1.1 --- /dev/null Thu Jan 01 00:00:00 1970 +0000
1.2 +++ b/os/textandloc/fontservices/textshaperplugin/IcuSource/common/util.h Fri Jun 15 03:10:57 2012 +0200
1.3 @@ -0,0 +1,248 @@
1.4 +/*
1.5 +**********************************************************************
1.6 +* Copyright (c) 2001-2005, International Business Machines
1.7 +* Corporation and others. All Rights Reserved.
1.8 +**********************************************************************
1.9 +* Date Name Description
1.10 +* 11/19/2001 aliu Creation.
1.11 +**********************************************************************
1.12 +*/
1.13 +#ifndef ICU_UTIL_H
1.14 +#define ICU_UTIL_H
1.15 +
1.16 +#include "unicode/utypes.h"
1.17 +#include "unicode/uobject.h"
1.18 +#include "unicode/unistr.h"
1.19 +
1.20 +//--------------------------------------------------------------------
1.21 +// class ICU_Utility
1.22 +// i18n utility functions, scoped into the class ICU_Utility.
1.23 +//--------------------------------------------------------------------
1.24 +
1.25 +U_NAMESPACE_BEGIN
1.26 +
1.27 +class UnicodeMatcher;
1.28 +
1.29 +class U_COMMON_API ICU_Utility /* not : public UObject because all methods are static */ {
1.30 + public:
1.31 +
1.32 + /**
1.33 + * Append a number to the given UnicodeString in the given radix.
1.34 + * Standard digits '0'-'9' are used and letters 'A'-'Z' for
1.35 + * radices 11 through 36.
1.36 + * @param result the digits of the number are appended here
1.37 + * @param n the number to be converted to digits; may be negative.
1.38 + * If negative, a '-' is prepended to the digits.
1.39 + * @param radix a radix from 2 to 36 inclusive.
1.40 + * @param minDigits the minimum number of digits, not including
1.41 + * any '-', to produce. Values less than 2 have no effect. One
1.42 + * digit is always emitted regardless of this parameter.
1.43 + * @return a reference to result
1.44 + */
1.45 + static UnicodeString& appendNumber(UnicodeString& result, int32_t n,
1.46 + int32_t radix = 10,
1.47 + int32_t minDigits = 1);
1.48 +
1.49 + /**
1.50 + * Return true if the character is NOT printable ASCII.
1.51 + *
1.52 + * This method should really be in UnicodeString (or similar). For
1.53 + * now, we implement it here and share it with friend classes.
1.54 + */
1.55 + static UBool isUnprintable(UChar32 c);
1.56 +
1.57 + /**
1.58 + * Escape unprintable characters using \uxxxx notation for U+0000 to
1.59 + * U+FFFF and \Uxxxxxxxx for U+10000 and above. If the character is
1.60 + * printable ASCII, then do nothing and return FALSE. Otherwise,
1.61 + * append the escaped notation and return TRUE.
1.62 + */
1.63 + static UBool escapeUnprintable(UnicodeString& result, UChar32 c);
1.64 +
1.65 + /**
1.66 + * Returns the index of a character, ignoring quoted text.
1.67 + * For example, in the string "abc'hide'h", the 'h' in "hide" will not be
1.68 + * found by a search for 'h'.
1.69 + * @param text text to be searched
1.70 + * @param start the beginning index, inclusive; <code>0 <= start
1.71 + * <= limit</code>.
1.72 + * @param limit the ending index, exclusive; <code>start <= limit
1.73 + * <= text.length()</code>.
1.74 + * @param c character to search for
1.75 + * @return Offset of the first instance of c, or -1 if not found.
1.76 + */
1.77 +//?FOR FUTURE USE. DISABLE FOR NOW for coverage reasons.
1.78 +// static int32_t quotedIndexOf(const UnicodeString& text,
1.79 +// int32_t start, int32_t limit,
1.80 +// UChar c);
1.81 +
1.82 + /**
1.83 + * Skip over a sequence of zero or more white space characters at pos.
1.84 + * @param advance if true, advance pos to the first non-white-space
1.85 + * character at or after pos, or str.length(), if there is none.
1.86 + * Otherwise leave pos unchanged.
