| 1 | <?php |
| 2 | /* |
| 3 | * $Id$ |
| 4 | * |
| 5 | * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS |
| 6 | * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT |
| 7 | * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR |
| 8 | * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT |
| 9 | * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, |
| 10 | * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT |
| 11 | * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, |
| 12 | * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY |
| 13 | * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT |
| 14 | * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE |
| 15 | * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. |
| 16 | * |
| 17 | * This software consists of voluntary contributions made by many individuals |
| 18 | * and is licensed under the LGPL. For more information, see |
| 19 | * <http://www.phpdoctrine.org>. |
| 20 | */ |
| 21 | |
| 22 | /** |
| 23 | * Doctrine_Query_Tokenizer |
| 24 | * |
| 25 | * @package Doctrine |
| 26 | * @subpackage Query |
| 27 | * @license http://www.opensource.org/licenses/lgpl-license.php LGPL |
| 28 | * @link www.phpdoctrine.org |
| 29 | * @since 1.0 |
| 30 | * @version $Revision$ |
| 31 | * @todo Give the tokenizer state, make it better work together with Doctrine_Query and maybe |
| 32 | * take out commonly used string manipulation methods |
| 33 | * into a stateless StringUtil? class. This tokenizer should be concerned with tokenizing |
| 34 | * DQL strings. |
| 35 | */ |
| 36 | class Doctrine_Query_Tokenizer |
| 37 | { |
| 38 | |
| 39 | /** |
| 40 | * tokenizeQuery |
| 41 | * splits the given dql query into an array where keys |
| 42 | * represent different query part names and values are |
| 43 | * arrays splitted using sqlExplode method |
| 44 | * |
| 45 | * example: |
| 46 | * |
| 47 | * parameter: |
| 48 | * $query = "SELECT u.* FROM User u WHERE u.name LIKE ?" |
| 49 | * returns: |
| 50 | * array('select' => array('u.*'), |
| 51 | * 'from' => array('User', 'u'), |
| 52 | * 'where' => array('u.name', 'LIKE', '?')) |
| 53 | * |
| 54 | * @param string $query DQL query |
| 55 | * @throws Doctrine_Query_Exception if some generic parsing error occurs |
| 56 | * @return array an array containing the query string parts |
| 57 | */ |
| 58 | public function tokenizeQuery($query) |
| 59 | { |
| 60 | $parts = array(); |
| 61 | $tokens = $this->sqlExplode($query, ' '); |
| 62 | |
| 63 | foreach ($tokens as $index => $token) { |
| 64 | $token = trim($token); |
| 65 | switch (strtolower($token)) { |
| 66 | case 'delete': |
| 67 | case 'update': |
| 68 | case 'select': |
| 69 | case 'set': |
| 70 | case 'from': |
| 71 | case 'where': |
| 72 | case 'limit': |
| 73 | case 'offset': |
| 74 | case 'having': |
| 75 | $p = $token; |
| 76 | //$parts[$token] = array(); |
| 77 | $parts[$token] = ''; |
| 78 | break; |
| 79 | case 'order': |
| 80 | case 'group': |
| 81 | $i = ($index + 1); |
| 82 | if (isset($tokens[$i]) && strtolower($tokens[$i]) === 'by') { |
| 83 | $p = $token; |
| 84 | $parts[$token] = ''; |
| 85 | //$parts[$token] = array(); |
| 86 | } else { |
| 87 | $parts[$p] .= "$token "; |
| 88 | //$parts[$p][] = $token; |
| 89 | } |
| 90 | break; |
| 91 | case 'by': |
| 92 | continue; |
| 93 | default: |
| 94 | if ( ! isset($p)) { |
| 95 | throw new Doctrine_Query_Tokenizer_Exception( |
| 96 | "Couldn't tokenize query. Encountered invalid token: '$token'."); |
| 97 | } |
| 98 | |
| 99 | $parts[$p] .= "$token "; |
| 100 | //$parts[$p][] = $token; |
| 101 | } |
| 102 | } |
| 103 | return $parts; |
| 104 | } |
| 105 | |
| 106 | /** |
| 107 | * trims brackets |
| 108 | * |
| 109 | * @param string $str |
| 110 | * @param string $e1 the first bracket, usually '(' |
