| 1 | /* |
| 2 | Copyright (c) 2016,2017 MariaDB |
| 3 | |
| 4 | This program is free software; you can redistribute it and/or modify |
| 5 | it under the terms of the GNU General Public License as published by |
| 6 | the Free Software Foundation; version 2 of the License. |
| 7 | |
| 8 | This program is distributed in the hope that it will be useful, |
| 9 | but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 10 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
| 11 | GNU General Public License for more details. |
| 12 | |
| 13 | You should have received a copy of the GNU General Public License |
| 14 | along with this program; if not, write to the Free Software |
| 15 | Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA */ |
| 16 | |
| 17 | #include "mariadb.h" |
| 18 | #include "item_windowfunc.h" |
| 19 | #include "sql_select.h" // test if group changed |
| 20 | |
| 21 | |
| 22 | bool |
| 23 | Item_window_func::resolve_window_name(THD *thd) |
| 24 | { |
| 25 | if (window_spec) |
| 26 | { |
| 27 | /* The window name has been already resolved */ |
| 28 | return false; |
| 29 | } |
| 30 | DBUG_ASSERT(window_name != NULL && window_spec == NULL); |
| 31 | const char *ref_name= window_name->str; |
| 32 | |
| 33 | /* !TODO: Add the code to resolve ref_name in outer queries */ |
| 34 | /* |
| 35 | First look for the deinition of the window with 'window_name' |
| 36 | in the current select |
| 37 | */ |
| 38 | List<Window_spec> curr_window_specs= |
| 39 | List<Window_spec> (thd->lex->current_select->window_specs); |
| 40 | List_iterator_fast<Window_spec> it(curr_window_specs); |
| 41 | Window_spec *win_spec; |
| 42 | while((win_spec= it++)) |
| 43 | { |
| 44 | const char *win_spec_name= win_spec->name(); |
| 45 | if (win_spec_name && |
| 46 | my_strcasecmp(system_charset_info, ref_name, win_spec_name) == 0) |
| 47 | { |
| 48 | window_spec= win_spec; |
| 49 | break; |
| 50 | } |
| 51 | } |
| 52 | |
| 53 | if (!window_spec) |
| 54 | { |
| 55 | my_error(ER_WRONG_WINDOW_SPEC_NAME, MYF(0), ref_name); |
| 56 | return true; |
| 57 | } |
| 58 | |
| 59 | return false; |
| 60 | } |
| 61 | |
| 62 | |
| 63 | void |
| 64 | Item_window_func::update_used_tables() |
| 65 | { |
| 66 | used_tables_cache= 0; |
| 67 | window_func()->update_used_tables(); |
| 68 | used_tables_cache|= window_func()->used_tables(); |
| 69 | for (ORDER *ord= window_spec->partition_list->first; ord; ord=ord->next) |
| 70 | { |
| 71 | Item *item= *ord->item; |
| 72 | item->update_used_tables(); |
| 73 | used_tables_cache|= item->used_tables(); |
| 74 | } |
| 75 | for (ORDER *ord= window_spec->order_list->first; ord; ord=ord->next) |
| 76 | { |
| 77 | Item *item= *ord->item; |
| 78 | item->update_used_tables(); |
| 79 | used_tables_cache|= item->used_tables(); |
| 80 | } |
| 81 | } |
| 82 | |
| 83 | |
| 84 | bool |
| 85 | Item_window_func::fix_fields(THD *thd, Item **ref) |
| 86 | { |
| 87 | DBUG_ASSERT(fixed == 0); |
| 88 | |
| 89 | enum_parsing_place place= thd->lex->current_select->context_analysis_place; |
| 90 | |
| 91 | if (!(place == SELECT_LIST || place == IN_ORDER_BY)) |
| 92 | { |
| 93 | my_error(ER_WRONG_PLACEMENT_OF_WINDOW_FUNCTION, MYF(0)); |
| 94 | return true; |
| 95 | } |
| 96 | |
| 97 | if (window_name && resolve_window_name(thd)) |
| 98 | return true; |
| 99 | |
| 100 | if (window_spec->window_frame && is_frame_prohibited()) |
| 101 | { |
| 102 | my_error(ER_NOT_ALLOWED_WINDOW_FRAME, MYF(0), window_func()->func_name()); |
| 103 | return true; |
| 104 | } |
| 105 | |
| 106 | if (window_spec->order_list->elements == 0 && is_order_list_mandatory()) |
| 107 | { |
