propertyValue;
\r propertyName = propertyName.substring(0,opPos);
\r if (posNotEqual < posColon) {
\r invert = true;
\r }
\r }
\r if (propertyName.endsWith("!")) {
\r propertyName = propertyName.substring(0, propertyName.length() - 1);
\r invert = !invert;
\r }
\r propertyValue = propertyValue.trim();
\r if (propertyValue.length() != 0) {
\r status = applyPropertyAlias0(propertyName, propertyValue, result);
\r } else {
\r try {
\r status = applyPropertyAlias0("gc", propertyName, result);
\r } catch (Exception e) {};
\r if (!status) {
\r try {
\r status = applyPropertyAlias0("sc", propertyName, result);
\r } catch (Exception e) {};
\r if (!status) {
\r try {
\r status = applyPropertyAlias0(propertyName, "Yes", result);
\r } catch (Exception e) {};
\r if (!status) {
\r status = applyPropertyAlias0(propertyName, "", result);
\r }
\r }
\r }
\r }
\r if (status && invert) {
\r result.complement();
\r }
\r return status;
\r }
\r\r static final HashMap<String,String[]> GC_REMAP = new HashMap();
\r {
\r GC_REMAP.put("c", "Cc Cf Cn Co Cs".split(" "));
\r GC_REMAP.put("other", GC_REMAP.get("c"));
\r \r GC_REMAP.put("l", "Ll Lm Lo Lt Lu".split(" "));
\r GC_REMAP.put("letter", GC_REMAP.get("l"));
\r \r GC_REMAP.put("lc", "Ll Lt Lu".split(" "));
\r GC_REMAP.put("casedletter", GC_REMAP.get("lc"));
\r \r GC_REMAP.put("m", "Mc Me Mn".split(" "));
\r GC_REMAP.put("mark", GC_REMAP.get("m"));
\r \r GC_REMAP.put("n", "Nd Nl No".split(" "));
\r GC_REMAP.put("number", GC_REMAP.get("n"));
\r \r GC_REMAP.put("p", "Pc Pd Pe Pf Pi Po Ps".split(" "));
\r GC_REMAP.put("punctuation", GC_REMAP.get("p"));
\r GC_REMAP.put("punct", GC_REMAP.get("p"));
\r \r GC_REMAP.put("s", "Sc Sk Sm So".split(" "));
\r GC_REMAP.put("symbol", GC_REMAP.get("s"));
\r \r GC_REMAP.put("z", "Zl Zp Zs".split(" "));
\r GC_REMAP.put("separator", GC_REMAP.get("z"));
\r }
\r \r public boolean applyPropertyAlias0(String propertyName,
\r String propertyValue, UnicodeSet result) {
\r result.clear();
\r UnicodeProperty prop = factory.getProperty(propertyName);
\r String canonicalName = prop.getName();
\r boolean isAge = UnicodeProperty.equalNames("Age", canonicalName);
\r\r // Hack for special GC values
\r if (canonicalName.equals("General_Category")) {
\r String[] parts = GC_REMAP.get(UnicodeProperty.toSkeleton(propertyValue));
\r if (parts != null) {
\r for (String part : parts) {
\r prop.getSet(part, result);
\r }
\r return true;
\r }
\r }
\r\r PatternMatcher patternMatcher = null;
\r if (propertyValue.length() > 1 && propertyValue.startsWith("/") && propertyValue.endsWith("/")) {
\r String fixedRegex = unicodeRegex.transform(propertyValue.substring(1, propertyValue.length() - 1));
\r patternMatcher = new UnicodeProperty.RegexMatcher().set(fixedRegex);
\r }
\r UnicodeProperty otherProperty = null;
\r boolean testCp = false;
\r if (propertyValue.length() > 1 && propertyValue.startsWith("@") && propertyValue.endsWith("@")) {
\r String otherPropName = propertyValue.substring(1, propertyValue.length() - 1).trim();
\r if ("cp".equalsIgnoreCase(otherPropName)) {
\r testCp = true;
\r } else {
\r otherProperty = factory.getProperty(otherPropName);
\r }
\r }
\r if (prop != null) {
\r UnicodeSet set;
\r if (testCp) {
\r set = new UnicodeSet();
\r for (int i = 0; i <= 0x10FFFF; ++i) {
\r if (UnicodeProperty.equals(i, prop.getValue(i))) {
\r set.add(i);
