Serhiy Storchaka | ab8bcb3 | 2016-07-03 13:26:52 +0300 | [diff] [blame] | 1 | #define PY_SSIZE_T_CLEAN |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 2 | #include "Python.h" |
Victor Stinner | a15e260 | 2020-04-08 02:01:56 +0200 | [diff] [blame] | 3 | #include "pycore_abstract.h" // _PyIndex_Check() |
Victor Stinner | 45876a9 | 2020-02-12 22:32:34 +0100 | [diff] [blame] | 4 | #include "pycore_bytes_methods.h" |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 5 | |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 6 | PyDoc_STRVAR_shared(_Py_isspace__doc__, |
| 7 | "B.isspace() -> bool\n\ |
| 8 | \n\ |
| 9 | Return True if all characters in B are whitespace\n\ |
| 10 | and there is at least one character in B, False otherwise."); |
| 11 | |
| 12 | PyObject* |
| 13 | _Py_bytes_isspace(const char *cptr, Py_ssize_t len) |
| 14 | { |
Antoine Pitrou | 9ed5f27 | 2013-08-13 20:18:52 +0200 | [diff] [blame] | 15 | const unsigned char *p |
Andy Lester | e6be9b5 | 2020-02-11 20:28:35 -0600 | [diff] [blame] | 16 | = (const unsigned char *) cptr; |
Antoine Pitrou | 9ed5f27 | 2013-08-13 20:18:52 +0200 | [diff] [blame] | 17 | const unsigned char *e; |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 18 | |
| 19 | /* Shortcut for single character strings */ |
Eric Smith | 6dc46f5 | 2009-04-27 20:39:49 +0000 | [diff] [blame] | 20 | if (len == 1 && Py_ISSPACE(*p)) |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 21 | Py_RETURN_TRUE; |
| 22 | |
| 23 | /* Special case for empty strings */ |
| 24 | if (len == 0) |
| 25 | Py_RETURN_FALSE; |
| 26 | |
| 27 | e = p + len; |
| 28 | for (; p < e; p++) { |
Antoine Pitrou | f95a1b3 | 2010-05-09 15:52:27 +0000 | [diff] [blame] | 29 | if (!Py_ISSPACE(*p)) |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 30 | Py_RETURN_FALSE; |
| 31 | } |
| 32 | Py_RETURN_TRUE; |
| 33 | } |
| 34 | |
| 35 | |
| 36 | PyDoc_STRVAR_shared(_Py_isalpha__doc__, |
| 37 | "B.isalpha() -> bool\n\ |
| 38 | \n\ |
| 39 | Return True if all characters in B are alphabetic\n\ |
| 40 | and there is at least one character in B, False otherwise."); |
| 41 | |
| 42 | PyObject* |
| 43 | _Py_bytes_isalpha(const char *cptr, Py_ssize_t len) |
| 44 | { |
Antoine Pitrou | 9ed5f27 | 2013-08-13 20:18:52 +0200 | [diff] [blame] | 45 | const unsigned char *p |
Andy Lester | e6be9b5 | 2020-02-11 20:28:35 -0600 | [diff] [blame] | 46 | = (const unsigned char *) cptr; |
Antoine Pitrou | 9ed5f27 | 2013-08-13 20:18:52 +0200 | [diff] [blame] | 47 | const unsigned char *e; |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 48 | |
| 49 | /* Shortcut for single character strings */ |
Eric Smith | 6dc46f5 | 2009-04-27 20:39:49 +0000 | [diff] [blame] | 50 | if (len == 1 && Py_ISALPHA(*p)) |
Antoine Pitrou | f95a1b3 | 2010-05-09 15:52:27 +0000 | [diff] [blame] | 51 | Py_RETURN_TRUE; |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 52 | |
| 53 | /* Special case for empty strings */ |
| 54 | if (len == 0) |
Antoine Pitrou | f95a1b3 | 2010-05-09 15:52:27 +0000 | [diff] [blame] | 55 | Py_RETURN_FALSE; |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 56 | |
| 57 | e = p + len; |
| 58 | for (; p < e; p++) { |
Antoine Pitrou | f95a1b3 | 2010-05-09 15:52:27 +0000 | [diff] [blame] | 59 | if (!Py_ISALPHA(*p)) |
| 60 | Py_RETURN_FALSE; |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 61 | } |
| 62 | Py_RETURN_TRUE; |
| 63 | } |
| 64 | |
| 65 | |
| 66 | PyDoc_STRVAR_shared(_Py_isalnum__doc__, |
| 67 | "B.isalnum() -> bool\n\ |
| 68 | \n\ |
| 69 | Return True if all characters in B are alphanumeric\n\ |
| 70 | and there is at least one character in B, False otherwise."); |
| 71 | |
| 72 | PyObject* |
| 73 | _Py_bytes_isalnum(const char *cptr, Py_ssize_t len) |
| 74 | { |
Antoine Pitrou | 9ed5f27 | 2013-08-13 20:18:52 +0200 | [diff] [blame] | 75 | const unsigned char *p |
Andy Lester | e6be9b5 | 2020-02-11 20:28:35 -0600 | [diff] [blame] | 76 | = (const unsigned char *) cptr; |
Antoine Pitrou | 9ed5f27 | 2013-08-13 20:18:52 +0200 | [diff] [blame] | 77 | const unsigned char *e; |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 78 | |
| 79 | /* Shortcut for single character strings */ |
Eric Smith | 6dc46f5 | 2009-04-27 20:39:49 +0000 | [diff] [blame] | 80 | if (len == 1 && Py_ISALNUM(*p)) |
Antoine Pitrou | f95a1b3 | 2010-05-09 15:52:27 +0000 | [diff] [blame] | 81 | Py_RETURN_TRUE; |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 82 | |
| 83 | /* Special case for empty strings */ |
| 84 | if (len == 0) |
Antoine Pitrou | f95a1b3 | 2010-05-09 15:52:27 +0000 | [diff] [blame] | 85 | Py_RETURN_FALSE; |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 86 | |
| 87 | e = p + len; |
| 88 | for (; p < e; p++) { |
