ucx/string.h

Mon, 14 Jul 2014 13:51:02 +0200

author
Olaf Wintermann <olaf.wintermann@gmail.com>
date
Mon, 14 Jul 2014 13:51:02 +0200
changeset 180
2185f19dcc45
parent 179
ee25d79a4187
child 183
6a694f8f0084
permissions
-rw-r--r--

added new sstrcat

olaf@20 1 /*
universe@103 2 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS HEADER.
olaf@20 3 *
universe@177 4 * Copyright 2014 Olaf Wintermann. All rights reserved.
universe@103 5 *
universe@103 6 * Redistribution and use in source and binary forms, with or without
universe@103 7 * modification, are permitted provided that the following conditions are met:
universe@103 8 *
universe@103 9 * 1. Redistributions of source code must retain the above copyright
universe@103 10 * notice, this list of conditions and the following disclaimer.
universe@103 11 *
universe@103 12 * 2. Redistributions in binary form must reproduce the above copyright
universe@103 13 * notice, this list of conditions and the following disclaimer in the
universe@103 14 * documentation and/or other materials provided with the distribution.
universe@103 15 *
universe@103 16 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
universe@103 17 * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
universe@103 18 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
universe@103 19 * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE
universe@103 20 * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
universe@103 21 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
universe@103 22 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
universe@103 23 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
universe@103 24 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
universe@103 25 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
universe@103 26 * POSSIBILITY OF SUCH DAMAGE.
olaf@20 27 */
universe@116 28 /**
universe@116 29 * Bounded string implementation.
universe@116 30 *
universe@116 31 * The UCX strings (<code>sstr_t</code>) provide an alternative to C strings.
universe@116 32 * The main difference to C strings is, that <code>sstr_t</code> does <b>not
universe@116 33 * need to be <code>NULL</code>-terminated</b>. Instead the length is stored
universe@116 34 * within the structure.
universe@116 35 *
universe@116 36 * When using <code>sstr_t</code>, developers must be full aware of what type
universe@116 37 * of string (<code>NULL</code>-terminated) or not) they are using, when
universe@116 38 * accessing the <code>char* ptr</code> directly.
universe@116 39 *
universe@116 40 * The UCX string module provides some common string functions, known from
universe@116 41 * standard libc, working with <code>sstr_t</code>.
universe@116 42 *
universe@116 43 * @file string.h
universe@116 44 * @author Mike Becker
universe@116 45 * @author Olaf Wintermann
universe@116 46 */
olaf@20 47
universe@116 48 #ifndef UCX_STRING_H
universe@116 49 #define UCX_STRING_H
olaf@20 50
universe@69 51 #include "ucx.h"
olaf@109 52 #include "allocator.h"
universe@38 53 #include <stddef.h>
universe@38 54
universe@116 55 /** Shortcut for a <code>sstr_t struct</code> literal. */
universe@116 56 #define ST(s) { (char*)s, sizeof(s)-1 }
universe@146 57
universe@116 58 /** Shortcut for the conversion of a C string to a <code>sstr_t</code>. */
universe@116 59 #define S(s) sstrn((char*)s, sizeof(s)-1)
olaf@20 60
olaf@20 61 #ifdef __cplusplus
olaf@20 62 extern "C" {
olaf@20 63 #endif
olaf@20 64
universe@116 65 /**
universe@116 66 * The UCX string structure.
universe@116 67 */
universe@116 68 typedef struct {
universe@116 69 /** A reference to the string (<b>not necessarily <code>NULL</code>
universe@116 70 * -terminated</b>) */
olaf@20 71 char *ptr;
universe@116 72 /** The length of the string */
olaf@20 73 size_t length;
olaf@20 74 } sstr_t;
olaf@20 75
universe@116 76 /**
universe@116 77 * Creates a new sstr_t based on a C string.
universe@116 78 *
universe@116 79 * The length is implicitly inferred by using a call to <code>strlen()</code>.
olaf@20 80 *
universe@116 81 * <b>Note:</b> the sstr_t will hold a <i>reference</i> to the C string. If you
universe@116 82 * do want a copy, use sstrdup() on the return value of this function.
universe@116 83 *
universe@116 84 * @param cstring the C string to wrap
universe@116 85 * @return a new sstr_t containing the C string
universe@116 86 *
universe@116 87 * @see sstrn()
olaf@20 88 */
universe@116 89 sstr_t sstr(char *cstring);
olaf@20 90
universe@116 91 /**
universe@116 92 * Creates a new sstr_t of the specified length based on a C string.
