sclfunc.c 46.1 KB
Newer Older
H
Haojun Liao 已提交
1 2 3
#include "function.h"
#include "scalar.h"
#include "tdatablock.h"
4
#include "ttime.h"
H
Haojun Liao 已提交
5
#include "sclInt.h"
D
dapan1121 已提交
6
#include "sclvector.h"
7

G
Ganlin Zhao 已提交
8 9 10 11 12
typedef float (*_float_fn)(float);
typedef double (*_double_fn)(double);
typedef double (*_double_fn_2)(double, double);
typedef int (*_conv_fn)(int);
typedef void (*_trim_fn)(char *, char*, int32_t, int32_t);
G
Ganlin Zhao 已提交
13
typedef int16_t (*_len_fn)(char *, int32_t);
G
Ganlin Zhao 已提交
14

15
/** Math functions **/
16
static double tlog(double v, double base) {
G
Ganlin Zhao 已提交
17
  return log(v) / log(base);
18 19
}

20
int32_t absFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
H
Haojun Liao 已提交
21 22 23 24 25
  SColumnInfoData *pInputData  = pInput->columnData;
  SColumnInfoData *pOutputData = pOutput->columnData;

  int32_t type = GET_PARAM_TYPE(pInput);
  if (!IS_NUMERIC_TYPE(type)) {
26 27
    return TSDB_CODE_FAILED;
  }
28

H
Haojun Liao 已提交
29
  switch (type) {
30
    case TSDB_DATA_TYPE_FLOAT: {
H
Haojun Liao 已提交
31 32
      float *in  = (float *)pInputData->pData;
      float *out = (float *)pOutputData->pData;
33
      for (int32_t i = 0; i < pInput->numOfRows; ++i) {
H
Haojun Liao 已提交
34
        if (colDataIsNull_f(pInputData->nullbitmap, i)) {
35
          colDataSetNull_f(pOutputData->nullbitmap, i);
H
Haojun Liao 已提交
36 37
          continue;
        }
38
        out[i] = (in[i] >= 0)? in[i] : -in[i];
39
      }
H
Haojun Liao 已提交
40
      break;
41 42 43
    }

    case TSDB_DATA_TYPE_DOUBLE: {
H
Haojun Liao 已提交
44 45
      double *in  = (double *)pInputData->pData;
      double *out = (double *)pOutputData->pData;
46
      for (int32_t i = 0; i < pInput->numOfRows; ++i) {
H
Haojun Liao 已提交
47
        if (colDataIsNull_f(pInputData->nullbitmap, i)) {
48
          colDataSetNull_f(pOutputData->nullbitmap, i);
H
Haojun Liao 已提交
49 50
          continue;
        }
51
        out[i] = (in[i] >= 0)? in[i] : -in[i];
H
Haojun Liao 已提交
52 53
      }
      break;
54 55
    }

H
Haojun Liao 已提交
56
    case TSDB_DATA_TYPE_TINYINT: {
H
Haojun Liao 已提交
57 58
      int8_t *in  = (int8_t *)pInputData->pData;
      int8_t *out = (int8_t *)pOutputData->pData;
59
      for (int32_t i = 0; i < pInput->numOfRows; ++i) {
H
Haojun Liao 已提交
60
        if (colDataIsNull_f(pInputData->nullbitmap, i)) {
61
          colDataSetNull_f(pOutputData->nullbitmap, i);
H
Haojun Liao 已提交
62 63
          continue;
        }
64
        out[i] = (in[i] >= 0)? in[i] : -in[i];
65
      }
H
Haojun Liao 已提交
66 67
      break;
    }
68

H
Haojun Liao 已提交
69
    case TSDB_DATA_TYPE_SMALLINT: {
H
Haojun Liao 已提交
70 71
      int16_t *in  = (int16_t *)pInputData->pData;
      int16_t *out = (int16_t *)pOutputData->pData;
72
      for (int32_t i = 0; i < pInput->numOfRows; ++i) {
H
Haojun Liao 已提交
73
        if (colDataIsNull_f(pInputData->nullbitmap, i)) {
74
          colDataSetNull_f(pOutputData->nullbitmap, i);
H
Haojun Liao 已提交
75 76
          continue;
        }
77
        out[i] = (in[i] >= 0)? in[i] : -in[i];
78
      }
H
Haojun Liao 已提交
79 80
      break;
    }
81

H
Haojun Liao 已提交
82
    case TSDB_DATA_TYPE_INT: {
H
Haojun Liao 已提交
83 84
      int32_t *in  = (int32_t *)pInputData->pData;
      int32_t *out = (int32_t *)pOutputData->pData;
85
      for (int32_t i = 0; i < pInput->numOfRows; ++i) {
H
Haojun Liao 已提交
86
        if (colDataIsNull_f(pInputData->nullbitmap, i)) {
87
          colDataSetNull_f(pOutputData->nullbitmap, i);
H
Haojun Liao 已提交
88 89
          continue;
        }
90
        out[i] = (in[i] >= 0)? in[i] : -in[i];
91
      }
H
Haojun Liao 已提交
92 93
      break;
    }
94

H
Haojun Liao 已提交
95
    case TSDB_DATA_TYPE_BIGINT: {
H
Haojun Liao 已提交
96 97
      int64_t *in  = (int64_t *)pInputData->pData;
      int64_t *out = (int64_t *)pOutputData->pData;
98
      for (int32_t i = 0; i < pInput->numOfRows; ++i) {
H
Haojun Liao 已提交
99
        if (colDataIsNull_f(pInputData->nullbitmap, i)) {
100
          colDataSetNull_f(pOutputData->nullbitmap, i);
H
Haojun Liao 已提交
101 102
          continue;
        }
103
        out[i] = (in[i] >= 0)? in[i] : -in[i];
104
      }
H
Haojun Liao 已提交
105
      break;
106 107
    }

H
Haojun Liao 已提交
108 109
    default: {
      colDataAssign(pOutputData, pInputData, pInput->numOfRows);
110 111
    }
  }
112

H
Haojun Liao 已提交
113
  pOutput->numOfRows = pInput->numOfRows;
114
  return TSDB_CODE_SUCCESS;
115 116
}

