string.cpp

Go to the documentation of this file.
00001 /* $Id: string.cpp 19875 2010-05-21 16:03:29Z 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 != '\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   }
00151 
00152   *dst = '\0';
00153 }
00154 
00155 
00156 void str_strip_colours(char *str)
00157 {
00158   char *dst = str;
00159   WChar c;
00160   size_t len;
00161 
00162   for (len = Utf8Decode(&c, str); c != '\0'; len = Utf8Decode(&c, str)) {
00163     if (c < SCC_BLUE || c > SCC_BLACK) {
00164       /* Copy the character back. Even if dst is current the same as str
00165        * (i.e. no characters have been changed) this is quicker than
00166        * moving the pointers ahead by len */
00167       do {
00168         *dst++ = *str++;
00169       } while (--len != 0);
00170     } else {
00171       /* Just skip (strip) the colour codes */
00172       str += len;
00173     }
00174   }
00175   *dst = '\0';
00176 }
00177 
00186 void strtolower(char *str)
00187 {
00188   for (; *str != '\0'; str++) *str = tolower(*str);
00189 }
00190 
00198 bool IsValidChar(WChar key, CharSetFilter afilter)
00199 {
00200   switch (afilter) {
00201     case CS_ALPHANUMERAL:  return IsPrintable(key);
00202     case CS_NUMERAL:       return (key >= '0' && key <= '9');
00203     case CS_NUMERAL_SPACE: return (key >= '0' && key <= '9') || key == ' ';
00204     case CS_ALPHA:         return IsPrintable(key) && !(key >= '0' && key <= '9');
00205   }
00206 
00207   return false;
00208 }
00209 
00210 #ifdef WIN32
00211 /* Since version 3.14, MinGW Runtime has snprintf() and vsnprintf() conform to C99 but it's not the case for older versions */
00212 #if (__MINGW32_MAJOR_VERSION < 3) || ((__MINGW32_MAJOR_VERSION == 3) && (__MINGW32_MINOR_VERSION < 14))
00213 int CDECL snprintf(char *str, size_t size, const char *format, ...)
00214 {
00215   va_list ap;
00216   int ret;
00217 
00218   va_start(ap, format);
00219   ret = vsnprintf(str, size, format, ap);
00220   va_end(ap);
00221   return ret;
00222 }
00223 #endif /* MinGW Runtime < 3.14 */
00224 
00225 #ifdef _MSC_VER
00226 
00233 int CDECL vsnprintf(char *str, size_t size, const char *format, va_list ap)
00234 {
00235   if (size == 0) return 0;
00236 
00237   errno = 0;
00238   int ret = _vsnprintf(str, size, format, ap);
00239 
00240   if (ret < 0) {
00241     if (errno != ERANGE) {
00242       /* There's a formatting error, better get that looked
00243        * at properly instead of ignoring it. */
00244       NOT_REACHED();
00245     }
00246   } else if ((size_t)ret < size) {
00247     /* The buffer is big enough for the number of
00248      * characers stored (excluding null), i.e.
00249      * the string has been null-terminated. */
00250     return ret;
00251   }
00252 
00253   /* The buffer is too small for _vsnprintf to write the
00254    * null-terminator at its end and return size. */
00255   str[size - 1] = '\0';
00256   return (int)size;
00257 }
00258 #endif /* _MSC_VER */
00259 
00260 #endif /* WIN32 */
00261 
00271 int CDECL seprintf(char *str, const char *last, const char *format, ...)
00272 {
00273   va_list ap;
00274 
00275   va_start(ap, format);
00276   int ret = vseprintf(str, last, format, ap);
00277   va_end(ap);
00278   return ret;
00279 }
00280 
00281 
00287 char *md5sumToString(char *buf, const char *last, const uint8 md5sum[16])
00288 {
00289   char *p = buf;
00290 
00291   for (uint i = 0; i < 16; i++) {
00292     p += seprintf(p, last, "%02X", md5sum[i]);
00293   }
00294 
00295   return p;
00296 }
00297 
00298 
00299 /* UTF-8 handling routines */
00300 
00301 
00302 /* Decode and consume the next UTF-8 encoded character
00303  * @param c Buffer to place decoded character.
00304  * @param s Character stream to retrieve character from.
00305  * @return Number of characters in the sequence.