olaf@20 93 *
universe@116 94 * <b>Note:</b> the sstr_t will hold a <i>reference</i> to the C string. If you
universe@116 95 * do want a copy, use sstrdup() on the return value of this function.
universe@116 96 *
universe@116 97 * @param cstring the C string to wrap
universe@116 98 * @param length the length of the string
universe@116 99 * @return a new sstr_t containing the C string
universe@116 100 *
universe@116 101 * @see sstr()
universe@116 102 * @see S()
olaf@20 103 */
universe@116 104 sstr_t sstrn(char *cstring, size_t length);
olaf@20 105
olaf@20 106
universe@116 107 /**
universe@116 108 * Returns the cumulated length of all specified strings.
olaf@20 109 *
universe@116 110 * At least one string must be specified.
universe@116 111 *
universe@116 112 * <b>Attention:</b> if the count argument does not match the count of the
universe@116 113 * specified strings, the behavior is undefined.
universe@116 114 *
universe@116 115 * @param count the total number of specified strings (so at least 1)
universe@116 116 * @param string the first string
universe@116 117 * @param ... all other strings
universe@116 118 * @return the cumulated length of all strings
olaf@20 119 */
universe@116 120 size_t sstrnlen(size_t count, sstr_t string, ...);
olaf@20 121
universe@119 122 /**
olaf@180 123 * Concatenates strings.
olaf@180 124 *
olaf@180 125 * @param count the total number of strings to concatenate
olaf@180 126 * @param ... all strings
olaf@180 127 * @return the concatenated string
olaf@180 128 */
olaf@180 129 sstr_t sstrcat(size_t count, sstr_t s1, sstr_t s2, ...);
olaf@180 130 sstr_t sstrcat_a(UcxAllocator *a, size_t count, sstr_t s1, sstr_t s2, ...);
olaf@180 131
olaf@180 132
olaf@180 133 /**
universe@119 134 * Returns a substring starting at the specified location.
universe@119 135 *
universe@119 136 * <b>Attention:</b> the new string references the same memory area as the
universe@119 137 * input string and will <b>NOT</b> be <code>NULL</code>-terminated.
universe@119 138 * Use sstrdup() to get a copy.
universe@119 139 *
universe@119 140 * @param string input string
universe@119 141 * @param start start location of the substring
universe@119 142 * @return a substring of <code>string</code> starting at <code>start</code>
universe@119 143 *
universe@119 144 * @see sstrsubsl()
universe@119 145 * @see sstrchr()
universe@119 146 */
universe@119 147 sstr_t sstrsubs(sstr_t string, size_t start);
universe@119 148
universe@119 149 /**
universe@119 150 * Returns a substring with a maximum length starting at the specified location.
universe@119 151 *
universe@119 152 * <b>Attention:</b> the new string references the same memory area as the
universe@119 153 * input string and will <b>NOT</b> be <code>NULL</code>-terminated.
universe@119 154 * Use sstrdup() to get a copy.
universe@119 155 *
universe@119 156 * @param string input string
universe@119 157 * @param start start location of the substring
universe@119 158 * @param length the maximum length of the substring
universe@119 159 * @return a substring of <code>string</code> starting at <code>start</code>
universe@119 160 * with a maximum length of <code>length</code>
universe@119 161 *
universe@119 162 * @see sstrsubs()
universe@119 163 * @see sstrchr()
universe@119 164 */
universe@119 165 sstr_t sstrsubsl(sstr_t string, size_t start, size_t length);
universe@119 166
universe@119 167 /**
universe@119 168 * Returns a substring starting at the location of the first occurrence of the
universe@119 169 * specified character.
universe@119 170 *
universe@119 171 * If the string does not contain the character, an empty string is returned.
universe@119 172 *
universe@119 173 * @param string the string where to locate the character
universe@119 174 * @param chr the character to locate
universe@148 175 * @return a substring starting at the first location of <code>chr</code>
universe@119 176 *
universe@119 177 * @see sstrsubs()
universe@119 178 */
universe@119 179 sstr_t sstrchr(sstr_t string, int chr);
universe@119 180
universe@119 181 /**
universe@148 182 * Returns a substring starting at the location of the last occurrence of the
universe@148 183 * specified character.
universe@148 184 *
universe@148 185 * If the string does not contain the character, an empty string is returned.
universe@148 186 *
universe@148 187 * @param string the string where to locate the character
universe@148 188 * @param chr the character to locate
universe@148 189 * @return a substring starting at the last location of <code>chr</code>
universe@148 190 *
universe@148 191 * @see sstrsubs()
universe@148 192 */
universe@148 193 sstr_t sstrrchr(sstr_t string, int chr);
universe@148 194
universe@148 195 /**
universe@119 196 * Splits a string into parts by using a delimiter string.