117
static int32_t doScalarFunctionUnique(SScalarParam *pInput, int32_t inputNum, SScalarParam* pOutput, _double_fn valFn) {
H
Haojun Liao 已提交
118 119
  int32_t type = GET_PARAM_TYPE(pInput);
  if (inputNum != 1 || !IS_NUMERIC_TYPE(type)) {
120 121
    return TSDB_CODE_FAILED;
  }
122

123
  SColumnInfoData *pInputData = pInput->columnData;
H
Haojun Liao 已提交
124
  SColumnInfoData *pOutputData = pOutput->columnData;
125

126
  _getDoubleValue_fn_t getValueFn = getVectorDoubleValueFn(type);
127

128
  double *out = (double *)pOutputData->pData;
129

130 131 132
  for (int32_t i = 0; i < pInput->numOfRows; ++i) {
    if (colDataIsNull_f(pInputData->nullbitmap, i)) {
      colDataSetNull_f(pOutputData->nullbitmap, i);
133
      continue;
134
    }
135
    out[i] = valFn(getValueFn(pInputData->pData, i));
136
  }
137

H
Haojun Liao 已提交
138
  pOutput->numOfRows = pInput->numOfRows;
139
  return TSDB_CODE_SUCCESS;
140 141
}

142
static int32_t doScalarFunctionUnique2(SScalarParam *pInput, int32_t inputNum, SScalarParam* pOutput, _double_fn_2 valFn) {
G
Ganlin Zhao 已提交
143
  if (inputNum != 2 || !IS_NUMERIC_TYPE(GET_PARAM_TYPE(&pInput[0])) || !IS_NUMERIC_TYPE(GET_PARAM_TYPE(&pInput[1]))) {
144 145 146
    return TSDB_CODE_FAILED;
  }

G
Ganlin Zhao 已提交
147
  SColumnInfoData *pInputData[2];
H
Haojun Liao 已提交
148
  SColumnInfoData *pOutputData = pOutput->columnData;
G
Ganlin Zhao 已提交
149
  _getDoubleValue_fn_t getValueFn[2];
150

G
Ganlin Zhao 已提交
151 152 153 154
  for (int32_t i = 0; i < inputNum; ++i) {
    pInputData[i] = pInput[i].columnData;
    getValueFn[i]= getVectorDoubleValueFn(GET_PARAM_TYPE(&pInput[i]));
  }
155

156
  double *out = (double *)pOutputData->pData;
157

158
  for (int32_t i = 0; i < pInput->numOfRows; ++i) {
G
Ganlin Zhao 已提交
159 160
    if (colDataIsNull_f(pInputData[0]->nullbitmap, i) ||
        colDataIsNull_f(pInputData[1]->nullbitmap, 0)) {
161 162
      colDataSetNull_f(pOutputData->nullbitmap, i);
      continue;
163
    }
G
Ganlin Zhao 已提交
164
    out[i] = valFn(getValueFn[0](pInputData[0]->pData, i), getValueFn[1](pInputData[1]->pData, 0));
165
  }
166

H
Haojun Liao 已提交
167
  pOutput->numOfRows = pInput->numOfRows;
168 169 170
  return TSDB_CODE_SUCCESS;
}

171
static int32_t doScalarFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam* pOutput, _float_fn f1, _double_fn d1) {
H
Haojun Liao 已提交
172 173
  int32_t type = GET_PARAM_TYPE(pInput);
  if (inputNum != 1 || !IS_NUMERIC_TYPE(type)) {
174 175 176
    return TSDB_CODE_FAILED;
  }

H
Haojun Liao 已提交
177 178
  SColumnInfoData *pInputData  = pInput->columnData;
  SColumnInfoData *pOutputData = pOutput->columnData;
179

H
Haojun Liao 已提交
180 181 182 183
  switch (type) {
    case TSDB_DATA_TYPE_FLOAT: {
      float *in  = (float *)pInputData->pData;
      float *out = (float *)pOutputData->pData;
184

185
      for (int32_t i = 0; i < pInput->numOfRows; ++i) {
H
Haojun Liao 已提交
186
        if (colDataIsNull_f(pInputData->nullbitmap, i)) {
187
          colDataSetNull_f(pOutputData->nullbitmap, i);
H
Haojun Liao 已提交
188 189
          continue;
        }
190
        out[i] = f1(in[i]);
191
      }
H
Haojun Liao 已提交
192 193
      break;
    }
194

H
Haojun Liao 已提交
195 196 197
    case TSDB_DATA_TYPE_DOUBLE: {
      double *in  = (double *)pInputData->pData;
      double *out = (double *)pOutputData->pData;
198

199
      for (int32_t i = 0; i < pInput->numOfRows; ++i) {
H
Haojun Liao 已提交
200
        if (colDataIsNull_f(pInputData->nullbitmap, i)) {
201
          colDataSetNull_f(pOutputData->nullbitmap, i);
H
Haojun Liao 已提交
202 203
          continue;
        }
204
        out[i] = d1(in[i]);
205
      }
H
Haojun Liao 已提交
206 207 208 209 210
      break;
    }

    default: {
      colDataAssign(pOutputData, pInputData, pInput->numOfRows);
211
    }
212 213
  }

H
Haojun Liao 已提交
214
  pOutput->numOfRows = pInput->numOfRows;
215 216
  return TSDB_CODE_SUCCESS;
}
217

G
Ganlin Zhao 已提交
218
/** String functions **/
219
static int16_t tlength(char *input, int32_t type) {
G
Ganlin Zhao 已提交
220 221 222
  return varDataLen(input);
}

223
static int16_t tcharlength(char *input, int32_t type) {
G
Ganlin Zhao 已提交
224 225 226 227
  if (type == TSDB_DATA_TYPE_VARCHAR) {
    return varDataLen(input);
  } else { //NCHAR
    return varDataLen(input) / TSDB_NCHAR_SIZE;
G
Ganlin Zhao 已提交
228
  }
G
Ganlin Zhao 已提交
229
}
G
Ganlin Zhao 已提交
230

