string.cpp

Go to the documentation of this file.
00001 /* $Id: string.cpp 20428 2010-08-09 21:17:19Z rubidium $ */
00002 
00003 /*
00004  * This file is part of OpenTTD.
00005  * OpenTTD is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 2.
00006  * OpenTTD is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
00007  * See the GNU General Public License for more details. You should have received a copy of the GNU General Public License along with OpenTTD. If not, see <http://www.gnu.org/licenses/>.
00008  */
00009 
00012 #include "stdafx.h"
00013 #include "debug.h"
00014 #include "core/alloc_func.hpp"
00015 #include "core/math_func.hpp"
00016 #include "string_func.h"
00017 
00018 #include "table/control_codes.h"
00019 
00020 #include <stdarg.h>
00021 #include <ctype.h> /* required for tolower() */
00022 
00023 #ifdef _MSC_VER
00024 #include <errno.h> // required by vsnprintf implementation for MSVC
00025 #endif
00026 
00037 static int CDECL vseprintf(char *str, const char *last, const char *format, va_list ap)
00038 {
00039   ptrdiff_t diff = last - str;
00040   if (diff < 0) return 0;
00041   return min((int)diff, vsnprintf(str, diff + 1, format, ap));
00042 }
00043 
00044 void ttd_strlcat(char *dst, const char *src, size_t size)
00045 {
00046   assert(size > 0);
00047   while (size > 0 && *dst != '\0') {
00048     size--;
00049     dst++;
00050   }
00051 
00052   ttd_strlcpy(dst, src, size);
00053 }
00054 
00055 
00056 void ttd_strlcpy(char *dst, const char *src, size_t size)
00057 {
00058   assert(size > 0);
00059   while (--size > 0 && *src != '\0') {
00060     *dst++ = *src++;
00061   }
00062   *dst = '\0';
00063 }
00064 
00065 
00066 char *strecat(char *dst, const char *src, const char *last)
00067 {
00068   assert(dst <= last);
00069   while (*dst != '\0') {
00070     if (dst == last) return dst;
00071     dst++;
00072   }
00073 
00074   return strecpy(dst, src, last);
00075 }
00076 
00077 
00078 char *strecpy(char *dst, const char *src, const char *last)
00079 {
00080   assert(dst <= last);
00081   while (dst != last && *src != '\0') {
00082     *dst++ = *src++;
00083   }
00084   *dst = '\0';
00085 
00086   if (dst == last && *src != '\0') {
00087 #ifdef STRGEN
00088     error("String too long for destination buffer");
00089 #else /* STRGEN */
00090     DEBUG(misc, 0, "String too long for destination buffer");
00091 #endif /* STRGEN */
00092   }
00093   return dst;
00094 }
00095 
00096 
00097 char *CDECL str_fmt(const char *str, ...)
00098 {
00099   char buf[4096];
00100   va_list va;
00101 
00102   va_start(va, str);
00103   int len = vseprintf(buf, lastof(buf), str, va);
00104   va_end(va);
00105   char *p = MallocT<char>(len + 1);
00106   memcpy(p, buf, len + 1);
00107   return p;
00108 }
00109 
00110 
00111 void str_validate(char *str, const char *last, bool allow_newlines, bool ignore)
00112 {
00113   /* Assume the ABSOLUTE WORST to be in str as it comes from the outside. */
00114 
00115   char *dst = str;
00116   while (str <= last && *str != '\0') {
00117     size_t len = Utf8EncodedCharLen(*str);
00118     /* If the character is unknown, i.e. encoded length is 0
00119      * we assume worst case for the length check.