| 111 | * @param string $e2 the second bracket, usually ')' |
| 112 | */ |
| 113 | public function bracketTrim($str, $e1 = '(', $e2 = ')') |
| 114 | { |
| 115 | if (substr($str, 0, 1) === $e1 && substr($str, -1) === $e2) { |
| 116 | return substr($str, 1, -1); |
| 117 | } else { |
| 118 | return $str; |
| 119 | } |
| 120 | } |
| 121 | |
| 122 | /** |
| 123 | * bracketExplode |
| 124 | * |
| 125 | * example: |
| 126 | * |
| 127 | * parameters: |
| 128 | * $str = (age < 20 AND age > 18) AND email LIKE 'John@example.com' |
| 129 | * $d = ' AND ' |
| 130 | * $e1 = '(' |
| 131 | * $e2 = ')' |
| 132 | * |
| 133 | * would return an array: |
| 134 | * array("(age < 20 AND age > 18)", |
| 135 | * "email LIKE 'John@example.com'") |
| 136 | * |
| 137 | * @param string $str |
| 138 | * @param string $d the delimeter which explodes the string |
| 139 | * @param string $e1 the first bracket, usually '(' |
| 140 | * @param string $e2 the second bracket, usually ')' |
| 141 | * |
| 142 | */ |
| 143 | public function bracketExplode($str, $d = ' ', $e1 = '(', $e2 = ')') |
| 144 | { |
| 145 | if (is_array($d)) { |
| 146 | $a = preg_split('#('.implode('|', $d).')#', $str); |
| 147 | $d = stripslashes($d[0]); |
| 148 | } else { |
| 149 | $a = explode($d, $str); |
| 150 | } |
| 151 | |
| 152 | $i = 0; |
| 153 | $term = array(); |
| 154 | foreach($a as $key=>$val) { |
| 155 | if (empty($term[$i])) { |
| 156 | $term[$i] = trim($val); |
| 157 | $s1 = substr_count($term[$i], $e1); |
| 158 | $s2 = substr_count($term[$i], $e2); |
| 159 | |
| 160 | if ($s1 == $s2) { |
| 161 | $i++; |
| 162 | } |
| 163 | } else { |
| 164 | $term[$i] .= $d . trim($val); |
| 165 | $c1 = substr_count($term[$i], $e1); |
| 166 | $c2 = substr_count($term[$i], $e2); |
| 167 | |
| 168 | if ($c1 == $c2) { |
| 169 | $i++; |
| 170 | } |
| 171 | } |
| 172 | } |
| 173 | return $term; |
| 174 | } |
| 175 | |
| 176 | /** |
| 177 | * quoteExplode |
| 178 | * |
| 179 | * example: |
| 180 | * |
| 181 | * parameters: |
| 182 | * $str = email LIKE 'John@example.com' |
| 183 | * $d = ' LIKE ' |
| 184 | * |
| 185 | * would return an array: |
| 186 | * array("email", "LIKE", "'John@example.com'") |
| 187 | * |
| 188 | * @param string $str |
| 189 | * @param string $d the delimeter which explodes the string |
| 190 | */ |
| 191 | public function quoteExplode($str, $d = ' ') |
| 192 | { |
| 193 | if (is_array($d)) { |
| 194 | $a = preg_split('/('.implode('|', $d).')/', $str); |
| 195 | $d = stripslashes($d[0]); |
| 196 | } else { |
| 197 | $a = explode($d, $str); |
| 198 | } |
| 199 | |
| 200 | $i = 0; |
| 201 | $term = array(); |
| 202 | foreach ($a as $key => $val) { |
| 203 | if (empty($term[$i])) { |
| 204 | $term[$i] = trim($val); |
| 205 | |
| 206 | if ( ! (substr_count($term[$i], "'") & 1)) { |
| 207 | $i++; |
| 208 | } |
| 209 | } else { |
| 210 | $term[$i] .= $d . trim($val); |
| 211 | |
| 212 | if ( ! (substr_count($term[$i], "'") & 1)) { |
| 213 | $i++; |
| 214 | } |
| 215 | } |
| 216 | } |
| 217 | return $term; |
| 218 | } |
| 219 | |
| 220 | /** |
| 221 | * sqlExplode |
| 222 | * |
| 223 | * explodes a string into array using custom brackets and |
| 224 | * quote delimeters |
| 225 | * |
| 226 | * |
| 227 | * example: |
| 228 | * |
| 229 | * parameters: |
| 230 | * $str = "(age < 20 AND age > 18) AND name LIKE 'John Doe'" |
| 231 | * $d = ' ' |
| 232 | * $e1 = '(' |
| 233 | * $e2 = ')' |
| 234 | * |
| 235 | * would return an array: |
| 236 | * array('(age < 20 AND age > 18)', |
| 237 | * 'name', |
| 238 | * 'LIKE', |
| 239 | * 'John Doe') |
| 240 | * |
| 241 | * @param string $str |
| 242 | * @param string $d the delimeter which explodes the string |
| 243 | * @param string $e1 the first bracket, usually '(' |
| 244 | * @param string $e2 the second bracket, usually ')' |
| 245 | * |