231
static void tltrim(char *input, char *output, int32_t type, int32_t charLen) {
G
Ganlin Zhao 已提交
232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247
  int32_t numOfSpaces = 0;
  if (type == TSDB_DATA_TYPE_VARCHAR) {
    for (int32_t i = 0; i < charLen; ++i) {
      if (!isspace(*(varDataVal(input) + i))) {
        break;
      }
      numOfSpaces++;
    }
  } else { //NCHAR
    for (int32_t i = 0; i < charLen; ++i) {
      if (!iswspace(*((uint32_t *)varDataVal(input) + i))) {
        break;
      }
      numOfSpaces++;
    }
  }
G
Ganlin Zhao 已提交
248

G
Ganlin Zhao 已提交
249 250 251 252 253 254 255 256
  int32_t resLen;
  if (type == TSDB_DATA_TYPE_VARCHAR) {
    resLen = charLen - numOfSpaces;
    memcpy(varDataVal(output), varDataVal(input) + numOfSpaces, resLen);
  } else {
    resLen = (charLen - numOfSpaces) * TSDB_NCHAR_SIZE;
    memcpy(varDataVal(output), varDataVal(input) + numOfSpaces * TSDB_NCHAR_SIZE, resLen);
  }
G
Ganlin Zhao 已提交
257

G
Ganlin Zhao 已提交
258 259 260
  varDataSetLen(output, resLen);
}

261
static void trtrim(char *input, char *output, int32_t type, int32_t charLen) {
G
Ganlin Zhao 已提交
262 263 264 265 266 267 268 269 270
  int32_t numOfSpaces = 0;
  if (type == TSDB_DATA_TYPE_VARCHAR) {
    for (int32_t i = charLen - 1; i >= 0; --i) {
      if (!isspace(*(varDataVal(input) + i))) {
        break;
      }
      numOfSpaces++;
    }
  } else { //NCHAR
271
    for (int32_t i = charLen - 1; i >= 0; --i) {
G
Ganlin Zhao 已提交
272 273 274 275
      if (!iswspace(*((uint32_t *)varDataVal(input) + i))) {
        break;
      }
      numOfSpaces++;
G
Ganlin Zhao 已提交
276
    }
G
Ganlin Zhao 已提交
277
  }
G
Ganlin Zhao 已提交
278

G
Ganlin Zhao 已提交
279 280 281 282 283
  int32_t resLen;
  if (type == TSDB_DATA_TYPE_VARCHAR) {
    resLen = charLen - numOfSpaces;
  } else {
    resLen = (charLen - numOfSpaces) * TSDB_NCHAR_SIZE;
G
Ganlin Zhao 已提交
284
  }
G
Ganlin Zhao 已提交
285
  memcpy(varDataVal(output), varDataVal(input), resLen);
G
Ganlin Zhao 已提交
286

G
Ganlin Zhao 已提交
287
  varDataSetLen(output, resLen);
G
Ganlin Zhao 已提交
288 289
}

290
static int32_t doLengthFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput, _len_fn lenFn) {
G
Ganlin Zhao 已提交
291 292 293 294 295 296 297 298
  int32_t type = GET_PARAM_TYPE(pInput);
  if (inputNum != 1 || !IS_VAR_DATA_TYPE(type)) {
    return TSDB_CODE_FAILED;
  }

  SColumnInfoData *pInputData  = pInput->columnData;
  SColumnInfoData *pOutputData = pOutput->columnData;

299
  char *in = pInputData->pData + pInputData->varmeta.offset[0];
G
Ganlin Zhao 已提交
300 301 302
  int16_t *out = (int16_t *)pOutputData->pData;

  for (int32_t i = 0; i < pInput->numOfRows; ++i) {
303
    if (colDataIsNull_s(pInputData, i)) {
G
Ganlin Zhao 已提交
304 305 306 307
      colDataSetNull_f(pOutputData->nullbitmap, i);
      continue;
    }

G
Ganlin Zhao 已提交
308 309
    out[i] = lenFn(in, type);
    in += varDataTLen(in);
G
Ganlin Zhao 已提交
310 311 312 313 314
  }

  pOutput->numOfRows = pInput->numOfRows;
  return TSDB_CODE_SUCCESS;
}
315

316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331
static int32_t concatCopyHelper(const char *input, char *output, bool hasNcharCol, int32_t type, int16_t *dataLen) {
  if (hasNcharCol && type == TSDB_DATA_TYPE_VARCHAR) {
    TdUcs4 *newBuf = taosMemoryCalloc((varDataLen(input) + 1) * TSDB_NCHAR_SIZE, 1);
    bool ret = taosMbsToUcs4(varDataVal(input), varDataLen(input), newBuf, (varDataLen(input) + 1) * TSDB_NCHAR_SIZE, NULL);
    if (!ret) {
      taosMemoryFree(newBuf);
      return TSDB_CODE_FAILED;
    }
    memcpy(varDataVal(output) + *dataLen, newBuf, varDataLen(input) * TSDB_NCHAR_SIZE);
    *dataLen += varDataLen(input) * TSDB_NCHAR_SIZE;
    taosMemoryFree(newBuf);
  } else {
    memcpy(varDataVal(output) + *dataLen, varDataVal(input), varDataLen(input));
    *dataLen += varDataLen(input);
  }
  return TSDB_CODE_SUCCESS;
332
}
G
Ganlin Zhao 已提交
333

334 335 336 337 338 339 340 341 342 343 344 345 346
static int32_t getNumOfNullEntries(SColumnInfoData *pColumnInfoData, int32_t numOfRows) {
  int32_t numOfNulls = 0;
  if (!pColumnInfoData->hasNull) {
    return numOfNulls;
  }
  for (int i = 0; i < numOfRows; ++i) {
    if (pColumnInfoData->varmeta.offset[i] == -1) {
      numOfNulls++;
    }
  }
  return numOfNulls;
}

G
Ganlin Zhao 已提交
347 348 349 350 351 352 353
int32_t concatFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
  if (inputNum < 2 || inputNum > 8) { // concat accpet 2-8 input strings
    return TSDB_CODE_FAILED;
  }

  SColumnInfoData **pInputData = taosMemoryCalloc(inputNum, sizeof(SColumnInfoData *));
  SColumnInfoData *pOutputData = pOutput->columnData;
354
  char **input = taosMemoryCalloc(inputNum, POINTER_BYTES);
355
  char *outputBuf = NULL;
G
Ganlin Zhao 已提交
356

