1

/* $Id$ */

     2

     3

/** @file string.cpp */

     4

     5

#include "stdafx.h"

     6

#include "openttd.h"

     7

#include "functions.h"

     8

#include "string.h"

     9

#include "macros.h"

    10

#include "table/control_codes.h"

    11

#include "helpers.hpp"

    12

#include "debug.h"

    13

    14

#include <stdarg.h>

    15

#include <ctype.h> // required for tolower()

    16

    17

void ttd_strlcat(char *dst, const char *src, size_t size)

    18

{

    19

	assert(size > 0);

    20

	for (; size > 0 && *dst != '\0'; --size, ++dst) {}

    21

	assert(size > 0);

    22

	while (--size > 0 && *src != '\0') *dst++ = *src++;

    23

	*dst = '\0';

    24

}

    25

    26

    27

void ttd_strlcpy(char *dst, const char *src, size_t size)

    28

{

    29

	assert(size > 0);

    30

	while (--size > 0 && *src != '\0') *dst++ = *src++;

    31

	*dst = '\0';

    32

}

    33

    34

    35

char* strecat(char* dst, const char* src, const char* last)

    36

{

    37

	assert(dst <= last);

    38

	for (; *dst != '\0'; ++dst)

    39

		if (dst == last) return dst;

    40

	for (; *src != '\0' && dst != last; ++dst, ++src) *dst = *src;

    41

	*dst = '\0';

    42

	return strecpy(dst, src, last);

    43

}

    44

    45

    46

char* strecpy(char* dst, const char* src, const char* last)

    47

{

    48

	assert(dst <= last);

    49

	for (; *src != '\0' && dst != last; ++dst, ++src) *dst = *src;

    50

	*dst = '\0';

    51

	if (dst == last && *src != '\0') {

    52

#ifdef STRGEN

    53

		error("String too long for destination buffer");

    54

#else /* STRGEN */

    55

		DEBUG(misc, 0, "String too long for destination buffer");

    56

		*dst = '\0';

    57

#endif /* STRGEN */

    58

	}

    59

	return dst;

    60

}

    61

    62

    63

char* CDECL str_fmt(const char* str, ...)

    64

{

    65

	char buf[4096];

    66

	va_list va;

    67

	int len;

    68

    69

	va_start(va, str);

    70

	len = vsnprintf(buf, lengthof(buf), str, va);

    71

	va_end(va);

    72

	char* p = MallocT<char>(len + 1);

    73

	if (p != NULL) memcpy(p, buf, len + 1);

    74

	return p;

    75

}

    76

    77

    78

void str_validate(char *str)

    79

{

    80

	char *dst = str;

    81

	WChar c;

    82

	size_t len;

    83

    84

	for (len = Utf8Decode(&c, str); c != '\0'; len = Utf8Decode(&c, str)) {

    85

		if (IsPrintable(c) && (c < SCC_SPRITE_START || c > SCC_SPRITE_END ||

    86

			IsValidChar(c - SCC_SPRITE_START, CS_ALPHANUMERAL))) {

    87

			/* Copy the character back. Even if dst is current the same as str

    88

			 * (i.e. no characters have been changed) this is quicker than

    89

			 * moving the pointers ahead by len */

    90

			do {

    91

				*dst++ = *str++;

    92

			} while (--len != 0);

    93

		} else {

    94

			/* Replace the undesirable character with a question mark */

    95

			str += len;

    96

			*dst++ = '?';

    97

		}

    98

	}

    99

   100

	*dst = '\0';

   101

}

   102

   103

   104

void str_strip_colours(char *str)

   105

{

   106

	char *dst = str;

   107

	WChar c;

   108

	size_t len;

   109

   110

	for (len = Utf8Decode(&c, str); c != '\0'; len = Utf8Decode(&c, str)) {

   111

		if (c < SCC_BLUE || c > SCC_BLACK) {

   112

			/* Copy the character back. Even if dst is current the same as str

   113

			 * (i.e. no characters have been changed) this is quicker than

   114

			 * moving the pointers ahead by len */

   115

			do {

   116

				*dst++ = *str++;

   117

			} while (--len != 0);

   118

		} else {

   119

			/* Just skip (strip) the colour codes */

   120

			str += len;

   121

		}

   122

	}

   123

	*dst = '\0';

   124

}

   125

   126

/** Convert a given ASCII string to lowercase.

