intl/icu/source/common/patternprops.h

changeset 0
6474c204b198
     1.1 --- /dev/null	Thu Jan 01 00:00:00 1970 +0000
     1.2 +++ b/intl/icu/source/common/patternprops.h	Wed Dec 31 06:09:35 2014 +0100
     1.3 @@ -0,0 +1,89 @@
     1.4 +/*
     1.5 +*******************************************************************************
     1.6 +*   Copyright (C) 2011, International Business Machines
     1.7 +*   Corporation and others.  All Rights Reserved.
     1.8 +*******************************************************************************
     1.9 +*   file name:  patternprops.h
    1.10 +*   encoding:   US-ASCII
    1.11 +*   tab size:   8 (not used)
    1.12 +*   indentation:4
    1.13 +*
    1.14 +*   created on: 2011mar13
    1.15 +*   created by: Markus W. Scherer
    1.16 +*/
    1.17 +
    1.18 +#ifndef __PATTERNPROPS_H__
    1.19 +#define __PATTERNPROPS_H__
    1.20 +
    1.21 +#include "unicode/utypes.h"
    1.22 +
    1.23 +U_NAMESPACE_BEGIN
    1.24 +
    1.25 +/**
    1.26 + * Implements the immutable Unicode properties Pattern_Syntax and Pattern_White_Space.
    1.27 + * Hardcodes these properties, does not load data, does not depend on other ICU classes.
    1.28 + * <p>
    1.29 + * Note: Both properties include ASCII as well as non-ASCII, non-Latin-1 code points,
    1.30 + * and both properties only include BMP code points (no supplementary ones).
    1.31 + * Pattern_Syntax includes some unassigned code points.
    1.32 + * <p>
    1.33 + * [:Pattern_White_Space:] =
    1.34 + *   [\u0009-\u000D\ \u0085\u200E\u200F\u2028\u2029]
    1.35 + * <p>
    1.36 + * [:Pattern_Syntax:] =
    1.37 + *   [!-/\:-@\[-\^`\{-~\u00A1-\u00A7\u00A9\u00AB\u00AC\u00AE
    1.38 + *    \u00B0\u00B1\u00B6\u00BB\u00BF\u00D7\u00F7
    1.39 + *    \u2010-\u2027\u2030-\u203E\u2041-\u2053\u2055-\u205E
    1.40 + *    \u2190-\u245F\u2500-\u2775\u2794-\u2BFF\u2E00-\u2E7F
    1.41 + *    \u3001-\u3003\u3008-\u3020\u3030\uFD3E\uFD3F\uFE45\uFE46]
    1.42 + * @author mscherer
    1.43 + */
    1.44 +class U_COMMON_API PatternProps {
    1.45 +public:
    1.46 +    /**
    1.47 +     * @return TRUE if c is a Pattern_Syntax code point.
    1.48 +     */
    1.49 +    static UBool isSyntax(UChar32 c);
    1.50 +
    1.51 +    /**
    1.52 +     * @return TRUE if c is a Pattern_Syntax or Pattern_White_Space code point.
    1.53 +     */
    1.54 +    static UBool isSyntaxOrWhiteSpace(UChar32 c);
    1.55 +
    1.56 +    /**
    1.57 +     * @return TRUE if c is a Pattern_White_Space character.
    1.58 +     */
    1.59 +    static UBool isWhiteSpace(UChar32 c);
    1.60 +
    1.61 +    /**
    1.62 +     * Skips over Pattern_White_Space starting at s.
    1.63 +     * @return The smallest pointer at or after s with a non-white space character.
    1.64 +     */
    1.65 +    static const UChar *skipWhiteSpace(const UChar *s, int32_t length);
    1.66 +
    1.67 +    /**
    1.68 +     * @return s except with leading and trailing Pattern_White_Space removed and length adjusted.
    1.69 +     */
    1.70 +    static const UChar *trimWhiteSpace(const UChar *s, int32_t &length);
    1.71 +
    1.72 +    /**
    1.73 +     * Tests whether the string contains a "pattern identifier", that is,
    1.74 +     * whether it contains only non-Pattern_White_Space, non-Pattern_Syntax characters.
    1.75 +     * @return TRUE if there are no Pattern_White_Space or Pattern_Syntax characters in s.
    1.76 +     */
    1.77 +    static UBool isIdentifier(const UChar *s, int32_t length);
    1.78 +
    1.79 +    /**
    1.80 +     * Skips over a "pattern identifier" starting at index s.
    1.81 +     * @return The smallest pointer at or after s with
    1.82 +     *         a Pattern_White_Space or Pattern_Syntax character.
    1.83 +     */
    1.84 +    static const UChar *skipIdentifier(const UChar *s, int32_t length);
    1.85 +
    1.86 +private:
    1.87 +    PatternProps();  // no constructor: all static methods
    1.88 +};
    1.89 +
    1.90 +U_NAMESPACE_END
    1.91 +
    1.92 +#endif  // __PATTERNPROPS_H__

mercurial