icuSources/common/util.h

   1 /*
   2 **********************************************************************
   3 *   Copyright (c) 2001-2005, International Business Machines
   4 *   Corporation and others.  All Rights Reserved.
   5 **********************************************************************
   6 *   Date        Name        Description
   7 *   11/19/2001  aliu        Creation.
   8 **********************************************************************
   9 */
  10 #ifndef ICU_UTIL_H
  11 #define ICU_UTIL_H
  12
  13 #include "unicode/utypes.h"
  14 #include "unicode/uobject.h"
  15 #include "unicode/unistr.h"
  16
  17 //--------------------------------------------------------------------
  18 // class ICU_Utility
  19 // i18n utility functions, scoped into the class ICU_Utility.
  20 //--------------------------------------------------------------------
  21
  22 U_NAMESPACE_BEGIN
  23
  24 class UnicodeMatcher;
  25
  26 class U_COMMON_API ICU_Utility /* not : public UObject because all methods are static */ {
  27  public:
  28
  29     /**
  30      * Append a number to the given UnicodeString in the given radix.
  31      * Standard digits '0'-'9' are used and letters 'A'-'Z' for
  32      * radices 11 through 36.
  33      * @param result the digits of the number are appended here
  34      * @param n the number to be converted to digits; may be negative.
  35      * If negative, a '-' is prepended to the digits.
  36      * @param radix a radix from 2 to 36 inclusive.
  37      * @param minDigits the minimum number of digits, not including
  38      * any '-', to produce.  Values less than 2 have no effect.  One
  39      * digit is always emitted regardless of this parameter.
  40      * @return a reference to result
  41      */
  42     static UnicodeString& appendNumber(UnicodeString& result, int32_t n,
  43                                        int32_t radix = 10,
  44                                        int32_t minDigits = 1);
  45
  46     /**
  47      * Return true if the character is NOT printable ASCII.
  48      *
  49      * This method should really be in UnicodeString (or similar).  For
  50      * now, we implement it here and share it with friend classes.
  51      */
  52     static UBool isUnprintable(UChar32 c);
  53
  54     /**
  55      * Escape unprintable characters using \uxxxx notation for U+0000 to
  56      * U+FFFF and \Uxxxxxxxx for U+10000 and above.  If the character is
  57      * printable ASCII, then do nothing and return FALSE.  Otherwise,
  58      * append the escaped notation and return TRUE.
  59      */
  60     static UBool escapeUnprintable(UnicodeString& result, UChar32 c);
  61
  62     /**
  63      * Returns the index of a character, ignoring quoted text.
  64      * For example, in the string "abc'hide'h", the 'h' in "hide" will not be
  65      * found by a search for 'h'.
  66      * @param text text to be searched
  67      * @param start the beginning index, inclusive; <code>0 <= start
  68      * <= limit</code>.
  69      * @param limit the ending index, exclusive; <code>start <= limit
  70      * <= text.length()</code>.
  71      * @param c character to search for
  72      * @return Offset of the first instance of c, or -1 if not found.
  73      */
  74 //?FOR FUTURE USE.  DISABLE FOR NOW for coverage reasons.
  75 //    static int32_t quotedIndexOf(const UnicodeString& text,
  76 //                                 int32_t start, int32_t limit,
  77 //                                 UChar c);
  78
  79     /**
  80      * Skip over a sequence of zero or more white space characters at pos.
  81      * @param advance if true, advance pos to the first non-white-space
  82      * character at or after pos, or str.length(), if there is none.
  83      * Otherwise leave pos unchanged.
  84      * @return the index of the first non-white-space character at or
  85      * after pos, or str.length(), if there is none.
  86      */
  87     static int32_t skipWhitespace(const UnicodeString& str, int32_t& pos,
  88                                   UBool advance = FALSE);
  89
  90     /**
  91      * Skip over whitespace in a Replaceable.  Whitespace is defined by
  92      * uprv_isRuleWhiteSpace().  Skipping may be done in the forward or
  93      * reverse direction.  In either case, the leftmost index will be
  94      * inclusive, and the rightmost index will be exclusive.  That is,
  95      * given a range defined as [start, limit), the call
  96      * skipWhitespace(text, start, limit) will advance start past leading
  97      * whitespace, whereas the call skipWhitespace(text, limit, start),
  98      * will back up limit past trailing whitespace.
  99      * @param text the text to be analyzed
 100      * @param pos either the start or limit of a range of 'text', to skip
 101      * leading or trailing whitespace, respectively
 102      * @param stop either the limit or start of a range of 'text', to skip
 103      * leading or trailing whitespace, respectively
 104      * @return the new start or limit, depending on what was passed in to
 105      * 'pos'
 106      */
 107 //?FOR FUTURE USE.  DISABLE FOR NOW for coverage reasons.
 108 //?    static int32_t skipWhitespace(const Replaceable& text,
 109 //?                                  int32_t pos, int32_t stop);
 110
 111     /**
 112      * Parse a single non-whitespace character 'ch', optionally
 113      * preceded by whitespace.
 114      * @param id the string to be parsed
 115      * @param pos INPUT-OUTPUT parameter.  On input, pos[0] is the
 116      * offset of the first character to be parsed.  On output, pos[0]
 117      * is the index after the last parsed character.  If the parse
 118      * fails, pos[0] will be unchanged.
 119      * @param ch the non-whitespace character to be parsed.
 120      * @return true if 'ch' is seen preceded by zero or more
 121      * whitespace characters.
