lib/util/charset/charset.h

   1 /*
   2    Unix SMB/CIFS implementation.
   3    charset defines
   4    Copyright (C) Andrew Tridgell 2001
   5    Copyright (C) Jelmer Vernooij 2002
   6
   7    This program is free software; you can redistribute it and/or modify
   8    it under the terms of the GNU General Public License as published by
   9    the Free Software Foundation; either version 3 of the License, or
  10    (at your option) any later version.
  11
  12    This program is distributed in the hope that it will be useful,
  13    but WITHOUT ANY WARRANTY; without even the implied warranty of
  14    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  15    GNU General Public License for more details.
  16
  17    You should have received a copy of the GNU General Public License
  18    along with this program.  If not, see <http://www.gnu.org/licenses/>.
  19 */
  20
  21 /* This is a public header file that is installed as part of Samba.
  22  * If you remove any functions or change their signature, update
  23  * the so version number. */
  24
  25 #ifndef __CHARSET_H__
  26 #define __CHARSET_H__
  27
  28 #include <talloc.h>
  29
  30 /* this defines the charset types used in samba */
  31 typedef enum {CH_UTF16LE=0, CH_UTF16=0, CH_UNIX, CH_DOS, CH_UTF8, CH_UTF16BE, CH_UTF16MUNGED} charset_t;
  32
  33 #define NUM_CHARSETS 7
  34
  35 /*
  36  * SMB UCS2 (16-bit unicode) internal type.
  37  * smb_ucs2_t is *always* in little endian format.
  38  */
  39
  40 typedef uint16_t smb_ucs2_t;
  41
  42 #ifdef WORDS_BIGENDIAN
  43 #define UCS2_SHIFT 8
  44 #else
  45 #define UCS2_SHIFT 0
  46 #endif
  47
  48 /* turn a 7 bit character into a ucs2 character */
  49 #define UCS2_CHAR(c) ((c) << UCS2_SHIFT)
  50
  51 /*
  52  *   for each charset we have a function that pulls from that charset to
  53  *     a ucs2 buffer, and a function that pushes to a ucs2 buffer
  54  *     */
  55
  56 struct charset_functions {
  57         const char *name;
  58         size_t (*pull)(void *, const char **inbuf, size_t *inbytesleft,
  59                                    char **outbuf, size_t *outbytesleft);
  60         size_t (*push)(void *, const char **inbuf, size_t *inbytesleft,
  61                                    char **outbuf, size_t *outbytesleft);
  62         bool samba_internal_charset;
  63 };
  64
  65 /* this type is used for manipulating unicode codepoints */
  66 typedef uint32_t codepoint_t;
  67
  68 #define INVALID_CODEPOINT ((codepoint_t)-1)
  69
  70 /* generic iconv conversion structure */
  71 typedef struct smb_iconv_s {
  72         size_t (*direct)(void *cd, const char **inbuf, size_t *inbytesleft,
  73                          char **outbuf, size_t *outbytesleft);
  74         size_t (*pull)(void *cd, const char **inbuf, size_t *inbytesleft,
  75                        char **outbuf, size_t *outbytesleft);
  76         size_t (*push)(void *cd, const char **inbuf, size_t *inbytesleft,
  77                        char **outbuf, size_t *outbytesleft);
  78         void *cd_direct, *cd_pull, *cd_push;
  79         char *from_name, *to_name;
  80 } *smb_iconv_t;
  81
  82 /* string manipulation flags */
  83 #define STR_TERMINATE 1
  84 #define STR_UPPER 2
  85 #define STR_ASCII 4
  86 #define STR_UNICODE 8
  87 #define STR_NOALIGN 16
  88 #define STR_NO_RANGE_CHECK 32
  89 #define STR_LEN8BIT 64
  90 #define STR_TERMINATE_ASCII 128 /* only terminate if ascii */
  91 #define STR_LEN_NOTERM 256 /* the length field is the unterminated length */
  92
  93 struct loadparm_context;
  94 struct smb_iconv_handle;
  95
  96 char *strchr_m(const char *s, char c);
  97 /**
  98  * Calculate the number of units (8 or 16-bit, depending on the
  99  * destination charset), that would be needed to convert the input
 100  * string which is expected to be in in src_charset encoding to the
 101  * destination charset (which should be a unicode charset).
