C++ UnicodeString::charAt方法代码示例

U_NAMESPACE_BEGIN

/**
 * Parse an integer at pos, either of the form \d+ or of the form
 * 0x[0-9A-Fa-f]+ or 0[0-7]+, that is, in standard decimal, hex,
 * or octal format.
 * @param pos INPUT-OUTPUT parameter.  On input, the first
 * character to parse.  On output, the character after the last
 * parsed character.
 */
int32_t ICU_Utility::parseInteger(const UnicodeString& rule, int32_t& pos, int32_t limit) {
    int32_t count = 0;
    int32_t value = 0;
    int32_t p = pos;
    int8_t radix = 10;

    if (p < limit && rule.charAt(p) == 48 /*0*/) {
        if (p+1 < limit && (rule.charAt(p+1) == 0x78 /*x*/ || rule.charAt(p+1) == 0x58 /*X*/)) {
            p += 2;
            radix = 16;
        }
        else {
            p++;
            count = 1;
            radix = 8;
        }
    }

    while (p < limit) {
        int32_t d = u_digit(rule.charAt(p++), radix);
        if (d < 0) {
            --p;
            break;
        }
        ++count;
        int32_t v = (value * radix) + d;
        if (v <= value) {
            // If there are too many input digits, at some point
            // the value will go negative, e.g., if we have seen
            // "0x8000000" already and there is another '0', when
            // we parse the next 0 the value will go negative.
            return 0;
        }
        value = v;
    }
    if (count > 0) {
        pos = p;
    }
    return value;
}

//
//   replaceCharRefs
//
//      replace the char entities <  & { ካ etc. in a string
//       with the corresponding actual character.
//
void
UXMLParser::replaceCharRefs(UnicodeString &s, UErrorCode &status) {
    UnicodeString result;
    UnicodeString replacement;
    int     i;

    mAmps.reset(s);
    // See the initialization for the regex matcher mAmps.
    //    Which entity we've matched is determined by which capture group has content,
    //      which is flaged by start() of that group not being -1.
    while (mAmps.find()) {
        if (mAmps.start(1, status) != -1) {
            replacement.setTo((UChar)x_AMP);
        } else if (mAmps.start(2, status) != -1) {
            replacement.setTo((UChar)x_LT);
        } else if (mAmps.start(3, status) != -1) {
            replacement.setTo((UChar)x_GT);
        } else if (mAmps.start(4, status) != -1) {
            replacement.setTo((UChar)x_APOS);
        } else if (mAmps.start(5, status) != -1) {
            replacement.setTo((UChar)x_QUOT);
        } else if (mAmps.start(6, status) != -1) {
            UnicodeString hexString = mAmps.group(6, status);
            UChar32 val = 0;
            for (i=0; i<hexString.length(); i++) {
                val = (val << 4) + u_digit(hexString.charAt(i), 16);
            }
            // TODO:  some verification that the character is valid
            replacement.setTo(val);
        } else if (mAmps.start(7, status) != -1) {
            UnicodeString decimalString = mAmps.group(7, status);
            UChar32 val = 0;
            for (i=0; i<decimalString.length(); i++) {
                val = val*10 + u_digit(decimalString.charAt(i), 10);
            }
            // TODO:  some verification that the character is valid
            replacement.setTo(val);
        } else {
            // An unrecognized &entity;  Leave it alone.
            //  TODO:  check that it really looks like an entity, and is not some
            //         random & in the text.
            replacement = mAmps.group((int32_t)0, status);
        }
        mAmps.appendReplacement(result, replacement, status);
    }
    mAmps.appendTail(result);
    s = result;
}

static void setCharField(JNIEnv* env, jobject obj, const char* fieldName, const UnicodeString& value) {
    if (value.length() == 0) {
        return;
    }
    jfieldID fid = env->GetFieldID(JniConstants::localeDataClass, fieldName, "C");
    env->SetCharField(obj, fid, value.charAt(0));
}

