include/iprt/ctype.h

   1 /** @file
   2  * IPRT - Simple character type classiciation and conversion.
   3  */
   4
   5 /*
   6  * Copyright (C) 2006-2024 Oracle and/or its affiliates.
   7  *
   8  * This file is part of VirtualBox base platform packages, as
   9  * available from https://www.virtualbox.org.
  10  *
  11  * This program is free software; you can redistribute it and/or
  12  * modify it under the terms of the GNU General Public License
  13  * as published by the Free Software Foundation, in version 3 of the
  14  * License.
  15  *
  16  * This program is distributed in the hope that it will be useful, but
  17  * WITHOUT ANY WARRANTY; without even the implied warranty of
  18  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  19  * General Public License for more details.
  20  *
  21  * You should have received a copy of the GNU General Public License
  22  * along with this program; if not, see <https://www.gnu.org/licenses>.
  23  *
  24  * The contents of this file may alternatively be used under the terms
  25  * of the Common Development and Distribution License Version 1.0
  26  * (CDDL), a copy of it is provided in the "COPYING.CDDL" file included
  27  * in the VirtualBox distribution, in which case the provisions of the
  28  * CDDL are applicable instead of those of the GPL.
  29  *
  30  * You may elect to license modified versions of this file under the
  31  * terms and conditions of either the GPL or the CDDL or both.
  32  *
  33  * SPDX-License-Identifier: GPL-3.0-only OR CDDL-1.0
  34  */
  35
  36 #ifndef IPRT_INCLUDED_ctype_h
  37 #define IPRT_INCLUDED_ctype_h
  38 #ifndef RT_WITHOUT_PRAGMA_ONCE
  39 # pragma once
  40 #endif
  41
  42 #include <iprt/types.h>
  43
  44 /** @name C locale predicates and conversions.
  45  *
  46  * For most practical purposes, this can safely be used when parsing UTF-8
  47  * strings.  Just keep in mind that we only deal with the first 127 chars and
  48  * that full correctness is only archived using the non-existing RTLocIs* API.
  49  *
  50  * @remarks Use the marcros, not the inlined functions.
  51  *
  52  * @remarks ASSUMES the source code includes the basic ASCII chars. This is a
  53  *          general IPRT assumption.
  54  * @{ */
  55 #define RT_C_IS_BLANK(ch)   RTLocCIsBlank((ch))
  56 #define RT_C_IS_ALNUM(ch)   RTLocCIsAlNum((ch))
  57 #define RT_C_IS_ALPHA(ch)   RTLocCIsAlpha((ch))
  58 #define RT_C_IS_CNTRL(ch)   RTLocCIsCntrl((ch))
  59 #define RT_C_IS_DIGIT(ch)   RTLocCIsDigit((ch))
  60 #define RT_C_IS_LOWER(ch)   RTLocCIsLower((ch))
  61 #define RT_C_IS_GRAPH(ch)   RTLocCIsGraph((ch))
  62 #define RT_C_IS_ODIGIT(ch)  RTLocCIsODigit((ch))
  63 #define RT_C_IS_PRINT(ch)   RTLocCIsPrint((ch))
  64 #define RT_C_IS_PUNCT(ch)   RTLocCIsPunct((ch))
  65 #define RT_C_IS_SPACE(ch)   RTLocCIsSpace((ch))
  66 #define RT_C_IS_UPPER(ch)   RTLocCIsUpper((ch))
  67 #define RT_C_IS_XDIGIT(ch)  RTLocCIsXDigit((ch))
  68
  69 #define RT_C_TO_LOWER(ch)   RTLocCToLower((ch))
  70 #define RT_C_TO_UPPER(ch)   RTLocCToUpper((ch))
  71
  72 /**
  73  * Checks for a blank character.
  74  *
  75  * @returns true / false.
  76  * @param   ch      The character to test.
  77  */
  78 DECL_FORCE_INLINE(bool) RTLocCIsBlank(int ch)
  79 {
  80     return ch == 0x20  /* space */
  81         || ch == 0x09; /* horizontal tab */
  82 }
  83
  84 /**
  85  * Checks for a control character.
  86  *
  87  * @returns true / false.
  88  * @param   ch      The character to test.
  89  *
  90  * @note    Will return true of ch is '\0'!
  91  */
  92 DECL_FORCE_INLINE(bool) RTLocCIsCntrl(int ch)
  93 {
  94     return (unsigned)ch < 32U /* 0..2f */
  95         || ch == 0x7f;
  96 }
  97
  98 /**
  99  * Checks for a decimal digit.
 100  *
 101  * @returns true / false.
 102  * @param   ch      The character to test.
 103  */
 104 DECL_FORCE_INLINE(bool) RTLocCIsDigit(int ch)
 105 {
 106     return (unsigned)ch - 0x30 < 10U; /* 30..39 */
 107 }
 108
 109 /**
 110  * Checks for a lower case character.
