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 "charstr.h" |
17 | #include "unicode/unistr.h" |
18 | #include "unicode/uobject.h" |
19 | #include "unicode/utypes.h" |
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 | * The tab, newline and linefeed characters are considered unprintable. |
59 | */ |
60 | static UBool isUnprintable(UChar32 c); |
61 | |
62 | /** |
63 | * @return true for control codes and for surrogate and noncharacter code points |
64 | */ |
65 | static UBool shouldAlwaysBeEscaped(UChar32 c); |
66 | |
67 | /** |
68 | * Escapes one unprintable code point using \uxxxx notation for U+0000 to |
69 | * U+FFFF and \Uxxxxxxxx for U+10000 and above. If the character is |
70 | * printable ASCII, then do nothing and return false. Otherwise, |
71 | * append the escaped notation and return true. |
72 | */ |
73 | static UBool escapeUnprintable(UnicodeString& result, UChar32 c); |
74 | |
75 | /** |
76 | * Escapes one code point using \uxxxx notation |
77 | * for U+0000 to U+FFFF and \Uxxxxxxxx for U+10000 and above. |
78 | * @return result |
79 | */ |
80 | static UnicodeString &escape(UnicodeString& result, UChar32 c); |
81 | |
82 | /** |
83 | * Returns the index of a character, ignoring quoted text. |
84 | * For example, in the string "abc'hide'h", the 'h' in "hide" will not be |
85 | * found by a search for 'h'. |
86 | * @param text text to be searched |
87 | * @param start the beginning index, inclusive; <code>0 <= start |
88 | * <= limit</code>. |
89 | * @param limit the ending index, exclusive; <code>start <= limit |
90 | * <= text.length()</code>. |
91 | * @param c character to search for |
92 | * @return Offset of the first instance of c, or -1 if not found. |
93 | */ |
94 | //?FOR FUTURE USE. DISABLE FOR NOW for coverage reasons. |
95 | // static int32_t quotedIndexOf(const UnicodeString& text, |
96 | // int32_t start, int32_t limit, |
97 | // char16_t c); |
98 | |
99 | /** |
100 | * Skip over a sequence of zero or more white space characters at pos. |
101 | * @param advance if true, advance pos to the first non-white-space |
102 | * character at or after pos, or str.length(), if there is none. |
103 | * Otherwise leave pos unchanged. |
104 | * @return the index of the first non-white-space character at or |
105 | * after pos, or str.length(), if there is none. |
106 | */ |
107 | static int32_t skipWhitespace(const UnicodeString& str, int32_t& pos, |
108 | UBool advance = false); |
109 | |
110 | /** |
111 | * Skip over Pattern_White_Space in a Replaceable. |
112 | * Skipping may be done in the forward or |
113 | * reverse direction. In either case, the leftmost index will be |
114 | * inclusive, and the rightmost index will be exclusive. That is, |
115 | * given a range defined as [start, limit), the call |
116 | * skipWhitespace(text, start, limit) will advance start past leading |
117 | * whitespace, whereas the call skipWhitespace(text, limit, start), |
118 | * will back up limit past trailing whitespace. |
119 | * @param text the text to be analyzed |
120 | * @param pos either the start or limit of a range of 'text', to skip |
121 | * leading or trailing whitespace, respectively |
122 | * @param stop either the limit or start of a range of 'text', to skip |
123 | * leading or trailing whitespace, respectively |
124 | * @return the new start or limit, depending on what was passed in to |
125 | * 'pos' |
126 | */ |
127 | //?FOR FUTURE USE. DISABLE FOR NOW for coverage reasons. |
128 | //? static int32_t skipWhitespace(const Replaceable& text, |
129 | //? int32_t pos, int32_t stop); |
130 | |
131 | /** |
132 | * Parse a single non-whitespace character 'ch', optionally |
133 | * preceded by whitespace. |
134 | * @param id the string to be parsed |
135 | * @param pos INPUT-OUTPUT parameter. On input, pos[0] is the |
136 | * offset of the first character to be parsed. On output, pos[0] |
137 | * is the index after the last parsed character. If the parse |
138 | * fails, pos[0] will be unchanged. |
139 | * @param ch the non-whitespace character to be parsed. |
140 | * @return true if 'ch' is seen preceded by zero or more |
141 | * whitespace characters. |
142 | */ |
143 | static UBool parseChar(const UnicodeString& id, int32_t& pos, char16_t ch); |
144 | |
145 | /** |
146 | * Parse a pattern string starting at offset pos. Keywords are |
147 | * matched case-insensitively. Spaces may be skipped and may be |
148 | * optional or required. Integer values may be parsed, and if |
149 | * they are, they will be returned in the given array. If |
150 | * successful, the offset of the next non-space character is |
151 | * returned. On failure, -1 is returned. |
152 | * @param pattern must only contain lowercase characters, which |
153 | * will match their uppercase equivalents as well. A space |
154 | * character matches one or more required spaces. A '~' character |