| 246 | * @return array |
| 247 | */ |
| 248 | public function sqlExplode($str, $d = ' ', $e1 = '(', $e2 = ')') |
| 249 | { |
| 250 | if ($d == ' ') { |
| 251 | $d = array(' ', '\s'); |
| 252 | } |
| 253 | if (is_array($d)) { |
| 254 | $d = array_map('preg_quote', $d); |
| 255 | |
| 256 | if (in_array(' ', $d)) { |
| 257 | $d[] = '\s'; |
| 258 | } |
| 259 | |
| 260 | $split = '#(' . implode('|', $d) . ')#'; |
| 261 | |
| 262 | $str = preg_split($split, $str); |
| 263 | $d = stripslashes($d[0]); |
| 264 | } else { |
| 265 | $str = explode($d, $str); |
| 266 | } |
| 267 | |
| 268 | $i = 0; |
| 269 | $term = array(); |
| 270 | |
| 271 | foreach ($str as $key => $val) { |
| 272 | if (empty($term[$i])) { |
| 273 | $term[$i] = trim($val); |
| 274 | |
| 275 | $s1 = substr_count($term[$i], $e1); |
| 276 | $s2 = substr_count($term[$i], $e2); |
| 277 | |
| 278 | if (strpos($term[$i], '(') !== false) { |
| 279 | if ($s1 == $s2) { |
| 280 | $i++; |
| 281 | } |
| 282 | } else { |
| 283 | if ( ! (substr_count($term[$i], "'") & 1) && |
| 284 | ! (substr_count($term[$i], "\"") & 1)) { |
| 285 | $i++; |
| 286 | } |
| 287 | } |
| 288 | } else { |
| 289 | $term[$i] .= $d . trim($val); |
| 290 | $c1 = substr_count($term[$i], $e1); |
| 291 | $c2 = substr_count($term[$i], $e2); |
| 292 | |
| 293 | if (strpos($term[$i], '(') !== false) { |
| 294 | if ($c1 == $c2) { |
| 295 | $i++; |
| 296 | } |
| 297 | } else { |
| 298 | if ( ! (substr_count($term[$i], "'") & 1) && |
| 299 | ! (substr_count($term[$i], "\"") & 1)) { |
| 300 | $i++; |
| 301 | } |
| 302 | } |
| 303 | } |
| 304 | } |
| 305 | return $term; |
| 306 | } |
| 307 | |
| 308 | /** |
| 309 | * clauseExplode |
| 310 | * |
| 311 | * explodes a string into array using custom brackets and |
| 312 | * quote delimeters |
| 313 | * |
| 314 | * |
| 315 | * example: |
| 316 | * |
| 317 | * parameters: |
| 318 | * $str = "(age < 20 AND age > 18) AND name LIKE 'John Doe'" |
| 319 | * $d = ' ' |
| 320 | * $e1 = '(' |
| 321 | * $e2 = ')' |
| 322 | * |
| 323 | * would return an array: |
| 324 | * array('(age < 20 AND age > 18)', |
| 325 | * 'name', |
| 326 | * 'LIKE', |
| 327 | * 'John Doe') |
| 328 | * |
| 329 | * @param string $str |
| 330 | * @param string $d the delimeter which explodes the string |
| 331 | * @param string $e1 the first bracket, usually '(' |
| 332 | * @param string $e2 the second bracket, usually ')' |
| 333 | * |
| 334 | * @return array |
| 335 | */ |
| 336 | public function clauseExplode($str, array $d, $e1 = '(', $e2 = ')') |
| 337 | { |
| 338 | if (is_array($d)) { |
| 339 | $d = array_map('preg_quote', $d); |
| 340 | |
| 341 | if (in_array(' ', $d)) { |
| 342 | $d[] = '\s'; |
| 343 | } |
| 344 | |
| 345 | $split = '#(' . implode('|', $d) . ')#'; |
| 346 | |
| 347 | $str = preg_split($split, $str, -1, PREG_SPLIT_DELIM_CAPTURE); |
| 348 | } |
| 349 | |
| 350 | $i = 0; |
| 351 | $term = array(); |
| 352 | |
| 353 | foreach ($str as $key => $val) { |
| 354 | if ($key & 1) { |
| 355 | if (isset($term[($i - 1)]) && ! is_array($term[($i - 1)])) { |
| 356 | $term[($i - 1)] = array($term[($i - 1)], $val); |
| 357 | } |
| 358 | continue; |
| 359 | } |
| 360 | if (empty($term[$i])) { |
| 361 | $term[$i] = $val; |
| 362 | } else { |
| 363 | $term[$i] .= $str[($key - 1)] . $val; |
| 364 | } |
| 365 | |
| 366 | $c1 = substr_count($term[$i], $e1); |
| 367 | $c2 = substr_count($term[$i], $e2); |
| 368 | |
| 369 | if (strpos($term[$i], '(') !== false) { |
| 370 | if ($c1 == $c2) { |
| 371 | $i++; |
| 372 | } |
| 373 | } else { |
| 374 | if ( ! (substr_count($term[$i], "'") & 1) && |
| 375 | ! (substr_count($term[$i], "\"") & 1)) { |
| 376 | $i++; |
| 377 | } |
| 378 | } |
| 379 | } |
| 380 | $term[$i - 1] = array($term[$i - 1], ''); |
| 381 | |
| 382 | return $term; |
| 383 | } |
| 384 | } |