Victor Chang | 7322950 | 2020-09-17 13:39:19 +0100 | [diff] [blame] | 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 | /** Returns a bogus UnicodeString by value. */ |
| 50 | static inline UnicodeString makeBogusString() { |
| 51 | UnicodeString result; |
| 52 | result.setToBogus(); |
| 53 | return result; |
| 54 | } |
| 55 | |
| 56 | /** |
| 57 | * Return true if the character is NOT printable ASCII. |
| 58 | * |
| 59 | * This method should really be in UnicodeString (or similar). For |
| 60 | * now, we implement it here and share it with friend classes. |
| 61 | */ |
| 62 | static UBool isUnprintable(UChar32 c); |
| 63 | |
| 64 | /** |
| 65 | * Escape unprintable characters using \uxxxx notation for U+0000 to |
| 66 | * U+FFFF and \Uxxxxxxxx for U+10000 and above. If the character is |
Victor Chang | ce4bf3c | 2021-01-19 16:34:24 +0000 | [diff] [blame] | 67 | * printable ASCII, then do nothing and return false. Otherwise, |
| 68 | * append the escaped notation and return true. |
Victor Chang | 7322950 | 2020-09-17 13:39:19 +0100 | [diff] [blame] | 69 | */ |
| 70 | static UBool escapeUnprintable(UnicodeString& result, UChar32 c); |
| 71 | |
| 72 | /** |
| 73 | * Returns the index of a character, ignoring quoted text. |
| 74 | * For example, in the string "abc'hide'h", the 'h' in "hide" will not be |
| 75 | * found by a search for 'h'. |
| 76 | * @param text text to be searched |
| 77 | * @param start the beginning index, inclusive; <code>0 <= start |
| 78 | * <= limit</code>. |
| 79 | * @param limit the ending index, exclusive; <code>start <= limit |
| 80 | * <= text.length()</code>. |
| 81 | * @param c character to search for |
| 82 | * @return Offset of the first instance of c, or -1 if not found. |
| 83 | */ |
| 84 | //?FOR FUTURE USE. DISABLE FOR NOW for coverage reasons. |
| 85 | // static int32_t quotedIndexOf(const UnicodeString& text, |
| 86 | // int32_t start, int32_t limit, |
| 87 | // UChar c); |
| 88 | |
| 89 | /** |
| 90 | * Skip over a sequence of zero or more white space characters at pos. |
| 91 | * @param advance if true, advance pos to the first non-white-space |
| 92 | * character at or after pos, or str.length(), if there is none. |
| 93 | * Otherwise leave pos unchanged. |
| 94 | * @return the index of the first non-white-space character at or |
| 95 | * after pos, or str.length(), if there is none. |
| 96 | */ |
| 97 | static int32_t skipWhitespace(const UnicodeString& str, int32_t& pos, |
Victor Chang | ce4bf3c | 2021-01-19 16:34:24 +0000 | [diff] [blame] | 98 | UBool advance = false); |
Victor Chang | 7322950 | 2020-09-17 13:39:19 +0100 | [diff] [blame] | 99 | |
| 100 | /** |
| 101 | * Skip over Pattern_White_Space in a Replaceable. |
| 102 | * Skipping may be done in the forward or |
| 103 | * reverse direction. In either case, the leftmost index will be |
| 104 | * inclusive, and the rightmost index will be exclusive. That is, |
| 105 | * given a range defined as [start, limit), the call |
| 106 | * skipWhitespace(text, start, limit) will advance start past leading |
| 107 | * whitespace, whereas the call skipWhitespace(text, limit, start), |
| 108 | * will back up limit past trailing whitespace. |
| 109 | * @param text the text to be analyzed |
| 110 | * @param pos either the start or limit of a range of 'text', to skip |
| 111 | * leading or trailing whitespace, respectively |
| 112 | * @param stop either the limit or start of a range of 'text', to skip |
| 113 | * leading or trailing whitespace, respectively |
| 114 | * @return the new start or limit, depending on what was passed in to |
| 115 | * 'pos' |
| 116 | */ |
| 117 | //?FOR FUTURE USE. DISABLE FOR NOW for coverage reasons. |
| 118 | //? static int32_t skipWhitespace(const Replaceable& text, |
| 119 | //? int32_t pos, int32_t stop); |
| 120 | |
| 121 | /** |
| 122 | * Parse a single non-whitespace character 'ch', optionally |
| 123 | * preceded by whitespace. |
| 124 | * @param id the string to be parsed |
| 125 | * @param pos INPUT-OUTPUT parameter. On input, pos[0] is the |
| 126 | * offset of the first character to be parsed. On output, pos[0] |
| 127 | * is the index after the last parsed character. If the parse |
| 128 | * fails, pos[0] will be unchanged. |
| 129 | * @param ch the non-whitespace character to be parsed. |
| 130 | * @return true if 'ch' is seen preceded by zero or more |
| 131 | * whitespace characters. |
| 132 | */ |
| 133 | static UBool parseChar(const UnicodeString& id, int32_t& pos, UChar ch); |
| 134 | |
| 135 | /** |
| 136 | * Parse a pattern string starting at offset pos. Keywords are |
| 137 | * matched case-insensitively. Spaces may be skipped and may be |
| 138 | * optional or required. Integer values may be parsed, and if |
| 139 | * they are, they will be returned in the given array. If |
| 140 | * successful, the offset of the next non-space character is |
| 141 | * returned. On failure, -1 is returned. |
| 142 | * @param pattern must only contain lowercase characters, which |
| 143 | * will match their uppercase equivalents as well. A space |
| 144 | * character matches one or more required spaces. A '~' character |