357
  int32_t inputLen = 0;
358
  int32_t numOfRows = 0;
359
  bool hasNcharCol = false;
360
  for (int32_t i = 0; i < inputNum; ++i) {
361 362
    int32_t type = GET_PARAM_TYPE(&pInput[i]);
    if (!IS_VAR_DATA_TYPE(type)) {
363 364
      return TSDB_CODE_FAILED;
    }
365 366 367
    if (type == TSDB_DATA_TYPE_NCHAR) {
      hasNcharCol = true;
    }
368 369 370 371
    if (pInput[i].numOfRows > numOfRows) {
      numOfRows = pInput[i].numOfRows;
    }
  }
G
Ganlin Zhao 已提交
372 373
  for (int32_t i = 0; i < inputNum; ++i) {
    pInputData[i] = pInput[i].columnData;
374
    input[i] = pInputData[i]->pData + pInputData[i]->varmeta.offset[0];
375 376 377 378
    int32_t factor = 1;
    if (hasNcharCol && (GET_PARAM_TYPE(&pInput[i]) == TSDB_DATA_TYPE_VARCHAR)) {
      factor = TSDB_NCHAR_SIZE;
    }
379 380

    int32_t numOfNulls = getNumOfNullEntries(pInputData[i], pInput[i].numOfRows);
381
    if (pInput[i].numOfRows == 1) {
382
      inputLen += (pInputData[i]->varmeta.length - VARSTR_HEADER_SIZE) * factor * (numOfRows - numOfNulls);
383
    } else {
384
      inputLen += pInputData[i]->varmeta.length - (numOfRows - numOfNulls) * VARSTR_HEADER_SIZE;
385
    }
386 387
  }

388 389 390
  int32_t outputLen = inputLen + numOfRows * VARSTR_HEADER_SIZE;
  outputBuf = taosMemoryCalloc(outputLen, 1);
  char *output = outputBuf;
G
Ganlin Zhao 已提交
391 392

  bool hasNull = false;
393
  for (int32_t k = 0; k < numOfRows; ++k) {
G
Ganlin Zhao 已提交
394
    for (int32_t i = 0; i < inputNum; ++i) {
395 396
      if (colDataIsNull_s(pInputData[i], k)) {
        colDataAppendNULL(pOutputData, k);
G
Ganlin Zhao 已提交
397 398 399 400 401 402 403 404 405 406 407
        hasNull = true;
        break;
      }
    }

    if (hasNull) {
      continue;
    }

    int16_t dataLen = 0;
    for (int32_t i = 0; i < inputNum; ++i) {
408 409 410 411
      int32_t ret = concatCopyHelper(input[i], output, hasNcharCol, GET_PARAM_TYPE(&pInput[i]), &dataLen);
      if (ret != TSDB_CODE_SUCCESS) {
        return ret;
      }
412 413 414
      if (pInput[i].numOfRows != 1) {
        input[i] += varDataTLen(input[i]);
      }
G
Ganlin Zhao 已提交
415
    }
416
    varDataSetLen(output, dataLen);
417 418
    colDataAppend(pOutputData, k, output, false);
    output += varDataTLen(output);
G
Ganlin Zhao 已提交
419 420
  }

421
  pOutput->numOfRows = numOfRows;
422
  taosMemoryFree(input);
423
  taosMemoryFree(outputBuf);
G
Ganlin Zhao 已提交
424 425 426 427 428
  taosMemoryFree(pInputData);

  return TSDB_CODE_SUCCESS;
}

429

G
Ganlin Zhao 已提交
430 431 432 433 434 435 436
int32_t concatWsFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
  if (inputNum < 3 || inputNum > 9) { // concat accpet 3-9 input strings including the separator
    return TSDB_CODE_FAILED;
  }

  SColumnInfoData **pInputData = taosMemoryCalloc(inputNum, sizeof(SColumnInfoData *));
  SColumnInfoData *pOutputData = pOutput->columnData;
437
  char **input = taosMemoryCalloc(inputNum, POINTER_BYTES);
438
  char *outputBuf = NULL;
G
Ganlin Zhao 已提交
439

440
  int32_t inputLen = 0;
441
  int32_t numOfRows = 0;
442
  bool hasNcharCol = false;
443
  for (int32_t i = 1; i < inputNum; ++i) {
444 445
    int32_t type = GET_PARAM_TYPE(&pInput[i]);
    if (!IS_VAR_DATA_TYPE(GET_PARAM_TYPE(&pInput[i]))) {
446 447
      return TSDB_CODE_FAILED;
    }
448 449 450
    if (type == TSDB_DATA_TYPE_NCHAR) {
      hasNcharCol = true;
    }
451 452 453 454
    if (pInput[i].numOfRows > numOfRows) {
      numOfRows = pInput[i].numOfRows;
    }
  }
G
Ganlin Zhao 已提交
455 456
  for (int32_t i = 0; i < inputNum; ++i) {
    pInputData[i] = pInput[i].columnData;
457
    input[i] = pInputData[i]->pData + pInputData[i]->varmeta.offset[0];
458 459 460 461
    int32_t factor = 1;
    if (hasNcharCol && (GET_PARAM_TYPE(&pInput[i]) == TSDB_DATA_TYPE_VARCHAR)) {
      factor = TSDB_NCHAR_SIZE;
    }
462 463

    int32_t numOfNulls = getNumOfNullEntries(pInputData[i], pInput[i].numOfRows);
464 465
    if (i == 0) {
      // calculate required separator space
466
      inputLen += (pInputData[0]->varmeta.length - VARSTR_HEADER_SIZE) * (numOfRows - numOfNulls) * (inputNum - 2) * factor;
467
    } else if (pInput[i].numOfRows == 1) {
468
      inputLen += (pInputData[i]->varmeta.length - VARSTR_HEADER_SIZE) * (numOfRows - numOfNulls) * factor;
469
    } else {
470
      inputLen += pInputData[i]->varmeta.length - (numOfRows - numOfNulls) * VARSTR_HEADER_SIZE;
471 472 473
    }
  }

474 475 476
  int32_t outputLen = inputLen + numOfRows * VARSTR_HEADER_SIZE;
  outputBuf = taosMemoryCalloc(outputLen, 1);
  char *output = outputBuf;
G
Ganlin Zhao 已提交
477