universe@119 197 *
universe@119 198 * This function will return <code>NULL</code>, if one of the following happens:
universe@119 199 * <ul>
universe@119 200 * <li>the string length is zero</li>
universe@119 201 * <li>the delimeter length is zero</li>
universe@119 202 * <li>the string equals the delimeter</li>
universe@119 203 * <li>memory allocation fails</li>
universe@119 204 * </ul>
universe@119 205 *
universe@119 206 * The integer referenced by <code>count</code> is used as input and determines
universe@160 207 * the maximum size of the resulting array, i.e. the maximum count of splits to
universe@119 208 * perform + 1.
universe@119 209 *
universe@119 210 * The integer referenced by <code>count</code> is also used as output and is
universe@119 211 * set to
universe@119 212 * <ul>
universe@119 213 * <li>-2, on memory allocation errors</li>
universe@119 214 * <li>-1, if either the string or the delimiter is an empty string</li>
universe@119 215 * <li>0, if the string equals the delimiter</li>
universe@119 216 * <li>1, if the string does not contain the delimiter</li>
universe@160 217 * <li>the count of array items, otherwise</li>
universe@119 218 * </ul>
universe@119 219 *
universe@119 220 * If the string starts with the delimiter, the first item of the resulting
universe@160 221 * array will be an empty string.
universe@119 222 *
universe@119 223 * If the string ends with the delimiter and the maximum list size is not
universe@160 224 * exceeded, the last array item will be an empty string.
universe@119 225 *
universe@160 226 * <b>Attention:</b> The array pointer <b>AND</b> all sstr_t.ptr of the array
universe@125 227 * items must be manually passed to <code>free()</code>. Use sstrsplit_a() with
universe@119 228 * an allocator to managed memory, to avoid this.
olaf@20 229 *
universe@119 230 * @param string the string to split
universe@119 231 * @param delim the delimiter string
universe@160 232 * @param count IN: the maximum size of the resulting array (0 = no limit),
universe@160 233 * OUT: the actual size of the array
universe@160 234 * @return a sstr_t array containing the split strings or
universe@119 235 * <code>NULL</code> on error
universe@119 236 *
universe@125 237 * @see sstrsplit_a()
olaf@20 238 */
universe@173 239 sstr_t* sstrsplit(sstr_t string, sstr_t delim, ssize_t *count);
olaf@20 240
universe@119 241 /**
universe@119 242 * Performing sstrsplit() using an UcxAllocator.
universe@119 243 *
universe@119 244 * <i>Read the description of sstrsplit() for details.</i>
universe@119 245 *
universe@160 246 * The memory for the sstr_t.ptr pointers of the array items and the memory for
universe@119 247 * the sstr_t array itself are allocated by using the UcxAllocator.malloc()
universe@119 248 * function.
universe@119 249 *
universe@119 250 * <b>Note:</b> the allocator is not used for memory that is freed within the
universe@119 251 * same call of this function (locally scoped variables).
universe@119 252 *
universe@125 253 * @param allocator the UcxAllocator used for allocating memory
universe@119 254 * @param string the string to split
universe@119 255 * @param delim the delimiter string
universe@160 256 * @param count IN: the maximum size of the resulting array (0 = no limit),
universe@160 257 * OUT: the actual size of the array
universe@160 258 * @return a sstr_t array containing the split strings or
universe@119 259 * <code>NULL</code> on error
universe@119 260 *
universe@119 261 * @see sstrsplit()
olaf@20 262 */
universe@125 263 sstr_t* sstrsplit_a(UcxAllocator *allocator, sstr_t string, sstr_t delim,
universe@173 264 ssize_t *count);
olaf@20 265
universe@116 266 /**
universe@116 267 * Compares two UCX strings with standard <code>memcmp()</code>.
universe@116 268 *
universe@116 269 * At first it compares the sstr_t.length attribute of the two strings. The
universe@116 270 * <code>memcmp()</code> function is called, if and only if the lengths match.
universe@116 271 *
universe@116 272 * @param s1 the first string
universe@116 273 * @param s2 the second string
universe@116 274 * @return -1, if the length of s1 is less than the length of s2 or 1, if the
universe@116 275 * length of s1 is greater than the length of s2 or the result of
universe@116 276 * <code>memcmp()</code> otherwise (i.e. 0 if the strings match)
universe@116 277 */
olaf@68 278 int sstrcmp(sstr_t s1, sstr_t s2);
olaf@20 279
universe@116 280 /**
universe@149 281 * Compares two UCX strings ignoring the case.