| 145 | * matches zero or more optional spaces. A '#' character matches |
| 146 | * an integer and stores it in parsedInts, which the caller must |
| 147 | * ensure has enough capacity. |
| 148 | * @param parsedInts array to receive parsed integers. Caller |
| 149 | * must ensure that parsedInts.length is >= the number of '#' |
| 150 | * signs in 'pattern'. |
| 151 | * @return the position after the last character parsed, or -1 if |
| 152 | * the parse failed |
| 153 | */ |
| 154 | static int32_t parsePattern(const UnicodeString& rule, int32_t pos, int32_t limit, |
| 155 | const UnicodeString& pattern, int32_t* parsedInts); |
| 156 | |
| 157 | /** |
| 158 | * Parse a pattern string within the given Replaceable and a parsing |
| 159 | * pattern. Characters are matched literally and case-sensitively |
| 160 | * except for the following special characters: |
| 161 | * |
| 162 | * ~ zero or more Pattern_White_Space chars |
| 163 | * |
| 164 | * If end of pattern is reached with all matches along the way, |
| 165 | * pos is advanced to the first unparsed index and returned. |
| 166 | * Otherwise -1 is returned. |
| 167 | * @param pat pattern that controls parsing |
| 168 | * @param text text to be parsed, starting at index |
| 169 | * @param index offset to first character to parse |
| 170 | * @param limit offset after last character to parse |
| 171 | * @return index after last parsed character, or -1 on parse failure. |
| 172 | */ |
| 173 | static int32_t parsePattern(const UnicodeString& pat, |
| 174 | const Replaceable& text, |
| 175 | int32_t index, |
| 176 | int32_t limit); |
| 177 | |
| 178 | /** |
| 179 | * Parse an integer at pos, either of the form \d+ or of the form |
| 180 | * 0x[0-9A-Fa-f]+ or 0[0-7]+, that is, in standard decimal, hex, |
| 181 | * or octal format. |
| 182 | * @param pos INPUT-OUTPUT parameter. On input, the index of the first |
| 183 | * character to parse. On output, the index of the character after the |
| 184 | * last parsed character. |
| 185 | */ |
| 186 | static int32_t parseInteger(const UnicodeString& rule, int32_t& pos, int32_t limit); |
| 187 | |
| 188 | /** |
| 189 | * Parse an integer at pos using only ASCII digits. |
| 190 | * Base 10 only. |
| 191 | * @param pos INPUT-OUTPUT parameter. On input, the index of the first |
| 192 | * character to parse. On output, the index of the character after the |
| 193 | * last parsed character. |
| 194 | */ |
| 195 | static int32_t parseAsciiInteger(const UnicodeString& str, int32_t& pos); |
| 196 | |
| 197 | /** |
| 198 | * Parse a Unicode identifier from the given string at the given |
| 199 | * position. Return the identifier, or an empty string if there |
| 200 | * is no identifier. |
| 201 | * @param str the string to parse |
| 202 | * @param pos INPUT-OUPUT parameter. On INPUT, pos is the |
| 203 | * first character to examine. It must be less than str.length(), |
| 204 | * and it must not point to a whitespace character. That is, must |
| 205 | * have pos < str.length() and |
| 206 | * !UCharacter::isWhitespace(str.char32At(pos)). On |
| 207 | * OUTPUT, the position after the last parsed character. |
| 208 | * @return the Unicode identifier, or an empty string if there is |
| 209 | * no valid identifier at pos. |
| 210 | */ |
| 211 | static UnicodeString parseUnicodeIdentifier(const UnicodeString& str, int32_t& pos); |
| 212 | |
| 213 | /** |
| 214 | * Parse an unsigned 31-bit integer at the given offset. Use |
| 215 | * UCharacter.digit() to parse individual characters into digits. |
| 216 | * @param text the text to be parsed |
| 217 | * @param pos INPUT-OUTPUT parameter. On entry, pos is the |
| 218 | * offset within text at which to start parsing; it should point |
| 219 | * to a valid digit. On exit, pos is the offset after the last |
| 220 | * parsed character. If the parse failed, it will be unchanged on |
| 221 | * exit. Must be >= 0 on entry. |
| 222 | * @param radix the radix in which to parse; must be >= 2 and <= |
| 223 | * 36. |
| 224 | * @return a non-negative parsed number, or -1 upon parse failure. |
| 225 | * Parse fails if there are no digits, that is, if pos does not |
| 226 | * point to a valid digit on entry, or if the number to be parsed |
| 227 | * does not fit into a 31-bit unsigned integer. |
| 228 | */ |
| 229 | static int32_t parseNumber(const UnicodeString& text, |
| 230 | int32_t& pos, int8_t radix); |
| 231 | |
| 232 | static void appendToRule(UnicodeString& rule, |
| 233 | UChar32 c, |
| 234 | UBool isLiteral, |
| 235 | UBool escapeUnprintable, |
| 236 | UnicodeString& quoteBuf); |
| 237 | |
| 238 | static void appendToRule(UnicodeString& rule, |
| 239 | const UnicodeString& text, |
| 240 | UBool isLiteral, |
| 241 | UBool escapeUnprintable, |
| 242 | UnicodeString& quoteBuf); |
| 243 | |
| 244 | static void appendToRule(UnicodeString& rule, |
| 245 | const UnicodeMatcher* matcher, |
| 246 | UBool escapeUnprintable, |
| 247 | UnicodeString& quoteBuf); |
| 248 | |
| 249 | private: |
| 250 | // do not instantiate |
| 251 | ICU_Utility(); |
| 252 | }; |
| 253 | |
| 254 | U_NAMESPACE_END |
| 255 | |
| 256 | #endif |
| 257 | //eof |