1.87 + * @return the index of the first non-white-space character at or
1.88 + * after pos, or str.length(), if there is none.
1.89 + */
1.90 + static int32_t skipWhitespace(const UnicodeString& str, int32_t& pos,
1.91 + UBool advance = FALSE);
1.92 +
1.93 + /**
1.94 + * Skip over whitespace in a Replaceable. Whitespace is defined by
1.95 + * uprv_isRuleWhiteSpace(). Skipping may be done in the forward or
1.96 + * reverse direction. In either case, the leftmost index will be
1.97 + * inclusive, and the rightmost index will be exclusive. That is,
1.98 + * given a range defined as [start, limit), the call
1.99 + * skipWhitespace(text, start, limit) will advance start past leading
1.100 + * whitespace, whereas the call skipWhitespace(text, limit, start),
1.101 + * will back up limit past trailing whitespace.
1.102 + * @param text the text to be analyzed
1.103 + * @param pos either the start or limit of a range of 'text', to skip
1.104 + * leading or trailing whitespace, respectively
1.105 + * @param stop either the limit or start of a range of 'text', to skip
1.106 + * leading or trailing whitespace, respectively
1.107 + * @return the new start or limit, depending on what was passed in to
1.108 + * 'pos'
1.109 + */
1.110 +//?FOR FUTURE USE. DISABLE FOR NOW for coverage reasons.
1.111 +//? static int32_t skipWhitespace(const Replaceable& text,
1.112 +//? int32_t pos, int32_t stop);
1.113 +
1.114 + /**
1.115 + * Parse a single non-whitespace character 'ch', optionally
1.116 + * preceded by whitespace.
1.117 + * @param id the string to be parsed
1.118 + * @param pos INPUT-OUTPUT parameter. On input, pos[0] is the
1.119 + * offset of the first character to be parsed. On output, pos[0]
1.120 + * is the index after the last parsed character. If the parse
1.121 + * fails, pos[0] will be unchanged.
1.122 + * @param ch the non-whitespace character to be parsed.
1.123 + * @return true if 'ch' is seen preceded by zero or more
1.124 + * whitespace characters.
1.125 + */
1.126 + static UBool parseChar(const UnicodeString& id, int32_t& pos, UChar ch);
1.127 +
1.128 + /**
1.129 + * Parse a pattern string starting at offset pos. Keywords are
1.130 + * matched case-insensitively. Spaces may be skipped and may be
1.131 + * optional or required. Integer values may be parsed, and if
1.132 + * they are, they will be returned in the given array. If
1.133 + * successful, the offset of the next non-space character is
1.134 + * returned. On failure, -1 is returned.
1.135 + * @param pattern must only contain lowercase characters, which
1.136 + * will match their uppercase equivalents as well. A space
1.137 + * character matches one or more required spaces. A '~' character
1.138 + * matches zero or more optional spaces. A '#' character matches
1.139 + * an integer and stores it in parsedInts, which the caller must
1.140 + * ensure has enough capacity.
1.141 + * @param parsedInts array to receive parsed integers. Caller
1.142 + * must ensure that parsedInts.length is >= the number of '#'
1.143 + * signs in 'pattern'.
1.144 + * @return the position after the last character parsed, or -1 if
1.145 + * the parse failed
1.146 + */
1.147 + static int32_t parsePattern(const UnicodeString& rule, int32_t pos, int32_t limit,
1.148 + const UnicodeString& pattern, int32_t* parsedInts);
1.149 +
1.150 + /**
1.151 + * Parse a pattern string within the given Replaceable and a parsing
1.152 + * pattern. Characters are matched literally and case-sensitively
1.153 + * except for the following special characters:
1.154 + *
1.155 + * ~ zero or more uprv_isRuleWhiteSpace chars
1.156 + *
1.157 + * If end of pattern is reached with all matches along the way,
1.158 + * pos is advanced to the first unparsed index and returned.
1.159 + * Otherwise -1 is returned.
1.160 + * @param pat pattern that controls parsing
1.161 + * @param text text to be parsed, starting at index
1.162 + * @param index offset to first character to parse
1.163 + * @param limit offset after last character to parse
1.164 + * @return index after last parsed character, or -1 on parse failure.