   127

 * NOTE: only support ASCII characters, no UTF8 fancy. As currently

   128

 * the function is only used to lowercase data-filenames if they are

   129

 * not found, this is sufficient. If more, or general functionality is

   130

 * needed, look to r7271 where it was removed because it was broken when

   131

 * using certain locales: eg in Turkish the uppercase 'I' was converted to

   132

 * '?', so just revert to the old functionality

   133

 * @param str string to convert */

   134

void strtolower(char *str)

   135

{

   136

	for (; *str != '\0'; str++) *str = tolower(*str);

   137

}

   138

   139

/**

   140

 * Only allow certain keys. You can define the filter to be used. This makes

   141

 *  sure no invalid keys can get into an editbox, like BELL.

   142

 * @param key character to be checked

   143

 * @param afilter the filter to use

   144

 * @return true or false depending if the character is printable/valid or not

   145

 */

   146

bool IsValidChar(WChar key, CharSetFilter afilter)

   147

{

   148

	switch (afilter) {

   149

		case CS_ALPHANUMERAL: return IsPrintable(key);

   150

		case CS_NUMERAL:      return (key >= '0' && key <= '9');

   151

		case CS_ALPHA:        return IsPrintable(key) && !(key >= '0' && key <= '9');

   152

	}

   153

   154

	return false;

   155

}

   156

   157

#ifdef WIN32

   158

int CDECL snprintf(char *str, size_t size, const char *format, ...)

   159

{

   160

	va_list ap;

   161

	int ret;

   162

   163

	va_start(ap, format);

   164

	ret = vsnprintf(str, size, format, ap);

   165

	va_end(ap);

   166

	return ret;

   167

}

   168

   169

#ifdef _MSC_VER

   170

/* *nprintf broken, not POSIX compliant, MSDN description

   171

 * - If len < count, then len characters are stored in buffer, a null-terminator is appended, and len is returned.

   172

 * - If len = count, then len characters are stored in buffer, no null-terminator is appended, and len is returned.

   173

 * - If len > count, then count characters are stored in buffer, no null-terminator is appended, and a negative value is returned

   174

 */

   175

int CDECL vsnprintf(char *str, size_t size, const char *format, va_list ap)

   176

{

   177

	int ret;

   178

	ret = _vsnprintf(str, size, format, ap);

   179

	if (ret < 0 || ret == size) str[size - 1] = '\0';

   180

	return ret;

   181

}

   182

#endif /* _MSC_VER */

   183

   184

#endif /* WIN32 */

   185

   186

   187

/** Convert the md5sum to a hexadecimal string representation

   188

 * @param buf buffer to put the md5sum into

   189

 * @param last last character of buffer (usually lastof(buf))

   190

 * @param md5sum the md5sum itself

   191

 * @return a pointer to the next character after the md5sum */

   192

char *md5sumToString(char *buf, const char *last, const uint8 md5sum[16])

   193

{

   194

	char *p = buf;

   195

   196

	for (uint i = 0; i < 16; i++) {

   197

		p += snprintf(p, last + 1 - p, "%02X", md5sum[i]);

   198

		if (p >= last) break;

   199

	}

   200

   201

	return p;

   202

}

   203

   204

   205

/* UTF-8 handling routines */

   206

   207

   208

/* Decode and consume the next UTF-8 encoded character

   209

 * @param c Buffer to place decoded character.

   210

 * @param s Character stream to retrieve character from.

   211

 * @return Number of characters in the sequence.

