1 // 2016 and later: Unicode, Inc. and others. 2 // License & terms of use: http://www.unicode.org/copyright.html 3 /* 4 ********************************************************************** 5 * Copyright (c) 2001-2011, International Business Machines 6 * Corporation and others. All Rights Reserved. 7 ********************************************************************** 8 * Date Name Description 9 * 11/19/2001 aliu Creation. 10 ********************************************************************** 11 */ 12 13 #ifndef ICU_UTIL_H 14 #define ICU_UTIL_H 15 16 #include "unicode/utypes.h" 17 #include "unicode/uobject.h" 18 #include "unicode/unistr.h" 19 20 //-------------------------------------------------------------------- 21 // class ICU_Utility 22 // i18n utility functions, scoped into the class ICU_Utility. 23 //-------------------------------------------------------------------- 24 25 U_NAMESPACE_BEGIN 26 27 class UnicodeMatcher; 28 29 class U_COMMON_API ICU_Utility /* not : public UObject because all methods are static */ { 30 public: 31 32 /** 33 * Append a number to the given UnicodeString in the given radix. 34 * Standard digits '0'-'9' are used and letters 'A'-'Z' for 35 * radices 11 through 36. 36 * @param result the digits of the number are appended here 37 * @param n the number to be converted to digits; may be negative. 38 * If negative, a '-' is prepended to the digits. 39 * @param radix a radix from 2 to 36 inclusive. 40 * @param minDigits the minimum number of digits, not including 41 * any '-', to produce. Values less than 2 have no effect. One 42 * digit is always emitted regardless of this parameter. 43 * @return a reference to result 44 */ 45 static UnicodeString& appendNumber(UnicodeString& result, int32_t n, 46 int32_t radix = 10, 47 int32_t minDigits = 1); 48 49 /** 50 * Return true if the character is NOT printable ASCII. 51 * 52 * This method should really be in UnicodeString (or similar). For 53 * now, we implement it here and share it with friend classes. 54 */ 55 static UBool isUnprintable(UChar32 c); 56 57 /** 58 * Escape unprintable characters using \uxxxx notation for U+0000 to 59 * U+FFFF and \Uxxxxxxxx for U+10000 and above. If the character is 60 * printable ASCII, then do nothing and return FALSE. Otherwise, 61 * append the escaped notation and return TRUE. 62 */ 63 static UBool escapeUnprintable(UnicodeString& result, UChar32 c); 64 65 /** 66 * Returns the index of a character, ignoring quoted text. 67 * For example, in the string "abc'hide'h", the 'h' in "hide" will not be 68 * found by a search for 'h'. 69 * @param text text to be searched 70 * @param start the beginning index, inclusive; <code>0 <= start 71 * <= limit</code>. 72 * @param limit the ending index, exclusive; <code>start <= limit 73 * <= text.length()</code>. 74 * @param c character to search for 75 * @return Offset of the first instance of c, or -1 if not found. 76 */ 77 //?FOR FUTURE USE. DISABLE FOR NOW for coverage reasons. 78 // static int32_t quotedIndexOf(const UnicodeString& text, 79 // int32_t start, int32_t limit, 80 // UChar c); 81 82 /** 83 * Skip over a sequence of zero or more white space characters at pos. 84 * @param advance if true, advance pos to the first non-white-space 85 * character at or after pos, or str.length(), if there is none. 86 * Otherwise leave pos unchanged. 87 * @return the index of the first non-white-space character at or 88 * after pos, or str.length(), if there is none. 89 */ 90 static int32_t skipWhitespace(const UnicodeString& str, int32_t& pos, 91 UBool advance = FALSE); 92 93 /** 94 * Skip over Pattern_White_Space in a Replaceable. 95 * Skipping may be done in the forward or 96 * reverse direction. In either case, the leftmost index will be 97 * inclusive, and the rightmost index will be exclusive. That is, 98 * given a range defined as [start, limit), the call 99 * skipWhitespace(text, start, limit) will advance start past leading 100 * whitespace, whereas the call skipWhitespace(text, limit, start), 101 * will back up limit past trailing whitespace. 102 * @param text the text to be analyzed 103 * @param pos either the start or limit of a range of 'text', to skip 104 * leading or trailing whitespace, respectively 105 * @param stop either the limit or start of a range of 'text', to skip 106 * leading or trailing whitespace, respectively 107 * @return the new start or limit, depending on what was passed in to 108 * 'pos' 109 */ 110 //?FOR FUTURE USE. DISABLE FOR NOW for coverage reasons. 111 //? static int32_t skipWhitespace(const Replaceable& text, 112 //? int32_t pos, int32_t stop); 113 114 /** 115 * Parse a single non-whitespace character 'ch', optionally 116 * preceded by whitespace. 117 * @param id the string to be parsed 118 * @param pos INPUT-OUTPUT parameter. On input, pos[0] is the 119 * offset of the first character to be parsed. On output, pos[0] 120 * is the index after the last parsed character. If the parse 121 * fails, pos[0] will be unchanged. 