 102  */
 103 size_t strlen_m_ext_handle(struct smb_iconv_handle *ic,
 104                            const char *s, charset_t src_charset, charset_t dst_charset);
 105 size_t strlen_m_ext(const char *s, charset_t src_charset, charset_t dst_charset);
 106 size_t strlen_m_ext_term(const char *s, charset_t src_charset,
 107                          charset_t dst_charset);
 108 size_t strlen_m_term(const char *s);
 109 size_t strlen_m_term_null(const char *s);
 110 size_t strlen_m(const char *s);
 111 char *alpha_strcpy(char *dest, const char *src, const char *other_safe_chars, size_t maxlength);
 112 void string_replace_m(char *s, char oldc, char newc);
 113 bool strcsequal(const char *s1,const char *s2);
 114 bool strequal_m(const char *s1, const char *s2);
 115 int strncasecmp_m(const char *s1, const char *s2, size_t n);
 116 int strncasecmp_m_handle(struct smb_iconv_handle *iconv_handle,
 117                          const char *s1, const char *s2, size_t n);
 118 bool next_token(const char **ptr,char *buff, const char *sep, size_t bufsize);
 119 int strcasecmp_m_handle(struct smb_iconv_handle *iconv_handle,
 120                         const char *s1, const char *s2);
 121 int strcasecmp_m(const char *s1, const char *s2);
 122 size_t count_chars_m(const char *s, char c);
 123 char *strupper_talloc(TALLOC_CTX *ctx, const char *src);
 124 char *talloc_strdup_upper(TALLOC_CTX *ctx, const char *src);
 125 char *strupper_talloc_n_handle(struct smb_iconv_handle *iconv_handle,
 126                                 TALLOC_CTX *ctx, const char *src, size_t n);
 127 char *strupper_talloc_n(TALLOC_CTX *ctx, const char *src, size_t n);
 128  char *strlower_talloc_handle(struct smb_iconv_handle *iconv_handle,
 129                               TALLOC_CTX *ctx, const char *src);
 130 char *strlower_talloc(TALLOC_CTX *ctx, const char *src);
 131 bool strhasupper(const char *string);
 132 bool strhaslower_handle(struct smb_iconv_handle *ic,
 133                         const char *string);
 134 bool strhaslower(const char *string);
 135 bool strhasupper_handle(struct smb_iconv_handle *ic,
 136                         const char *string);
 137 char *strrchr_m(const char *s, char c);
 138 char *strchr_m(const char *s, char c);
 139 char *strstr_m(const char *src, const char *findstr);
 140
 141 bool push_ascii_talloc(TALLOC_CTX *ctx, char **dest, const char *src, size_t *converted_size);
 142 bool push_ucs2_talloc(TALLOC_CTX *ctx, smb_ucs2_t **dest, const char *src, size_t *converted_size);
 143 bool push_utf8_talloc(TALLOC_CTX *ctx, char **dest, const char *src, size_t *converted_size);
 144 bool pull_ascii_talloc(TALLOC_CTX *ctx, char **dest, const char *src, size_t *converted_size);
 145 bool pull_ucs2_talloc(TALLOC_CTX *ctx, char **dest, const smb_ucs2_t *src, size_t *converted_size);
 146 bool pull_utf8_talloc(TALLOC_CTX *ctx, char **dest, const char *src, size_t *converted_size);
 147 ssize_t push_string(void *dest, const char *src, size_t dest_len, int flags);
 148 ssize_t pull_string(char *dest, const void *src, size_t dest_len, size_t src_len, int flags);
 149
 150 bool convert_string_talloc(TALLOC_CTX *ctx,
 151                            charset_t from, charset_t to,
 152                            void const *src, size_t srclen,
 153                            void *dest, size_t *converted_size);
 154
 155 bool convert_string(charset_t from, charset_t to,
 156                       void const *src, size_t srclen,
 157                       void *dest, size_t destlen,
 158                       size_t *converted_size);
 159 bool convert_string_error(charset_t from, charset_t to,
 160                           void const *src, size_t srclen,
 161                           void *dest, size_t destlen,
 162                           size_t *converted_size);
 163
 164 extern struct smb_iconv_handle *global_iconv_handle;
 165 struct smb_iconv_handle *get_iconv_handle(void);
 166 struct smb_iconv_handle *get_iconv_testing_handle(TALLOC_CTX *mem_ctx,
 167                                                   const char *dos_charset,
 168                                                   const char *unix_charset,