| 108 | my_error(ER_NO_ORDER_LIST_IN_WINDOW_SPEC, MYF(0), window_func()->func_name()); |
| 109 | return true; |
| 110 | } |
| 111 | |
| 112 | /* |
| 113 | TODO: why the last parameter is 'ref' in this call? What if window_func |
| 114 | decides to substitute itself for something else and does *ref=.... ? |
| 115 | This will substitute *this (an Item_window_func object) with Item_sum |
| 116 | object. Is this the intent? |
| 117 | */ |
| 118 | if (window_func()->fix_fields(thd, ref)) |
| 119 | return true; |
| 120 | |
| 121 | const_item_cache= false; |
| 122 | with_window_func= true; |
| 123 | with_sum_func= false; |
| 124 | |
| 125 | fix_length_and_dec(); |
| 126 | |
| 127 | max_length= window_func()->max_length; |
| 128 | maybe_null= window_func()->maybe_null; |
| 129 | |
| 130 | fixed= 1; |
| 131 | set_phase_to_initial(); |
| 132 | return false; |
| 133 | } |
| 134 | |
| 135 | |
| 136 | /* |
| 137 | @detail |
| 138 | Window function evaluates its arguments when it is scanning the temporary |
| 139 | table in partition/order-by order. That is, arguments should be read from |
| 140 | the temporary table, not from the original base columns. |
| 141 | |
| 142 | In order for this to work, we need to call "split_sum_func" for each |
| 143 | argument. The effect of the call is: |
| 144 | 1. the argument is added into ref_pointer_array. This will cause the |
| 145 | argument to be saved in the temp.table |
| 146 | 2. argument item is replaced with an Item_ref object. this object refers |
| 147 | the argument through the ref_pointer_array. |
| 148 | |
| 149 | then, change_to_use_tmp_fields() will replace ref_pointer_array with an |
| 150 | array that points to the temp.table fields. |
| 151 | This way, when window_func attempts to evaluate its arguments, it will use |
| 152 | Item_ref objects which will read data from the temp.table. |
| 153 | |
| 154 | Note: Before window functions, aggregate functions never needed to do such |
| 155 | transformations on their arguments. This is because grouping operation |
| 156 | does not need to read from the temp.table. |
| 157 | (Q: what happens when we first sort and then do grouping in a |
| 158 | group-after-group mode? dont group by items read from temp.table, then?) |
| 159 | */ |
| 160 | |
| 161 | void Item_window_func::split_sum_func(THD *thd, Ref_ptr_array ref_pointer_array, |
| 162 | List<Item> &fields, uint flags) |
| 163 | { |
| 164 | for (uint i=0; i < window_func()->argument_count(); i++) |
| 165 | { |
| 166 | Item **p_item= &window_func()->arguments()[i]; |
| 167 | (*p_item)->split_sum_func2(thd, ref_pointer_array, fields, p_item, flags); |
| 168 | } |
| 169 | window_func()->setup_caches(thd); |
| 170 | } |
| 171 | |
| 172 | bool Item_window_func::check_result_type_of_order_item() |
| 173 | { |
| 174 | if (only_single_element_order_list()) |
| 175 | { |
| 176 | Item_result rtype= window_spec->order_list->first->item[0]->cmp_type(); |
| 177 | // TODO (varun) : support date type in percentile_cont function |
| 178 | if (rtype != REAL_RESULT && rtype != INT_RESULT && |
| 179 | rtype != DECIMAL_RESULT && rtype != TIME_RESULT) |
| 180 | { |
| 181 | my_error(ER_WRONG_TYPE_FOR_PERCENTILE_FUNC, MYF(0), window_func()->func_name()); |
| 182 | return TRUE; |
| 183 | } |
| 184 | setting_handler_for_percentile_functions(rtype); |
| 185 | } |
| 186 | return FALSE; |
| 187 | } |
| 188 | |
| 189 | /* |
| 190 | This must be called before attempting to compute the window function values. |
| 191 | @detail |
| 192 | If we attempt to do it in fix_fields(), partition_fields will refer |