Antoine Pitrou | f95a1b3 | 2010-05-09 15:52:27 +0000 | [diff] [blame] | 89 | if (!Py_ISALNUM(*p)) |
| 90 | Py_RETURN_FALSE; |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 91 | } |
| 92 | Py_RETURN_TRUE; |
| 93 | } |
| 94 | |
| 95 | |
INADA Naoki | a49ac99 | 2018-01-27 14:06:21 +0900 | [diff] [blame] | 96 | PyDoc_STRVAR_shared(_Py_isascii__doc__, |
| 97 | "B.isascii() -> bool\n\ |
| 98 | \n\ |
| 99 | Return True if B is empty or all characters in B are ASCII,\n\ |
| 100 | False otherwise."); |
| 101 | |
INADA Naoki | bea5706 | 2018-01-28 09:59:12 +0900 | [diff] [blame] | 102 | // Optimization is copied from ascii_decode in unicodeobject.c |
Ma Lin | a0c603c | 2020-10-18 22:48:38 +0800 | [diff] [blame] | 103 | /* Mask to quickly check whether a C 'size_t' contains a |
INADA Naoki | bea5706 | 2018-01-28 09:59:12 +0900 | [diff] [blame] | 104 | non-ASCII, UTF8-encoded char. */ |
Ma Lin | a0c603c | 2020-10-18 22:48:38 +0800 | [diff] [blame] | 105 | #if (SIZEOF_SIZE_T == 8) |
| 106 | # define ASCII_CHAR_MASK 0x8080808080808080ULL |
| 107 | #elif (SIZEOF_SIZE_T == 4) |
| 108 | # define ASCII_CHAR_MASK 0x80808080U |
INADA Naoki | bea5706 | 2018-01-28 09:59:12 +0900 | [diff] [blame] | 109 | #else |
Ma Lin | a0c603c | 2020-10-18 22:48:38 +0800 | [diff] [blame] | 110 | # error C 'size_t' size should be either 4 or 8! |
INADA Naoki | bea5706 | 2018-01-28 09:59:12 +0900 | [diff] [blame] | 111 | #endif |
| 112 | |
INADA Naoki | a49ac99 | 2018-01-27 14:06:21 +0900 | [diff] [blame] | 113 | PyObject* |
| 114 | _Py_bytes_isascii(const char *cptr, Py_ssize_t len) |
| 115 | { |
INADA Naoki | bea5706 | 2018-01-28 09:59:12 +0900 | [diff] [blame] | 116 | const char *p = cptr; |
| 117 | const char *end = p + len; |
Ma Lin | a0c603c | 2020-10-18 22:48:38 +0800 | [diff] [blame] | 118 | const char *aligned_end = (const char *) _Py_ALIGN_DOWN(end, SIZEOF_SIZE_T); |
INADA Naoki | bea5706 | 2018-01-28 09:59:12 +0900 | [diff] [blame] | 119 | |
| 120 | while (p < end) { |
| 121 | /* Fast path, see in STRINGLIB(utf8_decode) in stringlib/codecs.h |
| 122 | for an explanation. */ |
Ma Lin | a0c603c | 2020-10-18 22:48:38 +0800 | [diff] [blame] | 123 | if (_Py_IS_ALIGNED(p, SIZEOF_SIZE_T)) { |
INADA Naoki | bea5706 | 2018-01-28 09:59:12 +0900 | [diff] [blame] | 124 | /* Help allocation */ |
| 125 | const char *_p = p; |
| 126 | while (_p < aligned_end) { |
Ma Lin | a0c603c | 2020-10-18 22:48:38 +0800 | [diff] [blame] | 127 | size_t value = *(const size_t *) _p; |
INADA Naoki | bea5706 | 2018-01-28 09:59:12 +0900 | [diff] [blame] | 128 | if (value & ASCII_CHAR_MASK) { |
| 129 | Py_RETURN_FALSE; |
| 130 | } |
Ma Lin | a0c603c | 2020-10-18 22:48:38 +0800 | [diff] [blame] | 131 | _p += SIZEOF_SIZE_T; |
INADA Naoki | bea5706 | 2018-01-28 09:59:12 +0900 | [diff] [blame] | 132 | } |
| 133 | p = _p; |
| 134 | if (_p == end) |
| 135 | break; |
| 136 | } |
| 137 | if ((unsigned char)*p & 0x80) { |
INADA Naoki | a49ac99 | 2018-01-27 14:06:21 +0900 | [diff] [blame] | 138 | Py_RETURN_FALSE; |
| 139 | } |
INADA Naoki | bea5706 | 2018-01-28 09:59:12 +0900 | [diff] [blame] | 140 | p++; |
INADA Naoki | a49ac99 | 2018-01-27 14:06:21 +0900 | [diff] [blame] | 141 | } |
| 142 | Py_RETURN_TRUE; |
| 143 | } |
| 144 | |
INADA Naoki | bea5706 | 2018-01-28 09:59:12 +0900 | [diff] [blame] | 145 | #undef ASCII_CHAR_MASK |
| 146 | |
INADA Naoki | a49ac99 | 2018-01-27 14:06:21 +0900 | [diff] [blame] | 147 | |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 148 | PyDoc_STRVAR_shared(_Py_isdigit__doc__, |
| 149 | "B.isdigit() -> bool\n\ |
| 150 | \n\ |
| 151 | Return True if all characters in B are digits\n\ |
| 152 | and there is at least one character in B, False otherwise."); |
| 153 | |
| 154 | PyObject* |
| 155 | _Py_bytes_isdigit(const char *cptr, Py_ssize_t len) |
| 156 | { |
Antoine Pitrou | 9ed5f27 | 2013-08-13 20:18:52 +0200 | [diff] [blame] | 157 | const unsigned char *p |
Andy Lester | e6be9b5 | 2020-02-11 20:28:35 -0600 | [diff] [blame] | 158 | = (const unsigned char *) cptr; |
Antoine Pitrou | 9ed5f27 | 2013-08-13 20:18:52 +0200 | [diff] [blame] | 159 | const unsigned char *e; |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 160 | |
| 161 | /* Shortcut for single character strings */ |
Eric Smith | 6dc46f5 | 2009-04-27 20:39:49 +0000 | [diff] [blame] | 162 | if (len == 1 && Py_ISDIGIT(*p)) |
Antoine Pitrou | f95a1b3 | 2010-05-09 15:52:27 +0000 | [diff] [blame] | 163 | Py_RETURN_TRUE; |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 164 | |
| 165 | /* Special case for empty strings */ |
| 166 | if (len == 0) |
Antoine Pitrou | f95a1b3 | 2010-05-09 15:52:27 +0000 | [diff] [blame] | 167 | Py_RETURN_FALSE; |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 168 | |
| 169 | e = p + len; |
| 170 | for (; p < e; p++) { |