478
  for (int32_t k = 0; k < numOfRows; ++k) {
479 480
    if (colDataIsNull_s(pInputData[0], k)) {
      colDataAppendNULL(pOutputData, k);
G
Ganlin Zhao 已提交
481 482 483 484 485
      continue;
    }

    int16_t dataLen = 0;
    for (int32_t i = 1; i < inputNum; ++i) {
486
      if (colDataIsNull_s(pInputData[i], k)) {
G
Ganlin Zhao 已提交
487 488 489
        continue;
      }

490 491 492 493 494
      int32_t ret = concatCopyHelper(input[i], output, hasNcharCol, GET_PARAM_TYPE(&pInput[i]), &dataLen);
      if (ret != TSDB_CODE_SUCCESS) {
        return ret;
      }

495 496 497
      if (pInput[i].numOfRows != 1) {
        input[i] += varDataTLen(input[i]);
      }
G
Ganlin Zhao 已提交
498 499 500

      if (i < inputNum - 1) {
        //insert the separator
501
        char *sep = pInputData[0]->pData;
502 503 504 505
        int32_t ret = concatCopyHelper(sep, output, hasNcharCol, GET_PARAM_TYPE(&pInput[0]), &dataLen);
        if (ret != TSDB_CODE_SUCCESS) {
          return ret;
        }
G
Ganlin Zhao 已提交
506 507
      }
    }
508
    varDataSetLen(output, dataLen);
509 510
    colDataAppend(pOutputData, k, output, false);
    output += varDataTLen(output);
G
Ganlin Zhao 已提交
511 512
  }

513 514
  pOutput->numOfRows = numOfRows;
  taosMemoryFree(input);
515
  taosMemoryFree(outputBuf);
G
Ganlin Zhao 已提交
516 517 518 519 520
  taosMemoryFree(pInputData);

  return TSDB_CODE_SUCCESS;
}

521
static int32_t doCaseConvFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput, _conv_fn convFn) {
G
Ganlin Zhao 已提交
522 523 524 525 526 527 528 529
  int32_t type = GET_PARAM_TYPE(pInput);
  if (inputNum != 1 || !IS_VAR_DATA_TYPE(type)) {
    return TSDB_CODE_FAILED;
  }

  SColumnInfoData *pInputData  = pInput->columnData;
  SColumnInfoData *pOutputData = pOutput->columnData;

530
  char *input  = pInputData->pData + pInputData->varmeta.offset[0];
531 532 533 534 535
  char *output = NULL;

  int32_t outputLen = pInputData->varmeta.length;
  char *outputBuf = taosMemoryCalloc(outputLen, 1);
  output = outputBuf;
536

G
Ganlin Zhao 已提交
537
  for (int32_t i = 0; i < pInput->numOfRows; ++i) {
538 539
    if (colDataIsNull_s(pInputData, i)) {
      colDataAppendNULL(pOutputData, i);
G
Ganlin Zhao 已提交
540 541 542
      continue;
    }

543
    int32_t len = varDataLen(input);
G
Ganlin Zhao 已提交
544 545
    if (type == TSDB_DATA_TYPE_VARCHAR) {
      for (int32_t j = 0; j < len; ++j) {
546
        *(varDataVal(output) + j) = convFn(*(varDataVal(input) + j));
G
Ganlin Zhao 已提交
547 548 549
      }
    } else { //NCHAR
      for (int32_t j = 0; j < len / TSDB_NCHAR_SIZE; ++j) {
550
        *((uint32_t *)varDataVal(output) + j) = convFn(*((uint32_t *)varDataVal(input) + j));
G
Ganlin Zhao 已提交
551 552
      }
    }
553
    varDataSetLen(output, len);
554
    colDataAppend(pOutputData, i, output, false);
555 556
    input += varDataTLen(input);
    output += varDataTLen(output);
G
Ganlin Zhao 已提交
557 558 559
  }

  pOutput->numOfRows = pInput->numOfRows;
560
  taosMemoryFree(outputBuf);
G
Ganlin Zhao 已提交
561 562 563 564 565

  return TSDB_CODE_SUCCESS;
}


566
static int32_t doTrimFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput, _trim_fn trimFn) {
G
Ganlin Zhao 已提交
567 568 569 570 571 572 573 574
  int32_t type = GET_PARAM_TYPE(pInput);
  if (inputNum != 1 || !IS_VAR_DATA_TYPE(type)) {
    return TSDB_CODE_FAILED;
  }

  SColumnInfoData *pInputData  = pInput->columnData;
  SColumnInfoData *pOutputData = pOutput->columnData;

575
  char *input  = pInputData->pData + pInputData->varmeta.offset[0];
576 577 578 579 580
  char *output = NULL;

  int32_t outputLen = pInputData->varmeta.length;
  char *outputBuf = taosMemoryCalloc(outputLen, 1);
  output = outputBuf;
581

G
Ganlin Zhao 已提交
582
  for (int32_t i = 0; i < pInput->numOfRows; ++i) {
583 584
    if (colDataIsNull_s(pInputData, i)) {
      colDataAppendNULL(pOutputData, i);
G
Ganlin Zhao 已提交
585 586 587
      continue;
    }

588
    int32_t len = varDataLen(input);
G
Ganlin Zhao 已提交
589
    int32_t charLen = (type == TSDB_DATA_TYPE_VARCHAR) ? len : len / TSDB_NCHAR_SIZE;
590 591 592
    trimFn(input, output, type, charLen);

    varDataSetLen(output, len);
593
    colDataAppend(pOutputData, i, output, false);
594 595
    input += varDataTLen(input);
    output += varDataTLen(output);
G
Ganlin Zhao 已提交
596 597 598
  }

  pOutput->numOfRows = pInput->numOfRows;
599
  taosMemoryFree(outputBuf);
G
Ganlin Zhao 已提交
600 601 602 603 604

  return TSDB_CODE_SUCCESS;
}

int32_t substrFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
605
  if (inputNum != 2 && inputNum!= 3) {
G
Ganlin Zhao 已提交
606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626
    return TSDB_CODE_FAILED;
  }

  int32_t subPos = 0;
  GET_TYPED_DATA(subPos, int32_t, GET_PARAM_TYPE(&pInput[1]), pInput[1].columnData->pData);
  if (subPos == 0) { //subPos needs to be positive or negative values;
    return TSDB_CODE_FAILED;
  }

  int32_t subLen = INT16_MAX;
  if (inputNum == 3) {
    GET_TYPED_DATA(subLen, int32_t, GET_PARAM_TYPE(&pInput[2]), pInput[2].columnData->pData);
    if (subLen < 0) { //subLen cannot be negative
      return TSDB_CODE_FAILED;
    }
    subLen = (GET_PARAM_TYPE(pInput) == TSDB_DATA_TYPE_VARCHAR) ? subLen : subLen * TSDB_NCHAR_SIZE;
  }