00120      * The length check is needed to prevent Utf8Decode to read
00121      * over the terminating '\0' if that happens to be placed
00122      * within the encoding of an UTF8 character. */
00123     if ((len == 0 && str + 4 > last) || str + len > last) break;
00124 
00125     WChar c;
00126     len = Utf8Decode(&c, str);
00127     /* It's possible to encode the string termination character
00128      * into a multiple bytes. This prevents those termination
00129      * characters to be skipped */
00130     if (c == '\0') break;
00131 
00132     if (IsPrintable(c) && (c < SCC_SPRITE_START || c > SCC_SPRITE_END)) {
00133       /* Copy the character back. Even if dst is current the same as str
00134        * (i.e. no characters have been changed) this is quicker than
00135        * moving the pointers ahead by len */
00136       do {
00137         *dst++ = *str++;
00138       } while (--len != 0);
00139     } else if (allow_newlines && c == '\n') {
00140       *dst++ = *str++;
00141     } else {
00142       if (allow_newlines && c == '\r' && str[1] == '\n') {
00143         str += len;
00144         continue;
00145       }
00146       /* Replace the undesirable character with a question mark */
00147       str += len;
00148       if (!ignore) *dst++ = '?';
00149 
00150       /* In case of these two special cases assume that they really
00151        * mean SETX/SETXY and also "eat" the paramater. If this was
00152        * not the case the string was broken to begin with and this
00153        * would not break much more. */
00154       if (c == SCC_SETX) {
00155         str++;
00156       } else if (c == SCC_SETXY) {
00157         str += 2;
00158       }
00159     }
00160   }
00161 
00162   *dst = '\0';
00163 }
00164 
00165 
00166 void str_strip_colours(char *str)
00167 {
00168   char *dst = str;
00169   WChar c;
00170   size_t len;
00171 
00172   for (len = Utf8Decode(&c, str); c != '\0'; len = Utf8Decode(&c, str)) {
00173     if (c < SCC_BLUE || c > SCC_BLACK) {
00174       /* Copy the character back. Even if dst is current the same as str
00175        * (i.e. no characters have been changed) this is quicker than
00176        * moving the pointers ahead by len */
00177       do {
00178         *dst++ = *str++;
00179       } while (--len != 0);
00180     } else {
00181       /* Just skip (strip) the colour codes */
00182       str += len;
00183     }
00184   }
00185   *dst = '\0';
00186 }
00187 
00196 void strtolower(char *str)
00197 {
00198   for (; *str != '\0'; str++) *str = tolower(*str);
00199 }
00200 
00208 bool IsValidChar(WChar key, CharSetFilter afilter)
00209 {
00210   switch (afilter) {
00211     case CS_ALPHANUMERAL:  return IsPrintable(key);
00212     case CS_NUMERAL:       return (key >= '0' && key <= '9');
00213     case CS_NUMERAL_SPACE: return (key >= '0' && key <= '9') || key == ' ';
00214     case CS_ALPHA:         return IsPrintable(key) && !(key >= '0' && key <= '9');
00215   }
00216 
00217   return false;
00218 }
00219 
00220 #ifdef WIN32
00221 /* Since version 3.14, MinGW Runtime has snprintf() and vsnprintf() conform to C99 but it's not the case for older versions */
00222 #if (__MINGW32_MAJOR_VERSION < 3) || ((__MINGW32_MAJOR_VERSION == 3) && (__MINGW32_MINOR_VERSION < 14))
00223 int CDECL snprintf(char *str, size_t size, const char *format, ...)
00224 {
00225   va_list ap;
00226   int ret;
00227 
00228   va_start(ap, format);
00229   ret = vsnprintf(str, size, format, ap);
00230   va_end(ap);
00231   return ret;
00232 }
00233 #endif /* MinGW Runtime < 3.14 */
00234 
00235 #ifdef _MSC_VER
00236 
00243 int CDECL vsnprintf(char *str, size_t size, const char *format, va_list ap)
00244 {
00245   if (size == 0) return 0;
00246 
00247   errno = 0;
00248   int ret = _vsnprintf(str, size, format, ap);
00249 
00250   if (ret < 0) {
00251     if (errno != ERANGE) {
00252       /* There's a formatting error, better get that looked
00253        * at properly instead of ignoring it. */
00254       NOT_REACHED();
00255     }
00256   } else if ((size_t)ret < size) {
00257     /* The buffer is big enough for the number of
00258      * characers stored (excluding null), i.e.
00259      * the string has been null-terminated. */
00260     return ret;
00261   }
00262 
00263   /* The buffer is too small for _vsnprintf to write the
00264    * null-terminator at its end and return size. */
00265   str[size - 1] = '\0';
00266   return (int)size;
00267 }
00268 #endif /* _MSC_VER */
00269 
00270 #endif /* WIN32 */
00271 
00281 int CDECL seprintf(char *str, const char *last, const char *format, ...)
00282 {
00283   va_list ap;
00284 
00285   va_start(ap, format);
00286   int ret = vseprintf(str, last, format, ap);
00287   va_end(ap);
00288   return ret;
00289 }
00290 
00291 
00297 char *md5sumToString(char *buf, const char *last, const uint8 md5sum[16])
00298 {
00299   char *p = buf;
00300 
00301   for (uint i = 0; i < 16; i++) {
00302     p += seprintf(p, last, "%02X", md5sum[i]);
00303   }
00304 
00305   return p;
00306 }
00307 
00308 
00309 /* UTF-8 handling routines */
00310 
00311 
00312 /* Decode and consume the next UTF-8 encoded character
00313  * @param c Buffer to place decoded character.
00314  * @param s Character stream to retrieve character from.
00315  * @return Number of characters in the sequence.