\r }
\r }
\r } else if (otherProperty != null) {
\r set = new UnicodeSet();
\r for (int i = 0; i <= 0x10FFFF; ++i) {
\r String v1 = prop.getValue(i);
\r String v2 = otherProperty.getValue(i);
\r if (UnicodeProperty.equals(v1, v2)) {
\r set.add(i);
\r }
\r }
\r } else if (patternMatcher == null) {
\r if (!isValid(prop, propertyValue)) {
\r throw new IllegalArgumentException("The value '" + propertyValue + "' is illegal. Values for " + propertyName
\r + " must be in "
\r + prop.getAvailableValues() + " or in " + prop.getValueAliases());
\r }
\r if (isAge) {
\r set = prop.getSet(new ComparisonMatcher(propertyValue, Relation.geq));
\r } else {
\r set = prop.getSet(propertyValue);
\r }
\r } else if (isAge) {
\r set = new UnicodeSet();
\r List<String> values = prop.getAvailableValues();
\r for (String value : values) {
\r if (patternMatcher.matches(value)) {
\r for (String other : values) {
\r if (other.compareTo(value) <= 0) {
\r set.addAll(prop.getSet(other));
\r }
\r }
\r }
\r }
\r } else {
\r set = prop.getSet(patternMatcher);
\r }
\r result.addAll(set);
\r return true;
\r }
\r throw new IllegalArgumentException("Illegal property: " + propertyName);
\r }
\r\r \r\r private boolean isValid(UnicodeProperty prop, String propertyValue) {
\r// if (prop.getName().equals("General_Category")) {
\r// if (propertyValue)
\r// }
\r return prop.isValidValue(propertyValue);
\r }
\r\r public enum Relation {less, leq, equal, geq, greater}
\r\r public static class ComparisonMatcher implements PatternMatcher {
\r Relation relation;
\r static Comparator comparator = new UTF16.StringComparator(true, false,0);
\r\r String pattern;
\r\r public ComparisonMatcher(String pattern, Relation comparator) {
\r this.relation = comparator;
\r this.pattern = pattern;
\r }
\r\r public boolean matches(Object value) {
\r int comp = comparator.compare(pattern, value.toString());
\r switch (relation) {
\r case less: return comp < 0;
\r case leq: return comp <= 0;
\r default: return comp == 0;
\r case geq: return comp >= 0;
\r case greater: return comp > 0;
\r }
\r }
\r\r public PatternMatcher set(String pattern) {
\r this.pattern = pattern;
\r return this;
\r }
\r }
\r }
\ No newline at end of file
+/* *******************************************************************************
+ * Copyright (C) 1996-2011, Google, International Business Machines Corporation and *
+ * others. All Rights Reserved. *
+ *******************************************************************************
+ */
+package com.ibm.icu.dev.test.util;
+
+import java.util.Comparator;
+import java.util.HashMap;
+import java.util.List;
+import java.util.Locale;
+import java.util.Set;
+
+import com.ibm.icu.dev.test.util.UnicodeProperty.PatternMatcher;
+import com.ibm.icu.impl.UnicodeRegex;
+import com.ibm.icu.text.UTF16;
+import com.ibm.icu.text.UnicodeSet;
+
+/**
+ * Allows for overriding the parsing of UnicodeSet property patterns.
+ * <p>
+ * WARNING: If this UnicodePropertySymbolTable is used with {@code UnicodeSet.setDefaultXSymbolTable}, and the
+ * Unassigned characters (gc=Cn) are different than in ICU other than in ICU, you MUST call
+ * {@code UnicodeProperty.ResetCacheProperties} afterwards. If you then call {@code UnicodeSet.setDefaultXSymbolTable}
+ * with null to clear the value, you MUST also call {@code UnicodeProperty.ResetCacheProperties}.