Antoine Pitrou | f95a1b3 | 2010-05-09 15:52:27 +0000 | [diff] [blame] | 171 | if (!Py_ISDIGIT(*p)) |
| 172 | Py_RETURN_FALSE; |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 173 | } |
| 174 | Py_RETURN_TRUE; |
| 175 | } |
| 176 | |
| 177 | |
| 178 | PyDoc_STRVAR_shared(_Py_islower__doc__, |
| 179 | "B.islower() -> bool\n\ |
| 180 | \n\ |
| 181 | Return True if all cased characters in B are lowercase and there is\n\ |
| 182 | at least one cased character in B, False otherwise."); |
| 183 | |
| 184 | PyObject* |
| 185 | _Py_bytes_islower(const char *cptr, Py_ssize_t len) |
| 186 | { |
Antoine Pitrou | 9ed5f27 | 2013-08-13 20:18:52 +0200 | [diff] [blame] | 187 | const unsigned char *p |
Andy Lester | e6be9b5 | 2020-02-11 20:28:35 -0600 | [diff] [blame] | 188 | = (const unsigned char *) cptr; |
Antoine Pitrou | 9ed5f27 | 2013-08-13 20:18:52 +0200 | [diff] [blame] | 189 | const unsigned char *e; |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 190 | int cased; |
| 191 | |
| 192 | /* Shortcut for single character strings */ |
| 193 | if (len == 1) |
Antoine Pitrou | f95a1b3 | 2010-05-09 15:52:27 +0000 | [diff] [blame] | 194 | return PyBool_FromLong(Py_ISLOWER(*p)); |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 195 | |
| 196 | /* Special case for empty strings */ |
| 197 | if (len == 0) |
Antoine Pitrou | f95a1b3 | 2010-05-09 15:52:27 +0000 | [diff] [blame] | 198 | Py_RETURN_FALSE; |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 199 | |
| 200 | e = p + len; |
| 201 | cased = 0; |
| 202 | for (; p < e; p++) { |
Antoine Pitrou | f95a1b3 | 2010-05-09 15:52:27 +0000 | [diff] [blame] | 203 | if (Py_ISUPPER(*p)) |
| 204 | Py_RETURN_FALSE; |
| 205 | else if (!cased && Py_ISLOWER(*p)) |
| 206 | cased = 1; |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 207 | } |
| 208 | return PyBool_FromLong(cased); |
| 209 | } |
| 210 | |
| 211 | |
| 212 | PyDoc_STRVAR_shared(_Py_isupper__doc__, |
| 213 | "B.isupper() -> bool\n\ |
| 214 | \n\ |
| 215 | Return True if all cased characters in B are uppercase and there is\n\ |
| 216 | at least one cased character in B, False otherwise."); |
| 217 | |
| 218 | PyObject* |
| 219 | _Py_bytes_isupper(const char *cptr, Py_ssize_t len) |
| 220 | { |
Antoine Pitrou | 9ed5f27 | 2013-08-13 20:18:52 +0200 | [diff] [blame] | 221 | const unsigned char *p |
Andy Lester | e6be9b5 | 2020-02-11 20:28:35 -0600 | [diff] [blame] | 222 | = (const unsigned char *) cptr; |
Antoine Pitrou | 9ed5f27 | 2013-08-13 20:18:52 +0200 | [diff] [blame] | 223 | const unsigned char *e; |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 224 | int cased; |
| 225 | |
| 226 | /* Shortcut for single character strings */ |
| 227 | if (len == 1) |
Antoine Pitrou | f95a1b3 | 2010-05-09 15:52:27 +0000 | [diff] [blame] | 228 | return PyBool_FromLong(Py_ISUPPER(*p)); |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 229 | |
| 230 | /* Special case for empty strings */ |
| 231 | if (len == 0) |
Antoine Pitrou | f95a1b3 | 2010-05-09 15:52:27 +0000 | [diff] [blame] | 232 | Py_RETURN_FALSE; |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 233 | |
| 234 | e = p + len; |
| 235 | cased = 0; |
| 236 | for (; p < e; p++) { |
Antoine Pitrou | f95a1b3 | 2010-05-09 15:52:27 +0000 | [diff] [blame] | 237 | if (Py_ISLOWER(*p)) |
| 238 | Py_RETURN_FALSE; |
| 239 | else if (!cased && Py_ISUPPER(*p)) |
| 240 | cased = 1; |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 241 | } |
| 242 | return PyBool_FromLong(cased); |
| 243 | } |
| 244 | |
| 245 | |
| 246 | PyDoc_STRVAR_shared(_Py_istitle__doc__, |
| 247 | "B.istitle() -> bool\n\ |
| 248 | \n\ |
| 249 | Return True if B is a titlecased string and there is at least one\n\ |
| 250 | character in B, i.e. uppercase characters may only follow uncased\n\ |
| 251 | characters and lowercase characters only cased ones. Return False\n\ |
| 252 | otherwise."); |
| 253 | |
| 254 | PyObject* |
| 255 | _Py_bytes_istitle(const char *cptr, Py_ssize_t len) |
| 256 | { |
Antoine Pitrou | 9ed5f27 | 2013-08-13 20:18:52 +0200 | [diff] [blame] | 257 | const unsigned char *p |
Andy Lester | e6be9b5 | 2020-02-11 20:28:35 -0600 | [diff] [blame] | 258 | = (const unsigned char *) cptr; |
Antoine Pitrou | 9ed5f27 | 2013-08-13 20:18:52 +0200 | [diff] [blame] | 259 | const unsigned char *e; |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 260 | int cased, previous_is_cased; |
| 261 | |
| 262 | /* Shortcut for single character strings */ |
| 263 | if (len == 1) |
Antoine Pitrou | f95a1b3 | 2010-05-09 15:52:27 +0000 | [diff] [blame] | 264 | return PyBool_FromLong(Py_ISUPPER(*p)); |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 265 | |
| 266 | /* Special case for empty strings */ |
| 267 | if (len == 0) |