  SColumnInfoData *pInputData  = pInput->columnData;
  SColumnInfoData *pOutputData = pOutput->columnData;

627
  char *input  = pInputData->pData + pInputData->varmeta.offset[0];
628 629
  char *output = NULL;

630
  int32_t outputLen = pInputData->varmeta.length * pInput->numOfRows;
631 632
  char *outputBuf = taosMemoryCalloc(outputLen, 1);
  output = outputBuf;
633 634

  for (int32_t i = 0; i < pInput->numOfRows; ++i) {
635 636
    if (colDataIsNull_s(pInputData, i)) {
      colDataAppendNULL(pOutputData, i);
G
Ganlin Zhao 已提交
637 638 639
      continue;
    }

640
    int32_t len = varDataLen(input);
G
Ganlin Zhao 已提交
641 642 643 644 645 646 647 648 649 650
    int32_t startPosBytes;

    if (subPos > 0) {
      startPosBytes = (GET_PARAM_TYPE(pInput) == TSDB_DATA_TYPE_VARCHAR) ? subPos - 1 : (subPos - 1) * TSDB_NCHAR_SIZE;
      startPosBytes = MIN(startPosBytes, len);
    } else {
      startPosBytes = (GET_PARAM_TYPE(pInput) == TSDB_DATA_TYPE_VARCHAR) ? len + subPos : len + subPos * TSDB_NCHAR_SIZE;
      startPosBytes = MAX(startPosBytes, 0);
    }

651 652 653
    int32_t resLen = MIN(subLen, len - startPosBytes);
    if (resLen > 0) {
      memcpy(varDataVal(output), varDataVal(input) + startPosBytes, resLen);
G
Ganlin Zhao 已提交
654 655
    }

656
    varDataSetLen(output, resLen);
657
    colDataAppend(pOutputData, i , output, false);
658 659
    input += varDataTLen(input);
    output += varDataTLen(output);
G
Ganlin Zhao 已提交
660 661 662
  }

  pOutput->numOfRows = pInput->numOfRows;
663
  taosMemoryFree(outputBuf);
G
Ganlin Zhao 已提交
664 665 666 667

  return TSDB_CODE_SUCCESS;
}

G
Ganlin Zhao 已提交
668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820
int32_t castFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
  if (inputNum!= 3) {
    return TSDB_CODE_FAILED;
  }

  int16_t inputType  = pInput[0].columnData->info.type;
  int16_t outputType = *(int16_t *)pInput[1].columnData->pData;
  if (outputType != TSDB_DATA_TYPE_BIGINT && outputType != TSDB_DATA_TYPE_UBIGINT &&
      outputType != TSDB_DATA_TYPE_VARCHAR && outputType != TSDB_DATA_TYPE_NCHAR &&
      outputType != TSDB_DATA_TYPE_TIMESTAMP) {
    return TSDB_CODE_FAILED;
  }
  int64_t outputLen = *(int64_t *)pInput[2].columnData->pData;

  char *input = NULL;
  char *outputBuf = taosMemoryCalloc(outputLen * pInput[0].numOfRows, 1);
  char *output = outputBuf;
  if (IS_VAR_DATA_TYPE(inputType)) {
    input = pInput[0].columnData->pData + pInput[0].columnData->varmeta.offset[0];
  } else {
    input = pInput[0].columnData->pData;
  }

  for (int32_t i = 0; i < pInput[0].numOfRows; ++i) {
    if (colDataIsNull_s(pInput[0].columnData, i)) {
      colDataAppendNULL(pOutput->columnData, i);
      continue;
    }