 122      */
 123     static UBool parseChar(const UnicodeString& id, int32_t& pos, UChar ch);
 124
 125     /**
 126      * Parse a pattern string starting at offset pos.  Keywords are
 127      * matched case-insensitively.  Spaces may be skipped and may be
 128      * optional or required.  Integer values may be parsed, and if
 129      * they are, they will be returned in the given array.  If
 130      * successful, the offset of the next non-space character is
 131      * returned.  On failure, -1 is returned.
 132      * @param pattern must only contain lowercase characters, which
 133      * will match their uppercase equivalents as well.  A space
 134      * character matches one or more required spaces.  A '~' character
 135      * matches zero or more optional spaces.  A '#' character matches
 136      * an integer and stores it in parsedInts, which the caller must
 137      * ensure has enough capacity.
 138      * @param parsedInts array to receive parsed integers.  Caller
 139      * must ensure that parsedInts.length is >= the number of '#'
 140      * signs in 'pattern'.
 141      * @return the position after the last character parsed, or -1 if
 142      * the parse failed
 143      */
 144     static int32_t parsePattern(const UnicodeString& rule, int32_t pos, int32_t limit,
 145                                 const UnicodeString& pattern, int32_t* parsedInts);
 146
 147     /**
 148      * Parse a pattern string within the given Replaceable and a parsing
 149      * pattern.  Characters are matched literally and case-sensitively
 150      * except for the following special characters:
 151      *
 152      * ~  zero or more uprv_isRuleWhiteSpace chars
 153      *
 154      * If end of pattern is reached with all matches along the way,
 155      * pos is advanced to the first unparsed index and returned.
 156      * Otherwise -1 is returned.
 157      * @param pat pattern that controls parsing
 158      * @param text text to be parsed, starting at index
 159      * @param index offset to first character to parse
 160      * @param limit offset after last character to parse
 161      * @return index after last parsed character, or -1 on parse failure.
 162      */
 163     static int32_t parsePattern(const UnicodeString& pat,
 164                                 const Replaceable& text,
 165                                 int32_t index,
 166                                 int32_t limit);
 167
 168     /**
 169      * Parse an integer at pos, either of the form \d+ or of the form
 170      * 0x[0-9A-Fa-f]+ or 0[0-7]+, that is, in standard decimal, hex,
 171      * or octal format.
 172      * @param pos INPUT-OUTPUT parameter.  On input, the first
 173      * character to parse.  On output, the character after the last
 174      * parsed character.
 175      */
 176     static int32_t parseInteger(const UnicodeString& rule, int32_t& pos, int32_t limit);
 177
 178     /**
 179      * Parse a Unicode identifier from the given string at the given
 180      * position.  Return the identifier, or an empty string if there
 181      * is no identifier.
 182      * @param str the string to parse
 183      * @param pos INPUT-OUPUT parameter.  On INPUT, pos is the
 184      * first character to examine.  It must be less than str.length(),
 185      * and it must not point to a whitespace character.  That is, must
 186      * have pos < str.length() and
 187      * !UCharacter::isWhitespace(str.char32At(pos)).  On
 188      * OUTPUT, the position after the last parsed character.
 189      * @return the Unicode identifier, or an empty string if there is
 190      * no valid identifier at pos.
 191      */
 192     static UnicodeString parseUnicodeIdentifier(const UnicodeString& str, int32_t& pos);
 193
 194     /**
 195      * Parse an unsigned 31-bit integer at the given offset.  Use
 196      * UCharacter.digit() to parse individual characters into digits.
 197      * @param text the text to be parsed
 198      * @param pos INPUT-OUTPUT parameter.  On entry, pos is the
 199      * offset within text at which to start parsing; it should point
 200      * to a valid digit.  On exit, pos is the offset after the last
 201      * parsed character.  If the parse failed, it will be unchanged on
 202      * exit.  Must be >= 0 on entry.
 203      * @param radix the radix in which to parse; must be >= 2 and <=
 204      * 36.
 205      * @return a non-negative parsed number, or -1 upon parse failure.
 206      * Parse fails if there are no digits, that is, if pos does not
 207      * point to a valid digit on entry, or if the number to be parsed
 208      * does not fit into a 31-bit unsigned integer.
 209      */
 210     static int32_t parseNumber(const UnicodeString& text,
 211                                int32_t& pos, int8_t radix);
 212
 213     static void appendToRule(UnicodeString& rule,
 214                              UChar32 c,
 215                              UBool isLiteral,
 216                              UBool escapeUnprintable,
 217                              UnicodeString& quoteBuf);
 218
 219     static void appendToRule(UnicodeString& rule,
 220                              const UnicodeString& text,
 221                              UBool isLiteral,
 222                              UBool escapeUnprintable,
 223                              UnicodeString& quoteBuf);
 224
 225     static void appendToRule(UnicodeString& rule,
 226                              const UnicodeMatcher* matcher,
 227                              UBool escapeUnprintable,
 228                              UnicodeString& quoteBuf);
 229
 230 private:
 231     // do not instantiate
 232     ICU_Utility();
 233 };
 234
 235 U_NAMESPACE_END
 236
 237 /**
 238  * Is this character a "white space" in the sense of ICU rule parsers?
 239  * Equivalent to test for Pattern_White_Space Unicode property.
 240  * Stable set of characters, won't change.
 241  * See UAX #31 Identifier and Pattern Syntax: http://www.unicode.org/reports/tr31/
 242  * @internal
 243  */
 244 U_CAPI UBool U_EXPORT2
 245 uprv_isRuleWhiteSpace(UChar32 c);
 246
 247 #endif
 248 //eof