jars/icu4j-4_2_1-src/src/com/ibm/icu/dev/test/translit/RegexUtilitiesTest.java

   1 //##header\r
   2 //#if defined(FOUNDATION10) || defined(J2SE13) || defined(J2SE14)\r
   3 //#else\r
   4 /*\r
   5  *******************************************************************************\r
   6  * Copyright (C) 2009, International Business Machines Corporation and         *\r
   7  * others. All Rights Reserved.                                                *\r
   8  *******************************************************************************\r
   9  */\r
  10 package com.ibm.icu.dev.test.translit;\r
  11 \r
  12 import java.util.ArrayList;\r
  13 import java.util.List;\r
  14 import java.util.regex.Matcher;\r
  15 import java.util.regex.Pattern;\r
  16 \r
  17 import com.ibm.icu.dev.test.TestFmwk;\r
  18 import com.ibm.icu.impl.UnicodeRegex;\r
  19 import com.ibm.icu.lang.UCharacter;\r
  20 import com.ibm.icu.lang.UProperty;\r
  21 import com.ibm.icu.lang.UProperty.NameChoice;\r
  22 import com.ibm.icu.text.Transliterator;\r
  23 import com.ibm.icu.text.UTF16;\r
  24 import com.ibm.icu.text.UnicodeSet;\r
  25 \r
  26 /**\r
  27  * @author markdavis\r
  28  */\r
  29 public class RegexUtilitiesTest extends TestFmwk {\r
  30 \r
  31     public static void main(String[] args) throws Exception {\r
  32         new RegexUtilitiesTest().run(args);\r
  33     }\r
  34 \r
  35     /**\r
  36      * Check basic construction.\r
  37      */\r
  38     public void TestConstruction() {\r
  39         String[][] tests = {\r
  40                 {"a"},\r
  41                 {"a[a-z]b"},\r
  42                 {"[ba-z]", "[a-z]"},\r
  43                 {"q[ba-z]", "q[a-z]"},\r
  44                 {"[ba-z]q", "[a-z]q"},\r
  45                 {"a\\p{joincontrol}b", "a[\u200C\u200D]b"},\r
  46                 {"a\\P{joincontrol}b", "a[^\u200C\u200D]b"},\r
  47                 {"a[[:whitespace:]&[:Zl:]]b", "a[\\\u2028]b"},\r
  48                 {"a [[:bc=cs:]&[:wspace:]] b", "a [\u00A0\u202F] b"},\r
  49         };\r
  50         for (int i = 0; i < tests.length; ++i) {\r
  51             final String source = tests[i][0];\r
  52             String expected = tests[i].length == 1 ? source : tests[i][1];\r
  53             String actual = UnicodeRegex.fix(source);\r
  54             assertEquals(source, expected, actual);\r
  55         } \r
  56     }\r
  57 \r
  58     Transliterator hex = Transliterator.getInstance("hex");\r
  59 \r
  60     /**\r
  61      * Perform an exhaustive test on all Unicode characters to make sure that the UnicodeSet with each\r
  62      * character works.\r
  63      */\r
  64     public void TestCharacters() {\r
  65         UnicodeSet requiresQuote = new UnicodeSet("[\\$\\&\\-\\:\\[\\\\\\]\\^\\{\\}[:pattern_whitespace:]]");\r
  66         boolean skip = getInclusion() < 10;\r
  67         for (int cp = 0; cp < 0x110000; ++cp) {\r
  68             if (cp > 0xFF && skip && (cp % 37 != 0)) {\r
  69                 continue;\r
  70             }\r
  71             String cpString = UTF16.valueOf(cp);\r
  72             String s = requiresQuote.contains(cp) ? "\\" + cpString : cpString;\r
  73             String pattern = null;\r
  74             final String rawPattern = "[" + s + s + "]";\r
  75             try {\r
  76                 pattern = UnicodeRegex.fix(rawPattern);\r
  77             } catch (Exception e) {\r
  78                 errln(e.getMessage());\r
  79                 continue;\r
  80             }\r
  81             final String expected = "[" + s + "]";\r
  82             assertEquals("Doubled character works" + hex.transform(s), expected, pattern);\r
  83 \r