    switch(outputType) {
      case TSDB_DATA_TYPE_BIGINT: {
        if (inputType == TSDB_DATA_TYPE_BINARY) {
          memcpy(output, varDataVal(input), varDataLen(input));
          *(int64_t *)output = strtoll(output, NULL, 10);
        } else if (inputType == TSDB_DATA_TYPE_NCHAR) {
          char *newBuf = taosMemoryCalloc(1, outputLen * TSDB_NCHAR_SIZE + 1);
          int32_t len  = taosUcs4ToMbs((TdUcs4 *)varDataVal(input), varDataLen(input), newBuf);
          if (len < 0) {
            taosMemoryFree(newBuf);
            return TSDB_CODE_FAILED;
          }
          newBuf[len] = 0;
          *(int64_t *)output = strtoll(newBuf, NULL, 10);
          taosMemoryFree(newBuf);
        } else {
          GET_TYPED_DATA(*(int64_t *)output, int64_t, inputType, input);
        }
        break;
      }
      case TSDB_DATA_TYPE_UBIGINT: {
        if (inputType == TSDB_DATA_TYPE_BINARY) {
          memcpy(output, varDataVal(input), varDataLen(input));
          *(uint64_t *)output = strtoull(output, NULL, 10);
        } else if (inputType == TSDB_DATA_TYPE_NCHAR) {
          char *newBuf = taosMemoryCalloc(1, outputLen * TSDB_NCHAR_SIZE + 1);
          int32_t len = taosUcs4ToMbs((TdUcs4 *)varDataVal(input), varDataLen(input), newBuf);
          if (len < 0) {
            taosMemoryFree(newBuf);
            return TSDB_CODE_FAILED;
          }
          newBuf[len] = 0;
          *(uint64_t *)output = strtoull(newBuf, NULL, 10);
          taosMemoryFree(newBuf);
        } else {
          GET_TYPED_DATA(*(uint64_t *)output, uint64_t, inputType, input);
        }
        break;
      }
      case TSDB_DATA_TYPE_TIMESTAMP: {
        if (inputType == TSDB_DATA_TYPE_BINARY || inputType == TSDB_DATA_TYPE_NCHAR) {
          //not support
          return TSDB_CODE_FAILED;
        } else {
          GET_TYPED_DATA(*(int64_t *)output, int64_t, inputType, input);
        }
        break;
      }
      case TSDB_DATA_TYPE_BINARY: {
        if (inputType == TSDB_DATA_TYPE_BOOL) {
          int32_t len = sprintf(varDataVal(output), "%.*s", (int32_t)(outputLen - VARSTR_HEADER_SIZE), *(int8_t *)input ? "true" : "false");
          varDataSetLen(output, len);
        } else if (inputType == TSDB_DATA_TYPE_BINARY) {
          int32_t len = sprintf(varDataVal(output), "%.*s", (int32_t)(outputLen - VARSTR_HEADER_SIZE), varDataVal(input));
          varDataSetLen(output, len);
        } else if (inputType == TSDB_DATA_TYPE_BINARY || inputType == TSDB_DATA_TYPE_NCHAR) {
          //not support
          return TSDB_CODE_FAILED;
        } else {
          char tmp[400] = {0};
          NUM_TO_STRING(inputType, input, sizeof(tmp), tmp);
          int32_t len = (int32_t)strlen(tmp);
          len = (outputLen - VARSTR_HEADER_SIZE) > len ? len : (outputLen - VARSTR_HEADER_SIZE);
          memcpy(varDataVal(output), tmp, len);
          varDataSetLen(output, len);
        }
        break;
      }
      case TSDB_DATA_TYPE_NCHAR: {
        int32_t outputCharLen = (outputLen - VARSTR_HEADER_SIZE) / TSDB_NCHAR_SIZE;
        if (inputType == TSDB_DATA_TYPE_BOOL) {
          char tmp[8] = {0};
          int32_t len = sprintf(tmp, "%.*s", outputCharLen, *(int8_t *)input ? "true" : "false" );
          bool ret = taosMbsToUcs4(tmp, len, (TdUcs4 *)varDataVal(output), outputLen - VARSTR_HEADER_SIZE, &len);
          if (!ret) {
            return TSDB_CODE_FAILED;
          }
          varDataSetLen(output, len);
        } else if (inputType == TSDB_DATA_TYPE_BINARY) {
          int32_t len = outputCharLen > varDataLen(input) ? varDataLen(input) : outputCharLen;
          bool ret = taosMbsToUcs4(input + VARSTR_HEADER_SIZE, len, (TdUcs4 *)varDataVal(output), outputLen - VARSTR_HEADER_SIZE, &len);
          if (!ret) {
            return TSDB_CODE_FAILED;
          }
          varDataSetLen(output, len);
        } else if (inputType == TSDB_DATA_TYPE_NCHAR) {
          int32_t len = MIN(outputLen, varDataLen(input) + VARSTR_HEADER_SIZE);
          memcpy(output, input, len);
          varDataSetLen(output, len - VARSTR_HEADER_SIZE);
        } else {
          char tmp[400] = {0};
          NUM_TO_STRING(inputType, input, sizeof(tmp), tmp);
          int32_t len = (int32_t)strlen(tmp);
          len = outputCharLen > len ? len : outputCharLen;
          bool ret = taosMbsToUcs4(tmp, len, (TdUcs4 *)varDataVal(output), outputLen - VARSTR_HEADER_SIZE, &len);
          if (!ret) {
            return TSDB_CODE_FAILED;
          }
          varDataSetLen(output, len);
        }
        break;
      }
      default: {
        return TSDB_CODE_FAILED;
      }
    }

    colDataAppend(pOutput->columnData, i, output, false);
    if (IS_VAR_DATA_TYPE(inputType)) {
      input  += varDataTLen(input);
    } else {
      input  += tDataTypes[inputType].bytes;
    }
    if (IS_VAR_DATA_TYPE(outputType)) {
      output += varDataTLen(output);
    } else {
      output += tDataTypes[outputType].bytes;
    }
  }

  pOutput->numOfRows = pInput->numOfRows;
  taosMemoryFree(outputBuf);
  return TSDB_CODE_SUCCESS;
}
G
Ganlin Zhao 已提交
821

822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860
int32_t toISO8601Function(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
  int32_t type = GET_PARAM_TYPE(pInput);
  if (type != TSDB_DATA_TYPE_BIGINT && type != TSDB_DATA_TYPE_TIMESTAMP) {
    return TSDB_CODE_FAILED;
  }

  if (inputNum != 1) {
    return TSDB_CODE_FAILED;
  }

  char *input  = pInput[0].columnData->pData;
  for (int32_t i = 0; i < pInput[0].numOfRows; ++i) {
    if (colDataIsNull_s(pInput[0].columnData, i)) {
      colDataAppendNULL(pOutput->columnData, i);
      continue;
    }

    char fraction[20] = {0};
    bool hasFraction = false;
    NUM_TO_STRING(type, input, sizeof(fraction), fraction);
    int32_t tsDigits = (int32_t)strlen(fraction);

    char buf[64] = {0};
    int64_t timeVal;
    GET_TYPED_DATA(timeVal, int64_t, type, input);
    if (tsDigits > TSDB_TIME_PRECISION_SEC_DIGITS) {
      if (tsDigits == TSDB_TIME_PRECISION_MILLI_DIGITS) {
        timeVal = timeVal / 1000;
      } else if (tsDigits == TSDB_TIME_PRECISION_MICRO_DIGITS) {
        timeVal = timeVal / (1000 * 1000);
      } else if (tsDigits == TSDB_TIME_PRECISION_NANO_DIGITS) {
        timeVal = timeVal / (1000 * 1000 * 1000);
      } else {
        assert(0);
      }
      hasFraction = true;
      memmove(fraction, fraction + TSDB_TIME_PRECISION_SEC_DIGITS, TSDB_TIME_PRECISION_SEC_DIGITS);
    }