| 193 | to the original window function arguments. |
| 194 | We need it to refer to temp.table columns. |
| 195 | */ |
| 196 | |
| 197 | void Item_sum_rank::setup_window_func(THD *thd, Window_spec *window_spec) |
| 198 | { |
| 199 | /* TODO: move this into Item_window_func? */ |
| 200 | peer_tracker = new Group_bound_tracker(thd, window_spec->order_list); |
| 201 | peer_tracker->init(); |
| 202 | clear(); |
| 203 | } |
| 204 | |
| 205 | void Item_sum_dense_rank::setup_window_func(THD *thd, Window_spec *window_spec) |
| 206 | { |
| 207 | /* TODO: consider moving this && Item_sum_rank's implementation */ |
| 208 | peer_tracker = new Group_bound_tracker(thd, window_spec->order_list); |
| 209 | peer_tracker->init(); |
| 210 | clear(); |
| 211 | } |
| 212 | |
| 213 | void Item_sum_percentile_disc::setup_window_func(THD *thd, Window_spec *window_spec) |
| 214 | { |
| 215 | order_item= window_spec->order_list->first->item[0]; |
| 216 | if (!(value= order_item->get_cache(thd))) |
| 217 | return; |
| 218 | value->setup(thd, order_item); |
| 219 | value->store(order_item); |
| 220 | } |
| 221 | |
| 222 | void Item_sum_percentile_cont::setup_window_func(THD *thd, Window_spec *window_spec) |
| 223 | { |
| 224 | order_item= window_spec->order_list->first->item[0]; |
| 225 | /* TODO(varun): need to discuss and finalise what type should we |
| 226 | return for percentile cont functions |
| 227 | */ |
| 228 | if (!(ceil_value= order_item->get_cache(thd))) |
| 229 | return; |
| 230 | ceil_value->setup(thd, order_item); |
| 231 | ceil_value->store(order_item); |
| 232 | |
| 233 | if (!(floor_value= order_item->get_cache(thd))) |
| 234 | return; |
| 235 | floor_value->setup(thd, order_item); |
| 236 | floor_value->store(order_item); |
| 237 | } |
| 238 | bool Item_sum_percentile_cont::fix_fields(THD *thd, Item **ref) |
| 239 | { |
| 240 | bool res; |
| 241 | res= Item_sum_num::fix_fields(thd, ref); |
| 242 | if (res) |
| 243 | return res; |
| 244 | |
| 245 | switch(args[0]->cmp_type()) |
| 246 | { |
| 247 | case DECIMAL_RESULT: |
| 248 | case REAL_RESULT: |
| 249 | case INT_RESULT: |
| 250 | break; |
| 251 | default: |
| 252 | my_error(ER_WRONG_TYPE_OF_ARGUMENT, MYF(0), func_name()); |
| 253 | return TRUE; |
| 254 | } |
| 255 | return res; |
| 256 | } |
| 257 | bool Item_sum_percentile_disc::fix_fields(THD *thd, Item **ref) |
| 258 | { |
| 259 | bool res; |
| 260 | res= Item_sum_num::fix_fields(thd, ref); |
| 261 | if (res) |
| 262 | return res; |
| 263 | |
| 264 | switch(args[0]->cmp_type()) |
| 265 | { |
| 266 | case DECIMAL_RESULT: |
| 267 | case REAL_RESULT: |
| 268 | case INT_RESULT: |
| 269 | break; |
| 270 | default: |
| 271 | my_error(ER_WRONG_TYPE_OF_ARGUMENT, MYF(0), func_name()); |
| 272 | return TRUE; |
| 273 | } |
| 274 | return res; |
| 275 | |
| 276 | } |
| 277 | |
| 278 | bool Item_sum_dense_rank::add() |
| 279 | { |
| 280 | if (peer_tracker->check_if_next_group() || first_add) |
| 281 | { |
| 282 | first_add= false; |
| 283 | dense_rank++; |
| 284 | } |
| 285 | |
| 286 | return false; |
| 287 | } |
| 288 | |
| 289 | |
| 290 | bool Item_sum_rank::add() |
| 291 | { |
| 292 | row_number++; |
| 293 | if (peer_tracker->check_if_next_group()) |
| 294 | { |
| 295 | /* Row value changed */ |
| 296 | cur_rank= row_number; |
| 297 | } |
| 298 | return false; |
| 299 | } |
| 300 | |
| 301 | bool Item_sum_percent_rank::add() |
| 302 | { |
| 303 | row_number++; |
| 304 | if (peer_tracker->check_if_next_group()) |
| 305 | { |
| 306 | /* Row value changed. */ |
| 307 | cur_rank= row_number; |
| 308 | } |
| 309 | return false; |
| 310 | } |
| 311 | |