  84             // verify that we can create a regex pattern and use as expected\r
  85             String shouldNotMatch = UTF16.valueOf((cp + 1) % 0x110000);\r
  86             checkCharPattern(Pattern.compile(pattern), pattern, cpString, shouldNotMatch);\r
  87 \r
  88             // verify that the Pattern.compile works\r
  89             checkCharPattern(UnicodeRegex.compile(rawPattern), pattern, cpString, shouldNotMatch);\r
  90         }\r
  91     }\r
  92 \r
  93     /**\r
  94      * Check all integer Unicode properties to make sure they work.\r
  95      */\r
  96     public void TestUnicodeProperties() {\r
  97         final boolean skip = getInclusion() < 10;\r
  98         UnicodeSet temp = new UnicodeSet();\r
  99         for (int propNum = UProperty.INT_START; propNum < UProperty.INT_LIMIT; ++propNum) {\r
 100             if (skip && (propNum % 5 != 0)) {\r
 101                 continue;\r
 102             }\r
 103             String propName = UCharacter.getPropertyName(propNum, NameChoice.LONG);\r
 104             final int intPropertyMinValue = UCharacter.getIntPropertyMinValue(propNum);\r
 105             int intPropertyMaxValue = UCharacter.getIntPropertyMaxValue(propNum);\r
 106             if (skip) { // only test first if not exhaustive\r
 107                 intPropertyMaxValue = intPropertyMinValue;\r
 108             }\r
 109             for (int valueNum = intPropertyMinValue; valueNum <= intPropertyMaxValue; ++valueNum) {\r
 110                 // hack for getting property value name\r
 111                 String valueName = UCharacter.getPropertyValueName(propNum, valueNum, NameChoice.LONG);\r
 112                 if (valueName == null) {\r
 113                     valueName = UCharacter.getPropertyValueName(propNum, valueNum, NameChoice.SHORT);\r
 114                     if (valueName == null) {\r
 115                         valueName = Integer.toString(valueNum);\r
 116                     }\r
 117                 }\r
 118                 temp.applyIntPropertyValue(propNum, valueNum);\r
 119                 if (temp.size() == 0) {\r
 120                     continue;\r
 121                 }\r
 122                 final String prefix = "a";\r
 123                 final String suffix = "b";\r
 124                 String shouldMatch = prefix + UTF16.valueOf(temp.charAt(0)) + suffix;\r
 125                 temp.complement();\r
 126                 String shouldNotMatch = prefix + UTF16.valueOf(temp.charAt(0)) + suffix;\r
 127 \r
 128                 // posix style pattern\r
 129                 String rawPattern = prefix + "[:" + propName + "=" + valueName + ":]" + suffix;\r
 130                 String rawNegativePattern = prefix + "[:^" + propName + "=" + valueName + ":]" + suffix;\r
 131                 checkCharPattern(UnicodeRegex.compile(rawPattern), rawPattern, shouldMatch, shouldNotMatch);\r
 132                 checkCharPattern(UnicodeRegex.compile(rawNegativePattern), rawNegativePattern, shouldNotMatch, shouldMatch);\r
 133 \r
 134                 // perl style pattern\r
 135                 rawPattern = prefix + "\\p{" + propName + "=" + valueName + "}" + suffix;\r
 136                 rawNegativePattern = prefix + "\\P{" + propName + "=" + valueName + "}" + suffix;\r
 137                 checkCharPattern(UnicodeRegex.compile(rawPattern), rawPattern, shouldMatch, shouldNotMatch);\r
 138                 checkCharPattern(UnicodeRegex.compile(rawNegativePattern), rawNegativePattern, shouldNotMatch, shouldMatch);\r
 139             }\r
 140         }\r
 141     }\r
 142 \r
 143     public void TestBnf() {\r
 144         UnicodeRegex regex = new UnicodeRegex();\r
 145         final String[][] tests = {\r