 169                                                   bool use_builtin_handlers);
 170 smb_iconv_t get_conv_handle(struct smb_iconv_handle *ic,
 171                             charset_t from, charset_t to);
 172 const char *charset_name(struct smb_iconv_handle *ic, charset_t ch);
 173
 174 codepoint_t next_codepoint_ext(const char *str, charset_t src_charset,
 175                                size_t *size);
 176 codepoint_t next_codepoint(const char *str, size_t *size);
 177 ssize_t push_codepoint(char *str, codepoint_t c);
 178
 179 /* codepoints */
 180 codepoint_t next_codepoint_handle_ext(struct smb_iconv_handle *ic,
 181                             const char *str, charset_t src_charset,
 182                             size_t *size);
 183 codepoint_t next_codepoint_handle(struct smb_iconv_handle *ic,
 184                             const char *str, size_t *size);
 185 ssize_t push_codepoint_handle(struct smb_iconv_handle *ic,
 186                                 char *str, codepoint_t c);
 187
 188 codepoint_t toupper_m(codepoint_t val);
 189 codepoint_t tolower_m(codepoint_t val);
 190 bool islower_m(codepoint_t val);
 191 bool isupper_m(codepoint_t val);
 192 int codepoint_cmpi(codepoint_t c1, codepoint_t c2);
 193
 194 /* Iconv convenience functions */
 195 struct smb_iconv_handle *smb_iconv_handle_reinit(TALLOC_CTX *mem_ctx,
 196                                                            const char *dos_charset,
 197                                                            const char *unix_charset,
 198                                                            bool use_builtin_handlers,
 199                                                            struct smb_iconv_handle *old_ic);
 200
 201 bool convert_string_handle(struct smb_iconv_handle *ic,
 202                                 charset_t from, charset_t to,
 203                                 void const *src, size_t srclen,
 204                                 void *dest, size_t destlen, size_t *converted_size);
 205 bool convert_string_error_handle(struct smb_iconv_handle *ic,
 206                                  charset_t from, charset_t to,
 207                                  void const *src, size_t srclen,
 208                                  void *dest, size_t destlen,
 209                                  size_t *converted_size);
 210
 211 bool convert_string_talloc_handle(TALLOC_CTX *ctx,
 212                                        struct smb_iconv_handle *ic,
 213                                        charset_t from, charset_t to,
 214                                        void const *src, size_t srclen,
 215                                        void *dest, size_t *converted_size);
 216 /* iconv */
 217 smb_iconv_t smb_iconv_open(const char *tocode, const char *fromcode);
 218 int smb_iconv_close(smb_iconv_t cd);
 219 size_t smb_iconv(smb_iconv_t cd,
 220                  const char **inbuf, size_t *inbytesleft,
 221                  char **outbuf, size_t *outbytesleft);
 222 smb_iconv_t smb_iconv_open_ex(TALLOC_CTX *mem_ctx, const char *tocode,
 223                               const char *fromcode, bool use_builtin_handlers);
 224
 225 void load_case_tables(void);
 226 void load_case_tables_library(void);
 227
 228 /* The following definitions come from util_unistr_w.c  */
 229
 230 size_t strlen_w(const smb_ucs2_t *src);
 231 size_t strnlen_w(const smb_ucs2_t *src, size_t max);
 232 smb_ucs2_t *strchr_w(const smb_ucs2_t *s, smb_ucs2_t c);
 233 smb_ucs2_t *strchr_wa(const smb_ucs2_t *s, char c);
 234 smb_ucs2_t *strrchr_w(const smb_ucs2_t *s, smb_ucs2_t c);
 235 smb_ucs2_t *strnrchr_w(const smb_ucs2_t *s, smb_ucs2_t c, unsigned int n);
 236 smb_ucs2_t *strstr_w(const smb_ucs2_t *s, const smb_ucs2_t *ins);
 237 bool strlower_w(smb_ucs2_t *s);
 238 bool strupper_w(smb_ucs2_t *s);
 239 int strcasecmp_w(const smb_ucs2_t *a, const smb_ucs2_t *b);
 240 int strncasecmp_w(const smb_ucs2_t *a, const smb_ucs2_t *b, size_t len);
 241 int strcmp_wa(const smb_ucs2_t *a, const char *b);
 242 smb_ucs2_t toupper_w(smb_ucs2_t v);
 243
 244 #endif /* __CHARSET_H__ */