UBool
LocaleUtility::isFallbackOf(const UnicodeString& root, const UnicodeString& child)
{
    return child.indexOf(root) == 0 &&
      (child.length() == root.length() ||
       child.charAt(root.length()) == UNDERSCORE_CHAR);
}

static UnicodeString parseHex(const UnicodeString &in) {
    // Convert a series of hex numbers in a Unicode String to a string with the
    // corresponding characters.
    // The conversion is _really_ annoying.  There must be some function to just do it.
    UnicodeString result;
    UChar32 cc = 0;
    for (int32_t i=0; i<in.length(); i++) {
        UChar c = in.charAt(i);
        if (c == 0x20) {   // Space
            if (cc > 0) {
               result.append(cc);
               cc = 0;
            }
        } else if (c>=0x30 && c<=0x39) {
            cc = (cc<<4) + (c - 0x30);
        } else if ((c>=0x41 && c<=0x46) || (c>=0x61 && c<=0x66)) {
            cc = (cc<<4) + (c & 0x0f)+9;
        }
        // else do something with bad input.
    }
    if (cc > 0) {
        result.append(cc);
    }
    return result;
}

// fixQuotes unescapes single quotes. Don''t -> Don't. Letter 'j' -> Letter j.
// Modifies s in place.
static void fixQuotes(UnicodeString& s) {
  QuoteState state = OUTSIDE;
  int32_t len = s.length();
  int32_t dest = 0;
  for (int32_t i = 0; i < len; ++i) {
    UChar ch = s.charAt(i);
    if (ch == u_apos) {
      if (state == INSIDE_EMPTY) {
        s.setCharAt(dest, ch);
        ++dest;
      }
    } else {
      s.setCharAt(dest, ch);
      ++dest;
    }

    // Update state
    switch (state) {
      case OUTSIDE:
        state = ch == u_apos ? INSIDE_EMPTY : OUTSIDE;
        break;
      case INSIDE_EMPTY:
      case INSIDE_FULL:
        state = ch == u_apos ? OUTSIDE : INSIDE_FULL;
        break;
      default:
        break;
    }
  }
  s.truncate(dest);
}

void TextGroup::addRun(const UnicodeString &input, UBiDiDirection direction, int32_t start, int32_t end)
{
    std::string text;
    input.tempSubString(start, end - start).toUTF8String(text);
    printf("Hominlinx-->======TextGroup::addRun[%s]==== %d\n",text.c_str(), input.charAt(0) );

    runs_.emplace_back(text, script_, lang_, uciDirectionToHB(direction));
}

 TestReplaceable (const UnicodeString& text, 
                  const UnicodeString& newStyles) {
     chars = text;
     UnicodeString s;
     for (int i = 0; i < text.length(); ++i) {
         if (i < newStyles.length()) {
             s.append(newStyles.charAt(i));
         } else {
             if (text.charAt(i) == NO_STYLE_MARK) {
                 s.append(NO_STYLE);
             } else {
                 s.append((UChar)(i + 0x0031));
             }
         }
     }
     this->styles = s;
 }

/**
 * Parse a pattern string starting at offset pos.  Keywords are
 * matched case-insensitively.  Spaces may be skipped and may be
 * optional or required.  Integer values may be parsed, and if
 * they are, they will be returned in the given array.  If
 * successful, the offset of the next non-space character is
 * returned.  On failure, -1 is returned.
 * @param pattern must only contain lowercase characters, which
 * will match their uppercase equivalents as well.  A space
 * character matches one or more required spaces.  A '~' character
 * matches zero or more optional spaces.  A '#' character matches
 * an integer and stores it in parsedInts, which the caller must
 * ensure has enough capacity.
 * @param parsedInts array to receive parsed integers.  Caller
 * must ensure that parsedInts.length is >= the number of '#'
 * signs in 'pattern'.
 * @return the position after the last character parsed, or -1 if
 * the parse failed
 */
int32_t ICU_Utility::parsePattern(const UnicodeString& rule, int32_t pos, int32_t limit,
                              const UnicodeString& pattern, int32_t* parsedInts) {
    // TODO Update this to handle surrogates
    int32_t p;
    int32_t intCount = 0; // number of integers parsed
    for (int32_t i=0; i<pattern.length(); ++i) {
        UChar cpat = pattern.charAt(i);
        UChar c;
        switch (cpat) {
        case 32 /*' '*/:
            if (pos >= limit) {
                return -1;
            }
            c = rule.charAt(pos++);
            if (!PatternProps::isWhiteSpace(c)) {
                return -1;
            }
            // FALL THROUGH to skipWhitespace
            U_FALLTHROUGH;
        case 126 /*'~'*/:
            pos = skipWhitespace(rule, pos);
            break;
        case 35 /*'#'*/:
            p = pos;
            parsedInts[intCount++] = parseInteger(rule, p, limit);
            if (p == pos) {
                // Syntax error; failed to parse integer
                return -1;
            }
            pos = p;
            break;
        default:
            if (pos >= limit) {
                return -1;
            }
            c = (UChar) u_tolower(rule.charAt(pos++));
            if (c != cpat) {
                return -1;
            }
            break;
        }
    }
    return pos;
}