Antoine Pitrou | f95a1b3 | 2010-05-09 15:52:27 +0000 | [diff] [blame] | 268 | Py_RETURN_FALSE; |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 269 | |
| 270 | e = p + len; |
| 271 | cased = 0; |
| 272 | previous_is_cased = 0; |
| 273 | for (; p < e; p++) { |
Antoine Pitrou | 9ed5f27 | 2013-08-13 20:18:52 +0200 | [diff] [blame] | 274 | const unsigned char ch = *p; |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 275 | |
Antoine Pitrou | f95a1b3 | 2010-05-09 15:52:27 +0000 | [diff] [blame] | 276 | if (Py_ISUPPER(ch)) { |
| 277 | if (previous_is_cased) |
| 278 | Py_RETURN_FALSE; |
| 279 | previous_is_cased = 1; |
| 280 | cased = 1; |
| 281 | } |
| 282 | else if (Py_ISLOWER(ch)) { |
| 283 | if (!previous_is_cased) |
| 284 | Py_RETURN_FALSE; |
| 285 | previous_is_cased = 1; |
| 286 | cased = 1; |
| 287 | } |
| 288 | else |
| 289 | previous_is_cased = 0; |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 290 | } |
| 291 | return PyBool_FromLong(cased); |
| 292 | } |
| 293 | |
| 294 | |
| 295 | PyDoc_STRVAR_shared(_Py_lower__doc__, |
| 296 | "B.lower() -> copy of B\n\ |
| 297 | \n\ |
| 298 | Return a copy of B with all ASCII characters converted to lowercase."); |
| 299 | |
| 300 | void |
| 301 | _Py_bytes_lower(char *result, const char *cptr, Py_ssize_t len) |
| 302 | { |
Antoine Pitrou | 9b49192 | 2010-08-15 17:38:46 +0000 | [diff] [blame] | 303 | Py_ssize_t i; |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 304 | |
Antoine Pitrou | 9b49192 | 2010-08-15 17:38:46 +0000 | [diff] [blame] | 305 | for (i = 0; i < len; i++) { |
Antoine Pitrou | 94f6fa6 | 2012-01-08 16:22:46 +0100 | [diff] [blame] | 306 | result[i] = Py_TOLOWER((unsigned char) cptr[i]); |
Antoine Pitrou | 9b49192 | 2010-08-15 17:38:46 +0000 | [diff] [blame] | 307 | } |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 308 | } |
| 309 | |
| 310 | |
| 311 | PyDoc_STRVAR_shared(_Py_upper__doc__, |
| 312 | "B.upper() -> copy of B\n\ |
| 313 | \n\ |
| 314 | Return a copy of B with all ASCII characters converted to uppercase."); |
| 315 | |
| 316 | void |
| 317 | _Py_bytes_upper(char *result, const char *cptr, Py_ssize_t len) |
| 318 | { |
Antoine Pitrou | 9b49192 | 2010-08-15 17:38:46 +0000 | [diff] [blame] | 319 | Py_ssize_t i; |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 320 | |
Antoine Pitrou | 9b49192 | 2010-08-15 17:38:46 +0000 | [diff] [blame] | 321 | for (i = 0; i < len; i++) { |
Antoine Pitrou | 94f6fa6 | 2012-01-08 16:22:46 +0100 | [diff] [blame] | 322 | result[i] = Py_TOUPPER((unsigned char) cptr[i]); |
Antoine Pitrou | 9b49192 | 2010-08-15 17:38:46 +0000 | [diff] [blame] | 323 | } |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 324 | } |
| 325 | |
| 326 | |
| 327 | PyDoc_STRVAR_shared(_Py_title__doc__, |
| 328 | "B.title() -> copy of B\n\ |
| 329 | \n\ |
| 330 | Return a titlecased version of B, i.e. ASCII words start with uppercase\n\ |
| 331 | characters, all remaining cased characters have lowercase."); |
| 332 | |
| 333 | void |
Serhiy Storchaka | ef1585e | 2015-12-25 20:01:53 +0200 | [diff] [blame] | 334 | _Py_bytes_title(char *result, const char *s, Py_ssize_t len) |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 335 | { |
Antoine Pitrou | 9b49192 | 2010-08-15 17:38:46 +0000 | [diff] [blame] | 336 | Py_ssize_t i; |
| 337 | int previous_is_cased = 0; |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 338 | |
Antoine Pitrou | 9b49192 | 2010-08-15 17:38:46 +0000 | [diff] [blame] | 339 | for (i = 0; i < len; i++) { |
| 340 | int c = Py_CHARMASK(*s++); |
| 341 | if (Py_ISLOWER(c)) { |
| 342 | if (!previous_is_cased) |
| 343 | c = Py_TOUPPER(c); |
| 344 | previous_is_cased = 1; |
| 345 | } else if (Py_ISUPPER(c)) { |
| 346 | if (previous_is_cased) |
| 347 | c = Py_TOLOWER(c); |
| 348 | previous_is_cased = 1; |
| 349 | } else |
| 350 | previous_is_cased = 0; |
| 351 | *result++ = c; |
| 352 | } |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 353 | } |
| 354 | |
| 355 | |
| 356 | PyDoc_STRVAR_shared(_Py_capitalize__doc__, |
| 357 | "B.capitalize() -> copy of B\n\ |
| 358 | \n\ |
Senthil Kumaran | e51ee8a | 2010-07-05 12:00:56 +0000 | [diff] [blame] | 359 | Return a copy of B with only its first character capitalized (ASCII)\n\ |
| 360 | and the rest lower-cased."); |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 361 | |
| 362 | void |
Serhiy Storchaka | ef1585e | 2015-12-25 20:01:53 +0200 | [diff] [blame] | 363 | _Py_bytes_capitalize(char *result, const char *s, Py_ssize_t len) |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 364 | { |
Sergey Fedoseev | 593bb30 | 2018-09-07 09:54:49 +0500 | [diff] [blame] | 365 | if (len > 0) { |
| 366 | *result = Py_TOUPPER(*s); |
| 367 | _Py_bytes_lower(result + 1, s + 1, len - 1); |