| 312 | void Item_sum_percent_rank::setup_window_func(THD *thd, Window_spec *window_spec) |
| 313 | { |
| 314 | /* TODO: move this into Item_window_func? */ |
| 315 | peer_tracker = new Group_bound_tracker(thd, window_spec->order_list); |
| 316 | peer_tracker->init(); |
| 317 | clear(); |
| 318 | } |
| 319 | |
| 320 | |
| 321 | bool Item_sum_hybrid_simple::fix_fields(THD *thd, Item **ref) |
| 322 | { |
| 323 | DBUG_ASSERT(fixed == 0); |
| 324 | |
| 325 | if (init_sum_func_check(thd)) |
| 326 | return TRUE; |
| 327 | |
| 328 | for (uint i= 0; i < arg_count; i++) |
| 329 | { |
| 330 | Item *item= args[i]; |
| 331 | // 'item' can be changed during fix_fields |
| 332 | if ((!item->fixed && item->fix_fields(thd, args)) || |
| 333 | (item= args[i])->check_cols(1)) |
| 334 | return TRUE; |
| 335 | } |
| 336 | Type_std_attributes::set(args[0]); |
| 337 | for (uint i= 0; i < arg_count && !m_with_subquery; i++) |
| 338 | m_with_subquery|= args[i]->with_subquery(); |
| 339 | |
| 340 | Item *item2= args[0]->real_item(); |
| 341 | if (item2->type() == Item::FIELD_ITEM) |
| 342 | set_handler(item2->type_handler()); |
| 343 | else if (args[0]->cmp_type() == TIME_RESULT) |
| 344 | set_handler(item2->type_handler()); |
| 345 | else |
| 346 | set_handler_by_result_type(item2->result_type(), |
| 347 | max_length, collation.collation); |
| 348 | |
| 349 | switch (result_type()) { |
| 350 | case INT_RESULT: |
| 351 | case DECIMAL_RESULT: |
| 352 | case STRING_RESULT: |
| 353 | break; |
| 354 | case REAL_RESULT: |
| 355 | max_length= float_length(decimals); |
| 356 | break; |
| 357 | case ROW_RESULT: |
| 358 | case TIME_RESULT: |
| 359 | DBUG_ASSERT(0); // XXX(cvicentiu) Should this never happen? |
| 360 | return TRUE; |
| 361 | }; |
| 362 | setup_hybrid(thd, args[0]); |
| 363 | /* MIN/MAX can return NULL for empty set indepedent of the used column */ |
| 364 | maybe_null= 1; |
| 365 | result_field=0; |
| 366 | null_value=1; |
| 367 | fix_length_and_dec(); |
| 368 | |
| 369 | if (check_sum_func(thd, ref)) |
| 370 | return TRUE; |
| 371 | for (uint i= 0; i < arg_count; i++) |
| 372 | { |
| 373 | orig_args[i]= args[i]; |
| 374 | } |
| 375 | fixed= 1; |
| 376 | return FALSE; |
| 377 | } |
| 378 | |
| 379 | bool Item_sum_hybrid_simple::add() |
| 380 | { |
| 381 | value->store(args[0]); |
| 382 | value->cache_value(); |
| 383 | null_value= value->null_value; |
| 384 | return false; |
| 385 | } |
| 386 | |
| 387 | void Item_sum_hybrid_simple::setup_hybrid(THD *thd, Item *item) |
| 388 | { |
| 389 | if (!(value= item->get_cache(thd))) |
| 390 | return; |
| 391 | value->setup(thd, item); |
| 392 | value->store(item); |
| 393 | if (!item->const_item()) |
| 394 | value->set_used_tables(RAND_TABLE_BIT); |
| 395 | collation.set(item->collation); |
| 396 | } |
| 397 | |
| 398 | double Item_sum_hybrid_simple::val_real() |
| 399 | { |
| 400 | DBUG_ASSERT(fixed == 1); |
| 401 | if (null_value) |
| 402 | return 0.0; |
| 403 | double retval= value->val_real(); |
| 404 | if ((null_value= value->null_value)) |
| 405 | DBUG_ASSERT(retval == 0.0); |
| 406 | return retval; |
| 407 | } |
| 408 | |
| 409 | longlong Item_sum_hybrid_simple::val_int() |
| 410 | { |
| 411 | DBUG_ASSERT(fixed == 1); |
| 412 | if (null_value) |
| 413 | return 0; |
| 414 | longlong retval= value->val_int(); |
| 415 | if ((null_value= value->null_value)) |
| 416 | DBUG_ASSERT(retval == 0); |
| 417 | return retval; |
| 418 | } |
| 419 | |
| 420 | my_decimal *Item_sum_hybrid_simple::val_decimal(my_decimal *val) |
| 421 | { |
| 422 | DBUG_ASSERT(fixed == 1); |
| 423 | if (null_value) |
| 424 | return 0; |
| 425 | my_decimal *retval= value->val_decimal(val); |