 146                 {\r
 147                     "c = a wq;\n" +\r
 148                     "a = xyz;\n" +\r
 149                     "b = a a c;\n"\r
 150                 },\r
 151                 {\r
 152                     "c = a b;\n" +\r
 153                     "a = xyz;\n" +\r
 154                     "b = a a c;\n",\r
 155                     "Exception"\r
 156                 },\r
 157                 {\r
 158                     "uri = (?: (scheme) \\:)? (host) (?: \\? (query))? (?: \\u0023 (fragment))?;\n" +\r
 159                     "scheme = reserved+;\n" +\r
 160                     "host = // reserved+;\n" +\r
 161                     "query = [\\=reserved]+;\n" +\r
 162                     "fragment = reserved+;\n" +\r
 163                     "reserved = [[:ascii:][:sc=grek:]&[:alphabetic:]];\n",\r
 164                 "http://\u03B1\u03B2\u03B3?huh=hi#there"},\r
 165                 {\r
 166                     "langtagRegex.txt"\r
 167                 }\r
 168         };\r
 169         for (int i = 0; i < tests.length; ++i) {\r
 170             String test = tests[i][0];\r
 171             final boolean expectException = tests[i].length < 2 ? false : tests[i][1].equals("Exception");\r
 172             try {\r
 173                 String result;\r
 174                 if (test.endsWith(".txt")) {\r
 175                     java.io.InputStream is = RegexUtilitiesTest.class.getResourceAsStream(test);\r
 176                     List lines = UnicodeRegex.appendLines(new ArrayList(), is, "UTF-8");\r
 177                     result = regex.compileBnf(lines);\r
 178                 } else {\r
 179                     result = regex.compileBnf(test);\r
 180                 }\r
 181                 if (expectException) {\r
 182                     errln("Expected exception for " + test);\r
 183                     continue;\r
 184                 }\r
 185                 result = result.replaceAll("[0-9]+%", ""); // just so we can use the language subtag stuff\r
 186                 String resolved = regex.transform(result);\r
 187                 logln(resolved);\r
 188                 Matcher m = Pattern.compile(resolved, Pattern.COMMENTS).matcher("");\r
 189                 String checks = "";\r
 190                 for (int j = 1; j < tests[i].length; ++j) {\r
 191                     String check = tests[i][j];\r
 192                     if (!m.reset(check).matches()) {\r
 193                         checks = checks + "Fails " + check + "\n";\r
 194                     } else {\r
 195                         for (int k = 1; k <= m.groupCount(); ++k) {\r
 196                             checks += "(" + m.group(k) + ")";\r
 197                         }\r
 198                         checks += "\n";\r
 199                     }\r
 200                 }\r
 201                 logln("Result: " + result + "\n" + checks + "\n" + test);\r
 202             } catch (Exception e) {\r
 203                 if (!expectException) {\r
 204                     errln(e.getClass().getName() + ": " + e.getMessage());\r
 205                 }\r
 206                 continue;\r
 207             }\r
 208         }\r
 209     }\r
 210 \r
 211     /**\r
 212      * Utility for checking patterns\r
 213      */\r
 214     private void checkCharPattern(Pattern pat, String matchTitle, String shouldMatch, String shouldNotMatch) {\r
 215         Matcher matcher = pat.matcher(shouldMatch);\r
 216         assertTrue(matchTitle + " and " + shouldMatch, matcher.matches());\r
 217         matcher.reset(shouldNotMatch);\r
 218         assertFalse(matchTitle + " and " + shouldNotMatch, matcher.matches());\r
 219     }\r
 220 }\r
 221 //#endif\r