Antoine Pitrou | 9b49192 | 2010-08-15 17:38:46 +0000 | [diff] [blame] | 368 | } |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 369 | } |
| 370 | |
| 371 | |
| 372 | PyDoc_STRVAR_shared(_Py_swapcase__doc__, |
| 373 | "B.swapcase() -> copy of B\n\ |
| 374 | \n\ |
| 375 | Return a copy of B with uppercase ASCII characters converted\n\ |
| 376 | to lowercase ASCII and vice versa."); |
| 377 | |
| 378 | void |
Serhiy Storchaka | ef1585e | 2015-12-25 20:01:53 +0200 | [diff] [blame] | 379 | _Py_bytes_swapcase(char *result, const char *s, Py_ssize_t len) |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 380 | { |
Antoine Pitrou | 9b49192 | 2010-08-15 17:38:46 +0000 | [diff] [blame] | 381 | Py_ssize_t i; |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 382 | |
Antoine Pitrou | 9b49192 | 2010-08-15 17:38:46 +0000 | [diff] [blame] | 383 | for (i = 0; i < len; i++) { |
| 384 | int c = Py_CHARMASK(*s++); |
| 385 | if (Py_ISLOWER(c)) { |
| 386 | *result = Py_TOUPPER(c); |
Antoine Pitrou | f95a1b3 | 2010-05-09 15:52:27 +0000 | [diff] [blame] | 387 | } |
Antoine Pitrou | 9b49192 | 2010-08-15 17:38:46 +0000 | [diff] [blame] | 388 | else if (Py_ISUPPER(c)) { |
| 389 | *result = Py_TOLOWER(c); |
| 390 | } |
| 391 | else |
| 392 | *result = c; |
| 393 | result++; |
| 394 | } |
Gregory P. Smith | 60d241f | 2007-10-16 06:31:30 +0000 | [diff] [blame] | 395 | } |
| 396 | |
Georg Brandl | abc3877 | 2009-04-12 15:51:51 +0000 | [diff] [blame] | 397 | |
| 398 | PyDoc_STRVAR_shared(_Py_maketrans__doc__, |
| 399 | "B.maketrans(frm, to) -> translation table\n\ |
| 400 | \n\ |
Senthil Kumaran | 84e3ccc | 2011-06-27 09:06:45 -0700 | [diff] [blame] | 401 | Return a translation table (a bytes object of length 256) suitable\n\ |
| 402 | for use in the bytes or bytearray translate method where each byte\n\ |
| 403 | in frm is mapped to the byte at the same position in to.\n\ |
| 404 | The bytes objects frm and to must be of the same length."); |
Georg Brandl | abc3877 | 2009-04-12 15:51:51 +0000 | [diff] [blame] | 405 | |
Georg Brandl | abc3877 | 2009-04-12 15:51:51 +0000 | [diff] [blame] | 406 | PyObject * |
Serhiy Storchaka | 3dd3e26 | 2015-02-03 01:25:42 +0200 | [diff] [blame] | 407 | _Py_bytes_maketrans(Py_buffer *frm, Py_buffer *to) |
Georg Brandl | abc3877 | 2009-04-12 15:51:51 +0000 | [diff] [blame] | 408 | { |
Martin v. Löwis | 7252a6e | 2014-07-27 16:25:09 +0200 | [diff] [blame] | 409 | PyObject *res = NULL; |
Antoine Pitrou | 9b49192 | 2010-08-15 17:38:46 +0000 | [diff] [blame] | 410 | Py_ssize_t i; |
| 411 | char *p; |
Georg Brandl | abc3877 | 2009-04-12 15:51:51 +0000 | [diff] [blame] | 412 | |
Serhiy Storchaka | 3dd3e26 | 2015-02-03 01:25:42 +0200 | [diff] [blame] | 413 | if (frm->len != to->len) { |
Antoine Pitrou | 9b49192 | 2010-08-15 17:38:46 +0000 | [diff] [blame] | 414 | PyErr_Format(PyExc_ValueError, |
| 415 | "maketrans arguments must have same length"); |
Serhiy Storchaka | 3dd3e26 | 2015-02-03 01:25:42 +0200 | [diff] [blame] | 416 | return NULL; |
Antoine Pitrou | 9b49192 | 2010-08-15 17:38:46 +0000 | [diff] [blame] | 417 | } |
| 418 | res = PyBytes_FromStringAndSize(NULL, 256); |
Serhiy Storchaka | 3dd3e26 | 2015-02-03 01:25:42 +0200 | [diff] [blame] | 419 | if (!res) |
| 420 | return NULL; |
Antoine Pitrou | 9b49192 | 2010-08-15 17:38:46 +0000 | [diff] [blame] | 421 | p = PyBytes_AS_STRING(res); |
| 422 | for (i = 0; i < 256; i++) |
Antoine Pitrou | 47019e5 | 2010-08-15 17:41:31 +0000 | [diff] [blame] | 423 | p[i] = (char) i; |
Serhiy Storchaka | 3dd3e26 | 2015-02-03 01:25:42 +0200 | [diff] [blame] | 424 | for (i = 0; i < frm->len; i++) { |
| 425 | p[((unsigned char *)frm->buf)[i]] = ((char *)to->buf)[i]; |
Antoine Pitrou | 9b49192 | 2010-08-15 17:38:46 +0000 | [diff] [blame] | 426 | } |
Georg Brandl | abc3877 | 2009-04-12 15:51:51 +0000 | [diff] [blame] | 427 | |
Antoine Pitrou | 9b49192 | 2010-08-15 17:38:46 +0000 | [diff] [blame] | 428 | return res; |
Georg Brandl | abc3877 | 2009-04-12 15:51:51 +0000 | [diff] [blame] | 429 | } |
Serhiy Storchaka | dd40fc3 | 2016-05-04 22:23:26 +0300 | [diff] [blame] | 430 | |
| 431 | #define FASTSEARCH fastsearch |
| 432 | #define STRINGLIB(F) stringlib_##F |
| 433 | #define STRINGLIB_CHAR char |
| 434 | #define STRINGLIB_SIZEOF_CHAR 1 |
| 435 | |
| 436 | #include "stringlib/fastsearch.h" |
| 437 | #include "stringlib/count.h" |
| 438 | #include "stringlib/find.h" |
| 439 | |
| 440 | /* |
Oren Milman | 0042510 | 2017-03-13 00:37:05 +0200 | [diff] [blame] | 441 | Wraps stringlib_parse_args_finds() and additionally checks the first |
| 442 | argument type. |
Serhiy Storchaka | dd40fc3 | 2016-05-04 22:23:26 +0300 | [diff] [blame] | 443 | |
Oren Milman | 0042510 | 2017-03-13 00:37:05 +0200 | [diff] [blame] | 444 | In case the first argument is a bytes-like object, sets it to subobj, |