universe@149 282 *
universe@149 283 * At first it compares the sstr_t.length attribute of the two strings. If and
universe@149 284 * only if the lengths match, both strings are compared char by char ignoring
universe@149 285 * the case.
universe@149 286 *
universe@149 287 * @param s1 the first string
universe@149 288 * @param s2 the second string
universe@149 289 * @return -1, if the length of s1 is less than the length of s2 or 1, if the
universe@149 290 * length of s1 is greater than the length of s2 or the difference between the
universe@149 291 * first two differing characters otherwise (i.e. 0 if the strings match and
universe@149 292 * no characters differ)
universe@149 293 */
universe@149 294 int sstrcasecmp(sstr_t s1, sstr_t s2);
universe@149 295
universe@149 296 /**
universe@116 297 * Creates a duplicate of the specified string.
universe@116 298 *
universe@116 299 * The new sstr_t will contain a copy allocated by standard
universe@116 300 * <code>malloc()</code>. So developers <b>MUST</b> pass the sstr_t.ptr to
universe@116 301 * <code>free()</code>.
universe@116 302 *
universe@118 303 * The sstr_t.ptr of the return value will <i>always</i> be <code>NULL</code>-
universe@118 304 * terminated.
universe@118 305 *
universe@116 306 * @param string the string to duplicate
universe@118 307 * @return a duplicate of the string
universe@125 308 * @see sstrdup_a()
universe@116 309 */
universe@116 310 sstr_t sstrdup(sstr_t string);
olaf@20 311
universe@118 312 /**
universe@118 313 * Creates a duplicate of the specified string using an UcxAllocator.
universe@118 314 *
universe@118 315 * The new sstr_t will contain a copy allocated by the allocators
universe@118 316 * ucx_allocator_malloc function. So it is implementation depended, whether the
universe@118 317 * returned sstr_t.ptr pointer must be passed to the allocators
universe@118 318 * ucx_allocator_free function manually.
universe@118 319 *
universe@118 320 * The sstr_t.ptr of the return value will <i>always</i> be <code>NULL</code>-
universe@118 321 * terminated.
universe@118 322 *
universe@118 323 * @param allocator a valid instance of an UcxAllocator
universe@118 324 * @param string the string to duplicate
universe@118 325 * @return a duplicate of the string
universe@119 326 * @see sstrdup()
universe@118 327 */
universe@125 328 sstr_t sstrdup_a(UcxAllocator *allocator, sstr_t string);
universe@118 329
universe@118 330 /**
universe@118 331 * Omits leading and trailing spaces.
universe@118 332 *
universe@118 333 * This function returns a new sstr_t containing a trimmed version of the
universe@118 334 * specified string.
universe@118 335 *
universe@118 336 * <b>Note:</b> the new sstr_t references the same memory, thus you
universe@118 337 * <b>MUST NOT</b> pass the sstr_t.ptr of the return value to
universe@118 338 * <code>free()</code>. It is also highly recommended to avoid assignments like
universe@118 339 * <code>mystr = sstrtrim(mystr);</code> as you lose the reference to the
universe@118 340 * source string. Assignments of this type are only permitted, if the
universe@118 341 * sstr_t.ptr of the source string does not need to be freed or if another
universe@118 342 * reference to the source string exists.
universe@118 343 *
universe@118 344 * @param string the string that shall be trimmed
universe@118 345 * @return a new sstr_t containing the trimmed string
universe@118 346 */
olaf@96 347 sstr_t sstrtrim(sstr_t string);
olaf@96 348
universe@146 349 /**
universe@146 350 * Checks, if a string has a specific prefix.
universe@146 351 * @param string the string to check
universe@146 352 * @param prefix the prefix the string should have
universe@146 353 * @return 1, if and only if the string has the specified prefix, 0 otherwise
universe@146 354 */
universe@146 355 int sstrprefix(sstr_t string, sstr_t prefix);
universe@146 356
universe@146 357 /**
universe@146 358 * Checks, if a string has a specific suffix.
universe@146 359 * @param string the string to check
universe@146 360 * @param suffix the suffix the string should have
universe@146 361 * @return 1, if and only if the string has the specified suffix, 0 otherwise
universe@146 362 */
universe@146 363 int sstrsuffix(sstr_t string, sstr_t suffix);
universe@146 364
olaf@20 365 #ifdef __cplusplus
olaf@20 366 }
olaf@20 367 #endif
olaf@20 368
universe@116 369 #endif /* UCX_STRING_H */

mercurial