1.165 + */
1.166 + static int32_t parsePattern(const UnicodeString& pat,
1.167 + const Replaceable& text,
1.168 + int32_t index,
1.169 + int32_t limit);
1.170 +
1.171 + /**
1.172 + * Parse an integer at pos, either of the form \d+ or of the form
1.173 + * 0x[0-9A-Fa-f]+ or 0[0-7]+, that is, in standard decimal, hex,
1.174 + * or octal format.
1.175 + * @param pos INPUT-OUTPUT parameter. On input, the first
1.176 + * character to parse. On output, the character after the last
1.177 + * parsed character.
1.178 + */
1.179 + static int32_t parseInteger(const UnicodeString& rule, int32_t& pos, int32_t limit);
1.180 +
1.181 + /**
1.182 + * Parse a Unicode identifier from the given string at the given
1.183 + * position. Return the identifier, or an empty string if there
1.184 + * is no identifier.
1.185 + * @param str the string to parse
1.186 + * @param pos INPUT-OUPUT parameter. On INPUT, pos is the
1.187 + * first character to examine. It must be less than str.length(),
1.188 + * and it must not point to a whitespace character. That is, must
1.189 + * have pos < str.length() and
1.190 + * !UCharacter::isWhitespace(str.char32At(pos)). On
1.191 + * OUTPUT, the position after the last parsed character.
1.192 + * @return the Unicode identifier, or an empty string if there is
1.193 + * no valid identifier at pos.
1.194 + */
1.195 + static UnicodeString parseUnicodeIdentifier(const UnicodeString& str, int32_t& pos);
1.196 +
1.197 + /**
1.198 + * Parse an unsigned 31-bit integer at the given offset. Use
1.199 + * UCharacter.digit() to parse individual characters into digits.
1.200 + * @param text the text to be parsed
1.201 + * @param pos INPUT-OUTPUT parameter. On entry, pos is the
1.202 + * offset within text at which to start parsing; it should point
1.203 + * to a valid digit. On exit, pos is the offset after the last
1.204 + * parsed character. If the parse failed, it will be unchanged on
1.205 + * exit. Must be >= 0 on entry.
1.206 + * @param radix the radix in which to parse; must be >= 2 and <=
1.207 + * 36.
1.208 + * @return a non-negative parsed number, or -1 upon parse failure.
1.209 + * Parse fails if there are no digits, that is, if pos does not
1.210 + * point to a valid digit on entry, or if the number to be parsed
1.211 + * does not fit into a 31-bit unsigned integer.
1.212 + */
1.213 + static int32_t parseNumber(const UnicodeString& text,
1.214 + int32_t& pos, int8_t radix);
1.215 +
1.216 + static void appendToRule(UnicodeString& rule,
1.217 + UChar32 c,
1.218 + UBool isLiteral,
1.219 + UBool escapeUnprintable,
1.220 + UnicodeString& quoteBuf);
1.221 +
1.222 + static void appendToRule(UnicodeString& rule,
1.223 + const UnicodeString& text,
1.224 + UBool isLiteral,
1.225 + UBool escapeUnprintable,
1.226 + UnicodeString& quoteBuf);
1.227 +
1.228 + static void appendToRule(UnicodeString& rule,
1.229 + const UnicodeMatcher* matcher,
1.230 + UBool escapeUnprintable,
1.231 + UnicodeString& quoteBuf);
1.232 +
1.233 +private:
1.234 + // do not instantiate
1.235 + ICU_Utility();
1.236 +};
1.237 +
1.238 +U_NAMESPACE_END
1.239 +
1.240 +/**
1.241 + * Is this character a "white space" in the sense of ICU rule parsers?
1.242 + * Equivalent to test for Pattern_White_Space Unicode property.
1.243 + * Stable set of characters, won't change.
1.244 + * See UAX #31 Identifier and Pattern Syntax: http://www.unicode.org/reports/tr31/
1.245 + * @internal
1.246 + */
1.247 +U_CAPI UBool U_EXPORT2
1.248 +uprv_isRuleWhiteSpace(UChar32 c);
1.249 +
1.250 +#endif
1.251 +//eof