| 445 | and doesn't touch the byte parameter. |
| 446 | In case it is an integer in range(0, 256), writes the integer value |
| 447 | to byte, and sets subobj to NULL. |
| 448 | |
| 449 | The other parameters are similar to those of |
Serhiy Storchaka | dd40fc3 | 2016-05-04 22:23:26 +0300 | [diff] [blame] | 450 | stringlib_parse_args_finds(). |
| 451 | */ |
| 452 | |
| 453 | Py_LOCAL_INLINE(int) |
| 454 | parse_args_finds_byte(const char *function_name, PyObject *args, |
| 455 | PyObject **subobj, char *byte, |
| 456 | Py_ssize_t *start, Py_ssize_t *end) |
| 457 | { |
| 458 | PyObject *tmp_subobj; |
| 459 | Py_ssize_t ival; |
Serhiy Storchaka | dd40fc3 | 2016-05-04 22:23:26 +0300 | [diff] [blame] | 460 | |
| 461 | if(!stringlib_parse_args_finds(function_name, args, &tmp_subobj, |
| 462 | start, end)) |
| 463 | return 0; |
| 464 | |
Oren Milman | 0042510 | 2017-03-13 00:37:05 +0200 | [diff] [blame] | 465 | if (PyObject_CheckBuffer(tmp_subobj)) { |
Serhiy Storchaka | dd40fc3 | 2016-05-04 22:23:26 +0300 | [diff] [blame] | 466 | *subobj = tmp_subobj; |
| 467 | return 1; |
| 468 | } |
| 469 | |
Victor Stinner | a15e260 | 2020-04-08 02:01:56 +0200 | [diff] [blame] | 470 | if (!_PyIndex_Check(tmp_subobj)) { |
Oren Milman | 0042510 | 2017-03-13 00:37:05 +0200 | [diff] [blame] | 471 | PyErr_Format(PyExc_TypeError, |
| 472 | "argument should be integer or bytes-like object, " |
| 473 | "not '%.200s'", |
| 474 | Py_TYPE(tmp_subobj)->tp_name); |
| 475 | return 0; |
Serhiy Storchaka | dd40fc3 | 2016-05-04 22:23:26 +0300 | [diff] [blame] | 476 | } |
| 477 | |
Oren Milman | 0042510 | 2017-03-13 00:37:05 +0200 | [diff] [blame] | 478 | ival = PyNumber_AsSsize_t(tmp_subobj, NULL); |
| 479 | if (ival == -1 && PyErr_Occurred()) { |
| 480 | return 0; |
| 481 | } |
Serhiy Storchaka | dd40fc3 | 2016-05-04 22:23:26 +0300 | [diff] [blame] | 482 | if (ival < 0 || ival > 255) { |
| 483 | PyErr_SetString(PyExc_ValueError, "byte must be in range(0, 256)"); |
| 484 | return 0; |
| 485 | } |
| 486 | |
| 487 | *subobj = NULL; |
| 488 | *byte = (char)ival; |
| 489 | return 1; |
| 490 | } |
| 491 | |
| 492 | /* helper macro to fixup start/end slice values */ |
| 493 | #define ADJUST_INDICES(start, end, len) \ |
| 494 | if (end > len) \ |
| 495 | end = len; \ |
| 496 | else if (end < 0) { \ |
| 497 | end += len; \ |
| 498 | if (end < 0) \ |
| 499 | end = 0; \ |
| 500 | } \ |
| 501 | if (start < 0) { \ |
| 502 | start += len; \ |
| 503 | if (start < 0) \ |
| 504 | start = 0; \ |
| 505 | } |
| 506 | |
| 507 | Py_LOCAL_INLINE(Py_ssize_t) |
| 508 | find_internal(const char *str, Py_ssize_t len, |
| 509 | const char *function_name, PyObject *args, int dir) |
| 510 | { |
| 511 | PyObject *subobj; |
| 512 | char byte; |
| 513 | Py_buffer subbuf; |
| 514 | const char *sub; |
| 515 | Py_ssize_t sub_len; |
| 516 | Py_ssize_t start = 0, end = PY_SSIZE_T_MAX; |
| 517 | Py_ssize_t res; |
| 518 | |
| 519 | if (!parse_args_finds_byte(function_name, args, |
| 520 | &subobj, &byte, &start, &end)) |
| 521 | return -2; |
| 522 | |
| 523 | if (subobj) { |
| 524 | if (PyObject_GetBuffer(subobj, &subbuf, PyBUF_SIMPLE) != 0) |
| 525 | return -2; |
| 526 | |
| 527 | sub = subbuf.buf; |
| 528 | sub_len = subbuf.len; |
| 529 | } |
| 530 | else { |
| 531 | sub = &byte; |
| 532 | sub_len = 1; |
| 533 | } |
| 534 | |
| 535 | ADJUST_INDICES(start, end, len); |
| 536 | if (end - start < sub_len) |
| 537 | res = -1; |
| 538 | else if (sub_len == 1) { |
| 539 | if (dir > 0) |
| 540 | res = stringlib_find_char( |
| 541 | str + start, end - start, |
| 542 | *sub); |
| 543 | else |
| 544 | res = stringlib_rfind_char( |
| 545 | str + start, end - start, |
| 546 | *sub); |
| 547 | if (res >= 0) |
| 548 | res += start; |
| 549 | } |
| 550 | else { |
| 551 | if (dir > 0) |
| 552 | res = stringlib_find_slice( |
| 553 | str, len, |
| 554 | sub, sub_len, start, end); |
| 555 | else |
| 556 | res = stringlib_rfind_slice( |
| 557 | str, len, |
| 558 | sub, sub_len, start, end); |
| 559 | } |
| 560 | |
| 561 | if (subobj) |
| 562 | PyBuffer_Release(&subbuf); |
| 563 | |
| 564 | return res; |
| 565 | } |
| 566 | |
| 567 | PyDoc_STRVAR_shared(_Py_find__doc__, |
| 568 | "B.find(sub[, start[, end]]) -> int\n\ |
| 569 | \n\ |
| 570 | Return the lowest index in B where subsection sub is found,\n\ |
| 571 | such that sub is contained within B[start,end]. Optional\n\ |
| 572 | arguments start and end are interpreted as in slice notation.\n\ |
| 573 | \n\ |
| 574 | Return -1 on failure."); |
| 575 | |
| 576 | PyObject * |
| 577 | _Py_bytes_find(const char *str, Py_ssize_t len, PyObject *args) |
| 578 | { |
| 579 | Py_ssize_t result = find_internal(str, len, "find", args, +1); |
| 580 | if (result == -2) |
| 581 | return NULL; |
