icuSources/common/util.h

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