   212

 */

   213

size_t Utf8Decode(WChar *c, const char *s)

   214

{

   215

	assert(c != NULL);

   216

   217

	if (!HASBIT(s[0], 7)) {

   218

		/* Single byte character: 0xxxxxxx */

   219

		*c = s[0];

   220

		return 1;

   221

	} else if (GB(s[0], 5, 3) == 6) {

   222

		if (IsUtf8Part(s[1])) {

   223

			/* Double byte character: 110xxxxx 10xxxxxx */

   224

			*c = GB(s[0], 0, 5) << 6 | GB(s[1], 0, 6);

   225

			if (*c >= 0x80) return 2;

   226

		}

   227

	} else if (GB(s[0], 4, 4) == 14) {

   228

		if (IsUtf8Part(s[1]) && IsUtf8Part(s[2])) {

   229

			/* Triple byte character: 1110xxxx 10xxxxxx 10xxxxxx */

   230

			*c = GB(s[0], 0, 4) << 12 | GB(s[1], 0, 6) << 6 | GB(s[2], 0, 6);

   231

			if (*c >= 0x800) return 3;

   232

		}

   233

	} else if (GB(s[0], 3, 5) == 30) {

   234

		if (IsUtf8Part(s[1]) && IsUtf8Part(s[2]) && IsUtf8Part(s[3])) {

   235

			/* 4 byte character: 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx */

   236

			*c = GB(s[0], 0, 3) << 18 | GB(s[1], 0, 6) << 12 | GB(s[2], 0, 6) << 6 | GB(s[3], 0, 6);

   237

			if (*c >= 0x10000 && *c <= 0x10FFFF) return 4;

   238

		}

   239

	}

   240

   241

	//DEBUG(misc, 1, "[utf8] invalid UTF-8 sequence");

   242

	*c = '?';

   243

	return 1;

   244

}

   245

   246

   247

/* Encode a unicode character and place it in the buffer

   248

 * @param buf Buffer to place character.

   249

 * @param c   Unicode character to encode.

   250

 * @return Number of characters in the encoded sequence.

   251

 */

   252

size_t Utf8Encode(char *buf, WChar c)

   253

{

   254

	if (c < 0x80) {

   255

		*buf = c;

   256

		return 1;

   257

	} else if (c < 0x800) {

   258

		*buf++ = 0xC0 + GB(c,  6, 5);

   259

		*buf   = 0x80 + GB(c,  0, 6);

   260

		return 2;

   261

	} else if (c < 0x10000) {

   262

		*buf++ = 0xE0 + GB(c, 12, 4);

   263

		*buf++ = 0x80 + GB(c,  6, 6);

   264

		*buf   = 0x80 + GB(c,  0, 6);

   265

		return 3;

   266

	} else if (c < 0x110000) {

   267

		*buf++ = 0xF0 + GB(c, 18, 3);

   268

		*buf++ = 0x80 + GB(c, 12, 6);

   269

		*buf++ = 0x80 + GB(c,  6, 6);

   270

		*buf   = 0x80 + GB(c,  0, 6);

   271

		return 4;

   272

	}

   273

   274

	//DEBUG(misc, 1, "[utf8] can't UTF-8 encode value 0x%X", c);

   275

	*buf = '?';

   276

	return 1;

   277

}

   278

   279

/**

   280

 * Properly terminate an UTF8 string to some maximum length

   281

 * @param s string to check if it needs additional trimming

   282

 * @param maxlen the maximum length the buffer can have.

   283

 * @return the new length in bytes of the string (eg. strlen(new_string))

   284

 * @NOTE maxlen is the string length _INCLUDING_ the terminating '\0'

   285

 */

   286

size_t Utf8TrimString(char *s, size_t maxlen)

   287

{

   288

	size_t length = 0;

   289

   290

	for (const char *ptr = strchr(s, '\0'); *s != '\0';) {

   291

		size_t len = Utf8EncodedCharLen(*s);

   292

		/* Silently ignore invalid UTF8 sequences, our only concern trimming */

   293

		if (len == 0) len = 1;

   294

   295

		/* Take care when a hard cutoff was made for the string and

   296

		 * the last UTF8 sequence is invalid */

   297

		if (length + len >= maxlen || (s + len > ptr)) break;

   298

		s += len;

   299

		length += len;

   300

	}

   301

   302

	*s = '\0';

   303

	return length;

   304

}