122 * @param ch the non-whitespace character to be parsed. 123 * @return true if 'ch' is seen preceded by zero or more 124 * whitespace characters. 125 */ 126 static UBool parseChar(const UnicodeString& id, int32_t& pos, UChar ch); 127 128 /** 129 * Parse a pattern string starting at offset pos. Keywords are 130 * matched case-insensitively. Spaces may be skipped and may be 131 * optional or required. Integer values may be parsed, and if 132 * they are, they will be returned in the given array. If 133 * successful, the offset of the next non-space character is 134 * returned. On failure, -1 is returned. 135 * @param pattern must only contain lowercase characters, which 136 * will match their uppercase equivalents as well. A space 137 * character matches one or more required spaces. A '~' character 138 * matches zero or more optional spaces. A '#' character matches 139 * an integer and stores it in parsedInts, which the caller must 140 * ensure has enough capacity. 141 * @param parsedInts array to receive parsed integers. Caller 142 * must ensure that parsedInts.length is >= the number of '#' 143 * signs in 'pattern'. 144 * @return the position after the last character parsed, or -1 if 145 * the parse failed 146 */ 147 static int32_t parsePattern(const UnicodeString& rule, int32_t pos, int32_t limit, 148 const UnicodeString& pattern, int32_t* parsedInts); 149 150 /** 151 * Parse a pattern string within the given Replaceable and a parsing 152 * pattern. Characters are matched literally and case-sensitively 153 * except for the following special characters: 154 * 155 * ~ zero or more Pattern_White_Space chars 156 * 157 * If end of pattern is reached with all matches along the way, 158 * pos is advanced to the first unparsed index and returned. 159 * Otherwise -1 is returned. 160 * @param pat pattern that controls parsing 161 * @param text text to be parsed, starting at index 162 * @param index offset to first character to parse 163 * @param limit offset after last character to parse 164 * @return index after last parsed character, or -1 on parse failure. 165 */ 166 static int32_t parsePattern(const UnicodeString& pat, 167 const Replaceable& text, 168 int32_t index, 169 int32_t limit); 170 171 /** 172 * Parse an integer at pos, either of the form \d+ or of the form 173 * 0x[0-9A-Fa-f]+ or 0[0-7]+, that is, in standard decimal, hex, 174 * or octal format. 175 * @param pos INPUT-OUTPUT parameter. On input, the first 176 * character to parse. On output, the character after the last 177 * parsed character. 178 */ 179 static int32_t parseInteger(const UnicodeString& rule, int32_t& pos, int32_t limit); 180 181 /** 182 * Parse a Unicode identifier from the given string at the given 183 * position. Return the identifier, or an empty string if there 184 * is no identifier. 185 * @param str the string to parse 186 * @param pos INPUT-OUPUT parameter. On INPUT, pos is the 187 * first character to examine. It must be less than str.length(), 188 * and it must not point to a whitespace character. That is, must 189 * have pos < str.length() and 190 * !UCharacter::isWhitespace(str.char32At(pos)). On 191 * OUTPUT, the position after the last parsed character. 192 * @return the Unicode identifier, or an empty string if there is 193 * no valid identifier at pos. 194 */ 195 static UnicodeString parseUnicodeIdentifier(const UnicodeString& str, int32_t& pos); 196 197 /** 198 * Parse an unsigned 31-bit integer at the given offset. Use 199 * UCharacter.digit() to parse individual characters into digits. 200 * @param text the text to be parsed 201 * @param pos INPUT-OUTPUT parameter. On entry, pos is the 202 * offset within text at which to start parsing; it should point 203 * to a valid digit. On exit, pos is the offset after the last 204 * parsed character. If the parse failed, it will be unchanged on 205 * exit. Must be >= 0 on entry. 206 * @param radix the radix in which to parse; must be >= 2 and <= 207 * 36. 208 * @return a non-negative parsed number, or -1 upon parse failure. 209 * Parse fails if there are no digits, that is, if pos does not 210 * point to a valid digit on entry, or if the number to be parsed 211 * does not fit into a 31-bit unsigned integer. 212 */ 213 static int32_t parseNumber(const UnicodeString& text, 214 int32_t& pos, int8_t radix); 215 216 static void appendToRule(UnicodeString& rule, 217 UChar32 c, 218 UBool isLiteral, 219 UBool escapeUnprintable, 220 UnicodeString& quoteBuf); 221 222 static void appendToRule(UnicodeString& rule, 223 const UnicodeString& text, 224 UBool isLiteral, 225 UBool escapeUnprintable, 226 UnicodeString& quoteBuf); 227 228 static void appendToRule(UnicodeString& rule, 229 const UnicodeMatcher* matcher, 230 UBool escapeUnprintable, 231 UnicodeString& quoteBuf); 232 233 private: 234 // do not instantiate 235 ICU_Utility(); 236 }; 237 238 U_NAMESPACE_END 239 240 #endif 241 //eof 242