00306  */
00307 size_t Utf8Decode(WChar *c, const char *s)
00308 {
00309   assert(c != NULL);
00310 
00311   if (!HasBit(s[0], 7)) {
00312     /* Single byte character: 0xxxxxxx */
00313     *c = s[0];
00314     return 1;
00315   } else if (GB(s[0], 5, 3) == 6) {
00316     if (IsUtf8Part(s[1])) {
00317       /* Double byte character: 110xxxxx 10xxxxxx */
00318       *c = GB(s[0], 0, 5) << 6 | GB(s[1], 0, 6);
00319       if (*c >= 0x80) return 2;
00320     }
00321   } else if (GB(s[0], 4, 4) == 14) {
00322     if (IsUtf8Part(s[1]) && IsUtf8Part(s[2])) {
00323       /* Triple byte character: 1110xxxx 10xxxxxx 10xxxxxx */
00324       *c = GB(s[0], 0, 4) << 12 | GB(s[1], 0, 6) << 6 | GB(s[2], 0, 6);
00325       if (*c >= 0x800) return 3;
00326     }
00327   } else if (GB(s[0], 3, 5) == 30) {
00328     if (IsUtf8Part(s[1]) && IsUtf8Part(s[2]) && IsUtf8Part(s[3])) {
00329       /* 4 byte character: 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx */
00330       *c = GB(s[0], 0, 3) << 18 | GB(s[1], 0, 6) << 12 | GB(s[2], 0, 6) << 6 | GB(s[3], 0, 6);
00331       if (*c >= 0x10000 && *c <= 0x10FFFF) return 4;
00332     }
00333   }
00334 
00335   /* DEBUG(misc, 1, "[utf8] invalid UTF-8 sequence"); */
00336   *c = '?';
00337   return 1;
00338 }
00339 
00340 
00341 /* Encode a unicode character and place it in the buffer
00342  * @param buf Buffer to place character.
00343  * @param c   Unicode character to encode.
00344  * @return Number of characters in the encoded sequence.
00345  */
00346 size_t Utf8Encode(char *buf, WChar c)
00347 {
00348   if (c < 0x80) {
00349     *buf = c;
00350     return 1;
00351   } else if (c < 0x800) {
00352     *buf++ = 0xC0 + GB(c,  6, 5);
00353     *buf   = 0x80 + GB(c,  0, 6);
00354     return 2;
00355   } else if (c < 0x10000) {
00356     *buf++ = 0xE0 + GB(c, 12, 4);
00357     *buf++ = 0x80 + GB(c,  6, 6);
00358     *buf   = 0x80 + GB(c,  0, 6);
00359     return 3;
00360   } else if (c < 0x110000) {
00361     *buf++ = 0xF0 + GB(c, 18, 3);
00362     *buf++ = 0x80 + GB(c, 12, 6);
00363     *buf++ = 0x80 + GB(c,  6, 6);
00364     *buf   = 0x80 + GB(c,  0, 6);
00365     return 4;
00366   }
00367 
00368   /* DEBUG(misc, 1, "[utf8] can't UTF-8 encode value 0x%X", c); */
00369   *buf = '?';
00370   return 1;
00371 }
00372 
00380 size_t Utf8TrimString(char *s, size_t maxlen)
00381 {
00382   size_t length = 0;
00383 
00384   for (const char *ptr = strchr(s, '\0'); *s != '\0';) {
00385     size_t len = Utf8EncodedCharLen(*s);
00386     /* Silently ignore invalid UTF8 sequences, our only concern trimming */
00387     if (len == 0) len = 1;
00388 
00389     /* Take care when a hard cutoff was made for the string and
00390      * the last UTF8 sequence is invalid */
00391     if (length + len >= maxlen || (s + len > ptr)) break;
00392     s += len;
00393     length += len;
00394   }
00395 
00396   *s = '\0';
00397   return length;
00398 }
00399 
00400 #ifdef DEFINE_STRNDUP
00401 #include "core/math_func.hpp"
00402 char *strndup(const char *s, size_t len)
00403 {
00404   len = min(strlen(s), len);
00405   char *tmp = CallocT<char>(len + 1);
00406   memcpy(tmp, s, len);
00407   return tmp;
00408 }
00409 #endif /* DEFINE_STRNDUP */
00410 
00411 #ifdef DEFINE_STRCASESTR
00412 char *strcasestr(const char *haystack, const char *needle)
00413 {
00414   size_t hay_len = strlen(haystack);
00415   size_t needle_len = strlen(needle);
00416   while (hay_len >= needle_len) {
00417     if (strncasecmp(haystack, needle, needle_len) == 0) return const_cast<char *>(haystack);
00418 
00419     haystack++;
00420     hay_len--;
00421   }
00422 
00423   return NULL;
00424 }
00425 #endif /* DEFINE_STRCASESTR */

Generated on Sat Jun 19 17:14:54 2010 for OpenTTD by  doxygen 1.6.1