00316  */
00317 size_t Utf8Decode(WChar *c, const char *s)
00318 {
00319   assert(c != NULL);
00320 
00321   if (!HasBit(s[0], 7)) {
00322     /* Single byte character: 0xxxxxxx */
00323     *c = s[0];
00324     return 1;
00325   } else if (GB(s[0], 5, 3) == 6) {
00326     if (IsUtf8Part(s[1])) {
00327       /* Double byte character: 110xxxxx 10xxxxxx */
00328       *c = GB(s[0], 0, 5) << 6 | GB(s[1], 0, 6);
00329       if (*c >= 0x80) return 2;
00330     }
00331   } else if (GB(s[0], 4, 4) == 14) {
00332     if (IsUtf8Part(s[1]) && IsUtf8Part(s[2])) {
00333       /* Triple byte character: 1110xxxx 10xxxxxx 10xxxxxx */
00334       *c = GB(s[0], 0, 4) << 12 | GB(s[1], 0, 6) << 6 | GB(s[2], 0, 6);
00335       if (*c >= 0x800) return 3;
00336     }
00337   } else if (GB(s[0], 3, 5) == 30) {
00338     if (IsUtf8Part(s[1]) && IsUtf8Part(s[2]) && IsUtf8Part(s[3])) {
00339       /* 4 byte character: 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx */
00340       *c = GB(s[0], 0, 3) << 18 | GB(s[1], 0, 6) << 12 | GB(s[2], 0, 6) << 6 | GB(s[3], 0, 6);
00341       if (*c >= 0x10000 && *c <= 0x10FFFF) return 4;
00342     }
00343   }
00344 
00345   /* DEBUG(misc, 1, "[utf8] invalid UTF-8 sequence"); */
00346   *c = '?';
00347   return 1;
00348 }
00349 
00350 
00351 /* Encode a unicode character and place it in the buffer
00352  * @param buf Buffer to place character.
00353  * @param c   Unicode character to encode.
00354  * @return Number of characters in the encoded sequence.
00355  */
00356 size_t Utf8Encode(char *buf, WChar c)
00357 {
00358   if (c < 0x80) {
00359     *buf = c;
00360     return 1;
00361   } else if (c < 0x800) {
00362     *buf++ = 0xC0 + GB(c,  6, 5);
00363     *buf   = 0x80 + GB(c,  0, 6);
00364     return 2;
00365   } else if (c < 0x10000) {
00366     *buf++ = 0xE0 + GB(c, 12, 4);
00367     *buf++ = 0x80 + GB(c,  6, 6);
00368     *buf   = 0x80 + GB(c,  0, 6);
00369     return 3;
00370   } else if (c < 0x110000) {
00371     *buf++ = 0xF0 + GB(c, 18, 3);
00372     *buf++ = 0x80 + GB(c, 12, 6);
00373     *buf++ = 0x80 + GB(c,  6, 6);
00374     *buf   = 0x80 + GB(c,  0, 6);
00375     return 4;
00376   }
00377 
00378   /* DEBUG(misc, 1, "[utf8] can't UTF-8 encode value 0x%X", c); */
00379   *buf = '?';
00380   return 1;
00381 }
00382 
00390 size_t Utf8TrimString(char *s, size_t maxlen)
00391 {
00392   size_t length = 0;
00393 
00394   for (const char *ptr = strchr(s, '\0'); *s != '\0';) {
00395     size_t len = Utf8EncodedCharLen(*s);
00396     /* Silently ignore invalid UTF8 sequences, our only concern trimming */
00397     if (len == 0) len = 1;
00398 
00399     /* Take care when a hard cutoff was made for the string and
00400      * the last UTF8 sequence is invalid */
00401     if (length + len >= maxlen || (s + len > ptr)) break;
00402     s += len;
00403     length += len;
00404   }
00405 
00406   *s = '\0';
00407   return length;
00408 }
00409 
00410 #ifdef DEFINE_STRNDUP
00411 #include "core/math_func.hpp"
00412 char *strndup(const char *s, size_t len)
00413 {
00414   len = min(strlen(s), len);
00415   char *tmp = CallocT<char>(len + 1);
00416   memcpy(tmp, s, len);
00417   return tmp;
00418 }
00419 #endif /* DEFINE_STRNDUP */
00420 
00421 #ifdef DEFINE_STRCASESTR
00422 char *strcasestr(const char *haystack, const char *needle)
00423 {
00424   size_t hay_len = strlen(haystack);
00425   size_t needle_len = strlen(needle);
00426   while (hay_len >= needle_len) {
00427     if (strncasecmp(haystack, needle, needle_len) == 0) return const_cast<char *>(haystack);
00428 
00429     haystack++;
00430     hay_len--;
00431   }
00432 
00433   return NULL;
00434 }
00435 #endif /* DEFINE_STRCASESTR */

Generated on Mon Aug 30 19:37:00 2010 for OpenTTD by  doxygen 1.6.1