| 426 | if ((null_value= value->null_value)) |
| 427 | DBUG_ASSERT(retval == NULL); |
| 428 | return retval; |
| 429 | } |
| 430 | |
| 431 | String * |
| 432 | Item_sum_hybrid_simple::val_str(String *str) |
| 433 | { |
| 434 | DBUG_ASSERT(fixed == 1); |
| 435 | if (null_value) |
| 436 | return 0; |
| 437 | String *retval= value->val_str(str); |
| 438 | if ((null_value= value->null_value)) |
| 439 | DBUG_ASSERT(retval == NULL); |
| 440 | return retval; |
| 441 | } |
| 442 | |
| 443 | bool Item_sum_hybrid_simple::get_date(MYSQL_TIME *ltime, ulonglong fuzzydate) |
| 444 | { |
| 445 | DBUG_ASSERT(fixed == 1); |
| 446 | if (null_value) |
| 447 | return 0; |
| 448 | bool retval= value->get_date(ltime, fuzzydate); |
| 449 | if ((null_value= value->null_value)) |
| 450 | DBUG_ASSERT(retval == true); |
| 451 | return retval; |
| 452 | } |
| 453 | |
| 454 | Field *Item_sum_hybrid_simple::create_tmp_field(bool group, TABLE *table) |
| 455 | { |
| 456 | DBUG_ASSERT(0); |
| 457 | return NULL; |
| 458 | } |
| 459 | |
| 460 | void Item_sum_hybrid_simple::reset_field() |
| 461 | { |
| 462 | switch(result_type()) { |
| 463 | case STRING_RESULT: |
| 464 | { |
| 465 | char buff[MAX_FIELD_WIDTH]; |
| 466 | String tmp(buff,sizeof(buff),result_field->charset()),*res; |
| 467 | |
| 468 | res=args[0]->val_str(&tmp); |
| 469 | if (args[0]->null_value) |
| 470 | { |
| 471 | result_field->set_null(); |
| 472 | result_field->reset(); |
| 473 | } |
| 474 | else |
| 475 | { |
| 476 | result_field->set_notnull(); |
| 477 | result_field->store(res->ptr(),res->length(),tmp.charset()); |
| 478 | } |
| 479 | break; |
| 480 | } |
| 481 | case INT_RESULT: |
| 482 | { |
| 483 | longlong nr=args[0]->val_int(); |
| 484 | |
| 485 | if (maybe_null) |
| 486 | { |
| 487 | if (args[0]->null_value) |
| 488 | { |
| 489 | nr=0; |
| 490 | result_field->set_null(); |
| 491 | } |
| 492 | else |
| 493 | result_field->set_notnull(); |
| 494 | } |
| 495 | result_field->store(nr, unsigned_flag); |
| 496 | break; |
| 497 | } |
| 498 | case REAL_RESULT: |
| 499 | { |
| 500 | double nr= args[0]->val_real(); |
| 501 | |
| 502 | if (maybe_null) |
| 503 | { |
| 504 | if (args[0]->null_value) |
| 505 | { |
| 506 | nr=0.0; |
| 507 | result_field->set_null(); |
| 508 | } |
| 509 | else |
| 510 | result_field->set_notnull(); |
| 511 | } |
| 512 | result_field->store(nr); |
| 513 | break; |
| 514 | } |
| 515 | case DECIMAL_RESULT: |
| 516 | { |
| 517 | my_decimal value_buff, *arg_dec= args[0]->val_decimal(&value_buff); |
| 518 | |
| 519 | if (maybe_null) |
| 520 | { |
| 521 | if (args[0]->null_value) |
| 522 | result_field->set_null(); |
| 523 | else |
| 524 | result_field->set_notnull(); |
| 525 | } |
| 526 | /* |
| 527 | We must store zero in the field as we will use the field value in |
| 528 | add() |
| 529 | */ |
| 530 | if (!arg_dec) // Null |
| 531 | arg_dec= &decimal_zero; |
| 532 | result_field->store_decimal(arg_dec); |
| 533 | break; |
| 534 | } |
| 535 | case ROW_RESULT: |
| 536 | case TIME_RESULT: |
| 537 | DBUG_ASSERT(0); |
| 538 | } |
| 539 | } |
| 540 | |
| 541 | void Item_sum_hybrid_simple::update_field() |
| 542 | { |
| 543 | DBUG_ASSERT(0); |
| 544 | } |
| 545 | |
| 546 | void Item_window_func::print(String *str, enum_query_type query_type) |
| 547 | { |
| 548 | window_func()->print(str, query_type); |
| 549 | str->append(" over " ); |
| 550 | #ifndef DBUG_OFF |
| 551 | if (!window_spec) // one can call dbug_print_item() anytime in gdb |
| 552 | str->append(window_name); |
| 553 | else |
| 554 | #endif |
| 555 | window_spec->print(str, query_type); |
| 556 | } |
| 557 | |