| 582 | return PyLong_FromSsize_t(result); |
| 583 | } |
| 584 | |
| 585 | PyDoc_STRVAR_shared(_Py_index__doc__, |
| 586 | "B.index(sub[, start[, end]]) -> int\n\ |
| 587 | \n\ |
Lisa Roach | 43ba886 | 2017-04-04 22:36:22 -0700 | [diff] [blame] | 588 | Return the lowest index in B where subsection sub is found,\n\ |
| 589 | such that sub is contained within B[start,end]. Optional\n\ |
| 590 | arguments start and end are interpreted as in slice notation.\n\ |
| 591 | \n\ |
| 592 | Raises ValueError when the subsection is not found."); |
Serhiy Storchaka | dd40fc3 | 2016-05-04 22:23:26 +0300 | [diff] [blame] | 593 | |
| 594 | PyObject * |
| 595 | _Py_bytes_index(const char *str, Py_ssize_t len, PyObject *args) |
| 596 | { |
| 597 | Py_ssize_t result = find_internal(str, len, "index", args, +1); |
| 598 | if (result == -2) |
| 599 | return NULL; |
| 600 | if (result == -1) { |
| 601 | PyErr_SetString(PyExc_ValueError, |
| 602 | "subsection not found"); |
| 603 | return NULL; |
| 604 | } |
| 605 | return PyLong_FromSsize_t(result); |
| 606 | } |
| 607 | |
| 608 | PyDoc_STRVAR_shared(_Py_rfind__doc__, |
| 609 | "B.rfind(sub[, start[, end]]) -> int\n\ |
| 610 | \n\ |
| 611 | Return the highest index in B where subsection sub is found,\n\ |
| 612 | such that sub is contained within B[start,end]. Optional\n\ |
| 613 | arguments start and end are interpreted as in slice notation.\n\ |
| 614 | \n\ |
| 615 | Return -1 on failure."); |
| 616 | |
| 617 | PyObject * |
| 618 | _Py_bytes_rfind(const char *str, Py_ssize_t len, PyObject *args) |
| 619 | { |
| 620 | Py_ssize_t result = find_internal(str, len, "rfind", args, -1); |
| 621 | if (result == -2) |
| 622 | return NULL; |
| 623 | return PyLong_FromSsize_t(result); |
| 624 | } |
| 625 | |
| 626 | PyDoc_STRVAR_shared(_Py_rindex__doc__, |
| 627 | "B.rindex(sub[, start[, end]]) -> int\n\ |
| 628 | \n\ |
Lisa Roach | 43ba886 | 2017-04-04 22:36:22 -0700 | [diff] [blame] | 629 | Return the highest index in B where subsection sub is found,\n\ |
| 630 | such that sub is contained within B[start,end]. Optional\n\ |
| 631 | arguments start and end are interpreted as in slice notation.\n\ |
| 632 | \n\ |
| 633 | Raise ValueError when the subsection is not found."); |
Serhiy Storchaka | dd40fc3 | 2016-05-04 22:23:26 +0300 | [diff] [blame] | 634 | |
| 635 | PyObject * |
| 636 | _Py_bytes_rindex(const char *str, Py_ssize_t len, PyObject *args) |
| 637 | { |
| 638 | Py_ssize_t result = find_internal(str, len, "rindex", args, -1); |
| 639 | if (result == -2) |
| 640 | return NULL; |
| 641 | if (result == -1) { |
| 642 | PyErr_SetString(PyExc_ValueError, |
| 643 | "subsection not found"); |
| 644 | return NULL; |
| 645 | } |
| 646 | return PyLong_FromSsize_t(result); |
| 647 | } |
| 648 | |
| 649 | PyDoc_STRVAR_shared(_Py_count__doc__, |
| 650 | "B.count(sub[, start[, end]]) -> int\n\ |
| 651 | \n\ |
| 652 | Return the number of non-overlapping occurrences of subsection sub in\n\ |
| 653 | bytes B[start:end]. Optional arguments start and end are interpreted\n\ |
| 654 | as in slice notation."); |
| 655 | |
| 656 | PyObject * |
| 657 | _Py_bytes_count(const char *str, Py_ssize_t len, PyObject *args) |
| 658 | { |
| 659 | PyObject *sub_obj; |
| 660 | const char *sub; |
| 661 | Py_ssize_t sub_len; |
| 662 | char byte; |
| 663 | Py_ssize_t start = 0, end = PY_SSIZE_T_MAX; |
| 664 | |
| 665 | Py_buffer vsub; |
| 666 | PyObject *count_obj; |
| 667 | |
| 668 | if (!parse_args_finds_byte("count", args, |
| 669 | &sub_obj, &byte, &start, &end)) |
| 670 | return NULL; |
| 671 | |
| 672 | if (sub_obj) { |
| 673 | if (PyObject_GetBuffer(sub_obj, &vsub, PyBUF_SIMPLE) != 0) |
| 674 | return NULL; |
| 675 | |
| 676 | sub = vsub.buf; |
| 677 | sub_len = vsub.len; |
| 678 | } |
| 679 | else { |
| 680 | sub = &byte; |
| 681 | sub_len = 1; |
| 682 | } |
| 683 | |
| 684 | ADJUST_INDICES(start, end, len); |
| 685 | |
| 686 | count_obj = PyLong_FromSsize_t( |
| 687 | stringlib_count(str + start, end - start, sub, sub_len, PY_SSIZE_T_MAX) |
| 688 | ); |
| 689 | |
| 690 | if (sub_obj) |
| 691 | PyBuffer_Release(&vsub); |
| 692 | |
| 693 | return count_obj; |
| 694 | } |
| 695 | |
| 696 | int |
| 697 | _Py_bytes_contains(const char *str, Py_ssize_t len, PyObject *arg) |
| 698 | { |
Serhiy Storchaka | f9efb8b | 2016-07-10 12:37:30 +0300 | [diff] [blame] | 699 | Py_ssize_t ival = PyNumber_AsSsize_t(arg, NULL); |
Serhiy Storchaka | dd40fc3 | 2016-05-04 22:23:26 +0300 | [diff] [blame] | 700 | if (ival == -1 && PyErr_Occurred()) { |
| 701 | Py_buffer varg; |
| 702 | Py_ssize_t pos; |
| 703 | PyErr_Clear(); |
| 704 | if (PyObject_GetBuffer(arg, &varg, PyBUF_SIMPLE) != 0) |
| 705 | return -1; |
| 706 | pos = stringlib_find(str, len, |
| 707 | varg.buf, varg.len, 0); |