UBool
SelectFormat::checkValidKeyword(const UnicodeString& argKeyword ) const{
    int32_t len = argKeyword.length();
    if (len < 1){
        return FALSE;
    }
    CharacterClass type = classifyCharacter(argKeyword.charAt(0));
    if( type != tStartKeyword ){
        return FALSE;
    }

    for (int32_t i = 0; i < argKeyword.length(); ++i) {
        type = classifyCharacter(argKeyword.charAt(i));
        if( type != tStartKeyword && type != tContinueKeyword ){
            return FALSE;
        }
    }
    return TRUE;
}

int32_t
MessagePattern::parseArgNumber(const UnicodeString &s, int32_t start, int32_t limit) {
    // If the identifier contains only ASCII digits, then it is an argument _number_
    // and must not have leading zeros (except "0" itself).
    // Otherwise it is an argument _name_.
    if(start>=limit) {
        return UMSGPAT_ARG_NAME_NOT_VALID;
    }
    int32_t number;
    // Defer numeric errors until we know there are only digits.
    UBool badNumber;
    UChar c=s.charAt(start++);
    if(c==0x30) {
        if(start==limit) {
            return 0;
        } else {
            number=0;
            badNumber=TRUE;  // leading zero
        }
    } else if(0x31<=c && c<=0x39) {
        number=c-0x30;
        badNumber=FALSE;
    } else {
        return UMSGPAT_ARG_NAME_NOT_NUMBER;
    }
    while(start<limit) {
        c=s.charAt(start++);
        if(0x30<=c && c<=0x39) {
            if(number>=INT32_MAX/10) {
                badNumber=TRUE;  // overflow
            }
            number=number*10+(c-0x30);
        } else {
            return UMSGPAT_ARG_NAME_NOT_NUMBER;
        }
    }
    // There are only ASCII digits.
    if(badNumber) {
        return UMSGPAT_ARG_NAME_NOT_VALID;
    } else {
        return number;
    }
}

U_NAMESPACE_BEGIN

int32_t ICU_Utility::parseInteger(const UnicodeString& rule, int32_t& pos, int32_t limit) {
    int32_t count = 0;
    int32_t value = 0;
    int32_t p = pos;
    int8_t radix = 10;

    if (p < limit && rule.charAt(p) == 48 /*0*/) {
        if (p+1 < limit && (rule.charAt(p+1) == 0x78 /*x*/ || rule.charAt(p+1) == 0x58 /*X*/)) {
            p += 2;
            radix = 16;
        }
        else {
            p++;
            count = 1;
            radix = 8;
        }
    }

    while (p < limit) {
        int32_t d = u_digit(rule.charAt(p++), radix);
        if (d < 0) {
            --p;
            break;
        }
        ++count;
        int32_t v = (value * radix) + d;
        if (v <= value) {
            // If there are too many input digits, at some point
            // the value will go negative, e.g., if we have seen
            // "0x8000000" already and there is another '0', when
            // we parse the next 0 the value will go negative.
            return 0;
        }
        value = v;
    }
    if (count > 0) {
        pos = p;
    }
    return value;
}

UBool checkEqual(const PluralRules &test, char *result, int32_t max) {
    UnicodeString key;
    UBool isEqual = TRUE;
    for (int32_t i=0; i<max; ++i) {
        key= test.select(i);
        if ( key.charAt(0)!=result[i] ) {
            isEqual = FALSE;
        }
    }
    return isEqual;
}

U_CFUNC void RBBI_DEBUG_printUnicodeString(const UnicodeString &s, int minWidth)
{
    int i;
    for (i=0; i<s.length(); i++) {
        RBBIDebugPrintf("%c", s.charAt(i));
        // putc(s.charAt(i), stdout);
    }
    for (i=s.length(); i<minWidth; i++) {
        RBBIDebugPrintf(" ");
    }
}

void UObjectTest::TestMFCCompatibility() {
#if U_HAVE_DEBUG_LOCATION_NEW
    /* Make sure that it compiles with MFC's debuggable new usage. */
    UnicodeString *str = new(__FILE__, __LINE__) UnicodeString();
    str->append((UChar)0x0040); // Is it usable?
    if(str->charAt(0) != 0x0040) {
        errln("debug new doesn't work.");
    }
    UnicodeString::operator delete(str, __FILE__, __LINE__);
#endif
}