155 | * matches zero or more optional spaces. A '#' character matches |
156 | * an integer and stores it in parsedInts, which the caller must |
157 | * ensure has enough capacity. |
158 | * @param parsedInts array to receive parsed integers. Caller |
159 | * must ensure that parsedInts.length is >= the number of '#' |
160 | * signs in 'pattern'. |
161 | * @return the position after the last character parsed, or -1 if |
162 | * the parse failed |
163 | */ |
164 | static int32_t parsePattern(const UnicodeString& rule, int32_t pos, int32_t limit, |
165 | const UnicodeString& pattern, int32_t* parsedInts); |
166 | |
167 | /** |
168 | * Parse a pattern string within the given Replaceable and a parsing |
169 | * pattern. Characters are matched literally and case-sensitively |
170 | * except for the following special characters: |
171 | * |
172 | * ~ zero or more Pattern_White_Space chars |
173 | * |
174 | * If end of pattern is reached with all matches along the way, |
175 | * pos is advanced to the first unparsed index and returned. |
176 | * Otherwise -1 is returned. |
177 | * @param pat pattern that controls parsing |
178 | * @param text text to be parsed, starting at index |
179 | * @param index offset to first character to parse |
180 | * @param limit offset after last character to parse |
181 | * @return index after last parsed character, or -1 on parse failure. |
182 | */ |
183 | static int32_t parsePattern(const UnicodeString& pat, |
184 | const Replaceable& text, |
185 | int32_t index, |
186 | int32_t limit); |
187 | |
188 | /** |
189 | * Parse an integer at pos, either of the form \d+ or of the form |
190 | * 0x[0-9A-Fa-f]+ or 0[0-7]+, that is, in standard decimal, hex, |
191 | * or octal format. |
192 | * @param pos INPUT-OUTPUT parameter. On input, the index of the first |
193 | * character to parse. On output, the index of the character after the |
194 | * last parsed character. |
195 | */ |
196 | static int32_t parseInteger(const UnicodeString& rule, int32_t& pos, int32_t limit); |
197 | |
198 | /** |
199 | * Parse an integer at pos using only ASCII digits. |
200 | * Base 10 only. |
201 | * @param pos INPUT-OUTPUT parameter. On input, the index of the first |
202 | * character to parse. On output, the index of the character after the |
203 | * last parsed character. |
204 | */ |
205 | static int32_t parseAsciiInteger(const UnicodeString& str, int32_t& pos); |
206 | |
207 | /** |
208 | * Parse a Unicode identifier from the given string at the given |
209 | * position. Return the identifier, or an empty string if there |
210 | * is no identifier. |
211 | * @param str the string to parse |
212 | * @param pos INPUT-OUTPUT parameter. On INPUT, pos is the |
213 | * first character to examine. It must be less than str.length(), |
214 | * and it must not point to a whitespace character. That is, must |
215 | * have pos < str.length() and |
216 | * !UCharacter::isWhitespace(str.char32At(pos)). On |
217 | * OUTPUT, the position after the last parsed character. |
218 | * @return the Unicode identifier, or an empty string if there is |
219 | * no valid identifier at pos. |
220 | */ |
221 | static UnicodeString parseUnicodeIdentifier(const UnicodeString& str, int32_t& pos); |
222 | |
223 | /** |
224 | * Parse an unsigned 31-bit integer at the given offset. Use |
225 | * UCharacter.digit() to parse individual characters into digits. |
226 | * @param text the text to be parsed |
227 | * @param pos INPUT-OUTPUT parameter. On entry, pos is the |
228 | * offset within text at which to start parsing; it should point |
229 | * to a valid digit. On exit, pos is the offset after the last |
230 | * parsed character. If the parse failed, it will be unchanged on |
231 | * exit. Must be >= 0 on entry. |
232 | * @param radix the radix in which to parse; must be >= 2 and <= |
233 | * 36. |
234 | * @return a non-negative parsed number, or -1 upon parse failure. |
235 | * Parse fails if there are no digits, that is, if pos does not |
236 | * point to a valid digit on entry, or if the number to be parsed |
237 | * does not fit into a 31-bit unsigned integer. |
238 | */ |
239 | static int32_t parseNumber(const UnicodeString& text, |
240 | int32_t& pos, int8_t radix); |
241 | |
242 | static void appendToRule(UnicodeString& rule, |
243 | UChar32 c, |
244 | UBool isLiteral, |
245 | UBool escapeUnprintable, |
246 | UnicodeString& quoteBuf); |
247 | |
248 | static void appendToRule(UnicodeString& rule, |
249 | const UnicodeString& text, |
250 | UBool isLiteral, |
251 | UBool escapeUnprintable, |
252 | UnicodeString& quoteBuf); |
253 | |
254 | static void appendToRule(UnicodeString& rule, |
255 | const UnicodeMatcher* matcher, |
256 | UBool escapeUnprintable, |
257 | UnicodeString& quoteBuf); |
258 | |
259 | private: |
260 | // do not instantiate |
261 | ICU_Utility() = delete; |
262 | }; |
263 | |
264 | U_NAMESPACE_END |
265 | |
266 | #endif |
267 | //eof |
268 | |