| 708 | PyBuffer_Release(&varg); |
| 709 | return pos >= 0; |
| 710 | } |
| 711 | if (ival < 0 || ival >= 256) { |
| 712 | PyErr_SetString(PyExc_ValueError, "byte must be in range(0, 256)"); |
| 713 | return -1; |
| 714 | } |
| 715 | |
| 716 | return memchr(str, (int) ival, len) != NULL; |
| 717 | } |
| 718 | |
| 719 | |
| 720 | /* Matches the end (direction >= 0) or start (direction < 0) of the buffer |
| 721 | * against substr, using the start and end arguments. Returns |
| 722 | * -1 on error, 0 if not found and 1 if found. |
| 723 | */ |
Benjamin Peterson | 621b430 | 2016-09-09 13:54:34 -0700 | [diff] [blame] | 724 | static int |
Serhiy Storchaka | dd40fc3 | 2016-05-04 22:23:26 +0300 | [diff] [blame] | 725 | tailmatch(const char *str, Py_ssize_t len, PyObject *substr, |
| 726 | Py_ssize_t start, Py_ssize_t end, int direction) |
| 727 | { |
| 728 | Py_buffer sub_view = {NULL, NULL}; |
| 729 | const char *sub; |
| 730 | Py_ssize_t slen; |
| 731 | |
| 732 | if (PyBytes_Check(substr)) { |
| 733 | sub = PyBytes_AS_STRING(substr); |
| 734 | slen = PyBytes_GET_SIZE(substr); |
| 735 | } |
| 736 | else { |
| 737 | if (PyObject_GetBuffer(substr, &sub_view, PyBUF_SIMPLE) != 0) |
| 738 | return -1; |
| 739 | sub = sub_view.buf; |
| 740 | slen = sub_view.len; |
| 741 | } |
| 742 | |
| 743 | ADJUST_INDICES(start, end, len); |
| 744 | |
| 745 | if (direction < 0) { |
| 746 | /* startswith */ |
Hai Shi | 24ddd9c | 2019-10-06 20:17:18 +0800 | [diff] [blame] | 747 | if (start > len - slen) |
Serhiy Storchaka | dd40fc3 | 2016-05-04 22:23:26 +0300 | [diff] [blame] | 748 | goto notfound; |
| 749 | } else { |
| 750 | /* endswith */ |
| 751 | if (end - start < slen || start > len) |
| 752 | goto notfound; |
| 753 | |
| 754 | if (end - slen > start) |
| 755 | start = end - slen; |
| 756 | } |
| 757 | if (end - start < slen) |
| 758 | goto notfound; |
| 759 | if (memcmp(str + start, sub, slen) != 0) |
| 760 | goto notfound; |
| 761 | |
| 762 | PyBuffer_Release(&sub_view); |
| 763 | return 1; |
| 764 | |
| 765 | notfound: |
| 766 | PyBuffer_Release(&sub_view); |
| 767 | return 0; |
| 768 | } |
| 769 | |
Benjamin Peterson | 621b430 | 2016-09-09 13:54:34 -0700 | [diff] [blame] | 770 | static PyObject * |
Serhiy Storchaka | dd40fc3 | 2016-05-04 22:23:26 +0300 | [diff] [blame] | 771 | _Py_bytes_tailmatch(const char *str, Py_ssize_t len, |
| 772 | const char *function_name, PyObject *args, |
| 773 | int direction) |
| 774 | { |
| 775 | Py_ssize_t start = 0; |
| 776 | Py_ssize_t end = PY_SSIZE_T_MAX; |
| 777 | PyObject *subobj; |
| 778 | int result; |
| 779 | |
| 780 | if (!stringlib_parse_args_finds(function_name, args, &subobj, &start, &end)) |
| 781 | return NULL; |
| 782 | if (PyTuple_Check(subobj)) { |
| 783 | Py_ssize_t i; |
| 784 | for (i = 0; i < PyTuple_GET_SIZE(subobj); i++) { |
| 785 | result = tailmatch(str, len, PyTuple_GET_ITEM(subobj, i), |
| 786 | start, end, direction); |
| 787 | if (result == -1) |
| 788 | return NULL; |
| 789 | else if (result) { |
| 790 | Py_RETURN_TRUE; |
| 791 | } |
| 792 | } |
| 793 | Py_RETURN_FALSE; |
| 794 | } |
| 795 | result = tailmatch(str, len, subobj, start, end, direction); |
| 796 | if (result == -1) { |
| 797 | if (PyErr_ExceptionMatches(PyExc_TypeError)) |
| 798 | PyErr_Format(PyExc_TypeError, |
| 799 | "%s first arg must be bytes or a tuple of bytes, " |
| 800 | "not %s", |
| 801 | function_name, Py_TYPE(subobj)->tp_name); |
| 802 | return NULL; |
| 803 | } |
| 804 | else |
| 805 | return PyBool_FromLong(result); |
| 806 | } |
| 807 | |
| 808 | PyDoc_STRVAR_shared(_Py_startswith__doc__, |
| 809 | "B.startswith(prefix[, start[, end]]) -> bool\n\ |
| 810 | \n\ |
| 811 | Return True if B starts with the specified prefix, False otherwise.\n\ |
| 812 | With optional start, test B beginning at that position.\n\ |
| 813 | With optional end, stop comparing B at that position.\n\ |
| 814 | prefix can also be a tuple of bytes to try."); |
| 815 | |
| 816 | PyObject * |
| 817 | _Py_bytes_startswith(const char *str, Py_ssize_t len, PyObject *args) |
| 818 | { |
| 819 | return _Py_bytes_tailmatch(str, len, "startswith", args, -1); |
| 820 | } |
| 821 | |
| 822 | PyDoc_STRVAR_shared(_Py_endswith__doc__, |
| 823 | "B.endswith(suffix[, start[, end]]) -> bool\n\ |
| 824 | \n\ |
| 825 | Return True if B ends with the specified suffix, False otherwise.\n\ |
| 826 | With optional start, test B beginning at that position.\n\ |
| 827 | With optional end, stop comparing B at that position.\n\ |
| 828 | suffix can also be a tuple of bytes to try."); |
| 829 | |
| 830 | PyObject * |
| 831 | _Py_bytes_endswith(const char *str, Py_ssize_t len, PyObject *args) |
| 832 | { |
| 833 | return _Py_bytes_tailmatch(str, len, "endswith", args, +1); |
| 834 | } |