+ *
+ * @author markdavis
+ */
+public class UnicodePropertySymbolTable extends UnicodeSet.XSymbolTable {
+ UnicodeRegex unicodeRegex;
+ final UnicodeProperty.Factory factory;
+
+ public UnicodePropertySymbolTable(UnicodeProperty.Factory factory) {
+ unicodeRegex = new UnicodeRegex().setSymbolTable(this);
+ this.factory = factory;
+ }
+
+
+ // public boolean applyPropertyAlias0(String propertyName,
+ // String propertyValue, UnicodeSet result) {
+ // if (!propertyName.contains("*")) {
+ // return applyPropertyAlias(propertyName, propertyValue, result);
+ // }
+ // String[] propertyNames = propertyName.split("[*]");
+ // for (int i = propertyNames.length - 1; i >= 0; ++i) {
+ // String pname = propertyNames[i];
+ //
+ // }
+ // return null;
+ // }
+
+ public boolean applyPropertyAlias(String propertyName,
+ String propertyValue, UnicodeSet result) {
+ boolean status = false;
+ boolean invert = false;
+ int posNotEqual = propertyName.indexOf('\u2260');
+ int posColon = propertyName.indexOf(':');
+ if (posNotEqual >= 0 || posColon >= 0) {
+ if (posNotEqual < 0) posNotEqual = propertyName.length();
+ if (posColon < 0) posColon = propertyName.length();
+ int opPos = posNotEqual < posColon ? posNotEqual : posColon;
+ propertyValue = propertyValue.length() == 0 ? propertyName.substring(opPos+1)
+ : propertyName.substring(opPos+1) + "=" + propertyValue;
+ propertyName = propertyName.substring(0,opPos);
+ if (posNotEqual < posColon) {
+ invert = true;
+ }
+ }
+ if (propertyName.endsWith("!")) {
+ propertyName = propertyName.substring(0, propertyName.length() - 1);
+ invert = !invert;
+ }
+ propertyValue = propertyValue.trim();
+ if (propertyValue.length() != 0) {
+ status = applyPropertyAlias0(propertyName, propertyValue, result);
+ } else {
+ try {
+ status = applyPropertyAlias0("gc", propertyName, result);
+ } catch (Exception e) {};
+ if (!status) {
+ try {
+ status = applyPropertyAlias0("sc", propertyName, result);
+ } catch (Exception e) {};
+ if (!status) {
+ try {
+ status = applyPropertyAlias0(propertyName, "Yes", result);
+ } catch (Exception e) {};
+ if (!status) {
+ status = applyPropertyAlias0(propertyName, "", result);
+ }
+ }
+ }
+ }
+ if (status && invert) {
+ result.complement();
+ }
+ return status;
+ }
+
+ static final HashMap<String,String[]> GC_REMAP = new HashMap();
+ {
+ GC_REMAP.put("c", "Cc Cf Cn Co Cs".split(" "));
+ GC_REMAP.put("other", GC_REMAP.get("c"));
+
+ GC_REMAP.put("l", "Ll Lm Lo Lt Lu".split(" "));
+ GC_REMAP.put("letter", GC_REMAP.get("l"));
+
+ GC_REMAP.put("lc", "Ll Lt Lu".split(" "));
+ GC_REMAP.put("casedletter", GC_REMAP.get("lc"));
+
+ GC_REMAP.put("m", "Mc Me Mn".split(" "));
+ GC_REMAP.put("mark", GC_REMAP.get("m"));
+
+ GC_REMAP.put("n", "Nd Nl No".split(" "));
+ GC_REMAP.put("number", GC_REMAP.get("n"));
+
+ GC_REMAP.put("p", "Pc Pd Pe Pf Pi Po Ps".split(" "));
+ GC_REMAP.put("punctuation", GC_REMAP.get("p"));
+ GC_REMAP.put("punct", GC_REMAP.get("p"));
+
+ GC_REMAP.put("s", "Sc Sk Sm So".split(" "));
+ GC_REMAP.put("symbol", GC_REMAP.get("s"));
+
+ GC_REMAP.put("z", "Zl Zp Zs".split(" "));
+ GC_REMAP.put("separator", GC_REMAP.get("z"));
+ }
+
+ public boolean applyPropertyAlias0(String propertyName,
+ String propertyValue, UnicodeSet result) {
+ result.clear();
+ UnicodeProperty prop = factory.getProperty(propertyName);
+ String canonicalName = prop.getName();