 111  *
 112  * @returns true / false.
 113  * @param   ch      The character to test.
 114  */
 115 DECL_FORCE_INLINE(bool) RTLocCIsLower(int ch)
 116 {
 117     return (unsigned)ch - 0x61U < 26U; /* 61..7a */
 118 }
 119
 120 /**
 121  * Checks for an octal digit.
 122  *
 123  * @returns true / false.
 124  * @param   ch      The character to test.
 125  */
 126 DECL_FORCE_INLINE(bool) RTLocCIsODigit(int ch)
 127 {
 128     return (unsigned)ch - 0x30 < 8U; /* 30..37 */
 129 }
 130
 131 /**
 132  * Checks for a printable character (whitespace included).
 133  *
 134  * @returns true / false.
 135  * @param   ch      The character to test.
 136  */
 137 DECL_FORCE_INLINE(bool) RTLocCIsPrint(int ch)
 138 {
 139     return (unsigned)ch - 0x20U < 95U; /* 20..7e */
 140 }
 141
 142 /**
 143  * Checks for punctuation (?).
 144  *
 145  * @returns true / false.
 146  * @param   ch      The character to test.
 147  */
 148 DECL_FORCE_INLINE(bool) RTLocCIsPunct(int ch)
 149 {
 150     return (unsigned)ch - 0x21U < 15U /* 21..2f */
 151         || (unsigned)ch - 0x3aU <  7U /* 3a..40 */
 152         || (unsigned)ch - 0x5bU <  6U /* 5a..60 */
 153         || (unsigned)ch - 0x7bU <  4U /* 7b..7e */;
 154 }
 155
 156 /**
 157  * Checks for a white-space character.
 158  *
 159  * @returns true / false.
 160  * @param   ch      The character to test.
 161  */
 162 DECL_FORCE_INLINE(bool) RTLocCIsSpace(int ch)
 163 {
 164     return ch == 0x20                 /* 20 (space) */
 165         || (unsigned)ch - 0x09U < 5U; /* 09..0d */
 166 }
 167
 168 /**
 169  * Checks for an upper case character.
 170  *
 171  * @returns true / false.
 172  * @param   ch      The character to test.
 173  */
 174 DECL_FORCE_INLINE(bool) RTLocCIsUpper(int ch)
 175 {
 176     return (unsigned)ch - 0x41 < 26U; /* 41..5a */
 177 }
 178
 179 /**
 180  * Checks for a hexadecimal digit.
 181  *
 182  * @returns true / false.
 183  * @param   ch      The character to test.
 184  */
 185 DECL_FORCE_INLINE(bool) RTLocCIsXDigit(int ch)
 186 {
 187     return (unsigned)ch - 0x30 < 10U /* 30..39 (0-9) */
 188         || (unsigned)ch - 0x41 < 6   /* 41..46 (A-F) */
 189         || (unsigned)ch - 0x61 < 6;  /* 61..66 (a-f) */
 190 }
 191
 192 /**
 193  * Checks for an alphabetic character.
 194  *
 195  * @returns true / false.
 196  * @param   ch      The character to test.
 197  */
 198 DECL_FORCE_INLINE(bool) RTLocCIsAlpha(int ch)
 199 {
 200     return RTLocCIsLower(ch) || RTLocCIsUpper(ch);
 201 }
 202
 203 /**
 204  * Checks for an alphanumerical character.
 205  *
 206  * @returns true / false.
 207  * @param   ch      The character to test.
 208  */
 209 DECL_FORCE_INLINE(bool) RTLocCIsAlNum(int ch)
 210 {
 211     return RTLocCIsDigit(ch) || RTLocCIsAlpha(ch);
 212 }
 213
 214 /**
 215  * Checks for a printable character whitespace excluded.
 216  *
 217  * @returns true / false.
 218  * @param   ch      The character to test.
 219  */
 220 DECL_FORCE_INLINE(bool) RTLocCIsGraph(int ch)
 221 {
 222     return RTLocCIsPrint(ch) && !RTLocCIsBlank(ch);
 223 }
 224
 225
 226 /**
 227  * Converts the character to lower case if applicable.
 228  *
 229  * @returns lower cased character or ch.
 230  * @param   ch      The character to test.
 231  */
 232 DECL_FORCE_INLINE(int) RTLocCToLower(int ch)
 233 {
 234     return RTLocCIsUpper(ch) ? (ch) + 0x20 : (ch);
 235 }
 236
 237 /**
 238  * Converts the character to upper case if applicable.
 239  *
 240  * @returns upper cased character or ch.
 241  * @param   ch      The character to test.
 242  */
 243 DECL_FORCE_INLINE(int) RTLocCToUpper(int ch)
 244 {
 245     return RTLocCIsLower(ch) ? (ch) - 0x20 : (ch);
 246 }
 247
 248
 249 /** @} */
 250
 251 #endif /* !IPRT_INCLUDED_ctype_h */