+ boolean isAge = UnicodeProperty.equalNames("Age", canonicalName);
+
+ // Hack for special GC values
+ if (canonicalName.equals("General_Category")) {
+ String[] parts = GC_REMAP.get(UnicodeProperty.toSkeleton(propertyValue));
+ if (parts != null) {
+ for (String part : parts) {
+ prop.getSet(part, result);
+ }
+ return true;
+ }
+ }
+
+ PatternMatcher patternMatcher = null;
+ if (propertyValue.length() > 1 && propertyValue.startsWith("/") && propertyValue.endsWith("/")) {
+ String fixedRegex = unicodeRegex.transform(propertyValue.substring(1, propertyValue.length() - 1));
+ patternMatcher = new UnicodeProperty.RegexMatcher().set(fixedRegex);
+ }
+ UnicodeProperty otherProperty = null;
+ boolean testCp = false;
+ if (propertyValue.length() > 1 && propertyValue.startsWith("@") && propertyValue.endsWith("@")) {
+ String otherPropName = propertyValue.substring(1, propertyValue.length() - 1).trim();
+ if ("cp".equalsIgnoreCase(otherPropName)) {
+ testCp = true;
+ } else {
+ otherProperty = factory.getProperty(otherPropName);
+ }
+ }
+ if (prop != null) {
+ UnicodeSet set;
+ if (testCp) {
+ set = new UnicodeSet();
+ for (int i = 0; i <= 0x10FFFF; ++i) {
+ if (UnicodeProperty.equals(i, prop.getValue(i))) {
+ set.add(i);
+ }
+ }
+ } else if (otherProperty != null) {
+ set = new UnicodeSet();
+ for (int i = 0; i <= 0x10FFFF; ++i) {
+ String v1 = prop.getValue(i);
+ String v2 = otherProperty.getValue(i);
+ if (UnicodeProperty.equals(v1, v2)) {
+ set.add(i);
+ }
+ }
+ } else if (patternMatcher == null) {
+ if (!isValid(prop, propertyValue)) {
+ throw new IllegalArgumentException("The value '" + propertyValue + "' is illegal. Values for " + propertyName
+ + " must be in "
+ + prop.getAvailableValues() + " or in " + prop.getValueAliases());
+ }
+ if (isAge) {
+ set = prop.getSet(new ComparisonMatcher(propertyValue, Relation.geq));
+ } else {
+ set = prop.getSet(propertyValue);
+ }
+ } else if (isAge) {
+ set = new UnicodeSet();
+ List<String> values = prop.getAvailableValues();
+ for (String value : values) {
+ if (patternMatcher.matches(value)) {
+ for (String other : values) {
+ if (other.compareTo(value) <= 0) {
+ set.addAll(prop.getSet(other));
+ }
+ }
+ }
+ }
+ } else {
+ set = prop.getSet(patternMatcher);
+ }
+ result.addAll(set);
+ return true;
+ }
+ throw new IllegalArgumentException("Illegal property: " + propertyName);
+ }
+
+
+
+ private boolean isValid(UnicodeProperty prop, String propertyValue) {
+// if (prop.getName().equals("General_Category")) {
+// if (propertyValue)
+// }
+ return prop.isValidValue(propertyValue);
+ }
+
+ public enum Relation {less, leq, equal, geq, greater}
+
+ public static class ComparisonMatcher implements PatternMatcher {
+ Relation relation;
+ static Comparator comparator = new UTF16.StringComparator(true, false,0);
+
+ String pattern;
+
+ public ComparisonMatcher(String pattern, Relation comparator) {
+ this.relation = comparator;
+ this.pattern = pattern;
+ }
+
+ public boolean matches(Object value) {
+ int comp = comparator.compare(pattern, value.toString());
+ switch (relation) {
+ case less: return comp < 0;
+ case leq: return comp <= 0;
+ default: return comp == 0;
+ case geq: return comp >= 0;
+ case greater: return comp > 0;
+ }
+ }
+
+ public PatternMatcher set(String pattern) {
+ this.pattern = pattern;
+ return this;
+ }
+ }
+ }
\ No newline at end of file
projects / icu / commitdiff