861
    struct tm *tmInfo = taosLocalTime((const time_t *)&timeVal, NULL);
862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892
    strftime(buf, sizeof(buf), "%Y-%m-%dT%H:%M:%S%z", tmInfo);
    int32_t len = (int32_t)strlen(buf);

    if (hasFraction) {
      int32_t fracLen = (int32_t)strlen(fraction) + 1;
      char *tzInfo = strchr(buf, '+');
      if (tzInfo) {
        memmove(tzInfo + fracLen, tzInfo, strlen(tzInfo));
      } else {
        tzInfo = strchr(buf, '-');
        memmove(tzInfo + fracLen, tzInfo, strlen(tzInfo));
      }

      char tmp[32];
      sprintf(tmp, ".%s", fraction);
      memcpy(tzInfo, tmp, fracLen);
      len += fracLen;
    }

    memmove(buf + VARSTR_HEADER_SIZE, buf, len);
    varDataSetLen(buf, len);

    colDataAppend(pOutput->columnData, i, buf, false);
    input   += tDataTypes[type].bytes;
  }

  pOutput->numOfRows = pInput->numOfRows;

  return TSDB_CODE_SUCCESS;
}

893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922
int32_t toUnixtimestampFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
  int32_t type = GET_PARAM_TYPE(pInput);
  int32_t timePrec = GET_PARAM_PRECISON(pInput);
  if (type != TSDB_DATA_TYPE_BINARY && type != TSDB_DATA_TYPE_NCHAR) {
    return TSDB_CODE_FAILED;
  }

  if (inputNum != 1) {
    return TSDB_CODE_FAILED;
  }

  char *input = pInput[0].columnData->pData + pInput[0].columnData->varmeta.offset[0];
  for (int32_t i = 0; i < pInput[0].numOfRows; ++i) {
    if (colDataIsNull_s(pInput[0].columnData, i)) {
      colDataAppendNULL(pOutput->columnData, i);
      continue;
    }

    int64_t timeVal = 0;
    convertStringToTimestamp(type, input, timePrec, &timeVal);

    colDataAppend(pOutput->columnData, i, (char *)&timeVal, false);
    input += varDataTLen(input);
  }

  pOutput->numOfRows = pInput->numOfRows;

  return TSDB_CODE_SUCCESS;
}

923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123
int32_t timeTruncateFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
  int32_t type = GET_PARAM_TYPE(&pInput[0]);
  int32_t timePrec = GET_PARAM_PRECISON(&pInput[0]);
  if (inputNum != 2) {
    return TSDB_CODE_FAILED;
  }

  if (type != TSDB_DATA_TYPE_BIGINT && type != TSDB_DATA_TYPE_TIMESTAMP &&
      type != TSDB_DATA_TYPE_BINARY && type != TSDB_DATA_TYPE_NCHAR) {
    return TSDB_CODE_FAILED;
  }

  if (GET_PARAM_TYPE(&pInput[1]) != TSDB_DATA_TYPE_BIGINT) { //time_unit
    return TSDB_CODE_FAILED;
  }

  int64_t timeUnit, timeVal = 0;
  GET_TYPED_DATA(timeUnit, int64_t, GET_PARAM_TYPE(&pInput[1]), pInput[1].columnData->pData);

  int64_t factor = (timePrec == TSDB_TIME_PRECISION_MILLI) ? 1000 :
                   (timePrec == TSDB_TIME_PRECISION_MICRO ? 1000000 : 1000000000);

  char *input = NULL;
  if (IS_VAR_DATA_TYPE(type)) {
    input = pInput[0].columnData->pData + pInput[0].columnData->varmeta.offset[0];
  } else {
    input = pInput[0].columnData->pData;
  }

  for (int32_t i = 0; i < pInput[0].numOfRows; ++i) {
    if (colDataIsNull_s(pInput[0].columnData, i)) {
      colDataAppendNULL(pOutput->columnData, i);
      continue;
    }

    if (IS_VAR_DATA_TYPE(type)) { /* datetime format strings */
      convertStringToTimestamp(type, input, TSDB_TIME_PRECISION_NANO, &timeVal);
      //If converted value is less than 10digits in second, use value in second instead
      int64_t timeValSec = timeVal / 1000000000;
      if (timeValSec < 1000000000) {
        timeVal = timeValSec;
      }
    } else if (type == TSDB_DATA_TYPE_BIGINT) { /* unix timestamp */
      GET_TYPED_DATA(timeVal, int64_t, type, input);
    } else if (type == TSDB_DATA_TYPE_TIMESTAMP) { /* timestamp column*/
      GET_TYPED_DATA(timeVal, int64_t, type, input);
      int64_t timeValSec = timeVal / factor;
      if (timeValSec < 1000000000) {
        timeVal = timeValSec;
      }
    }

    char buf[20] = {0};
    NUM_TO_STRING(TSDB_DATA_TYPE_BIGINT, &timeVal, sizeof(buf), buf);
    int32_t tsDigits = (int32_t)strlen(buf);
    timeUnit = timeUnit * 1000 / factor;

    switch (timeUnit) {
      case 0: { /* 1u */
        if (tsDigits == TSDB_TIME_PRECISION_NANO_DIGITS) {
          timeVal = timeVal / 1000 * 1000;
        //} else if (tsDigits == TSDB_TIME_PRECISION_MICRO_DIGITS) {
        //  //timeVal = timeVal / 1000;
        } else if (tsDigits <= TSDB_TIME_PRECISION_SEC_DIGITS) {
          timeVal = timeVal * factor;
        } else {
          timeVal = timeVal * 1;
        }
        break;
      }
      case 1: { /* 1a */
        if (tsDigits == TSDB_TIME_PRECISION_MILLI_DIGITS) {
          timeVal = timeVal * 1;
        } else if (tsDigits == TSDB_TIME_PRECISION_MICRO_DIGITS) {
          timeVal = timeVal / 1000 * 1000;
        } else if (tsDigits == TSDB_TIME_PRECISION_NANO_DIGITS) {
          timeVal = timeVal / 1000000 * 1000000;
        } else if (tsDigits <= TSDB_TIME_PRECISION_SEC_DIGITS){
          timeVal = timeVal * factor;
        } else {
          assert(0);
        }
        break;
      }
      case 1000: { /* 1s */
        if (tsDigits == TSDB_TIME_PRECISION_MILLI_DIGITS) {
          timeVal = timeVal / 1000 * 1000;
        } else if (tsDigits == TSDB_TIME_PRECISION_MICRO_DIGITS) {
          timeVal = timeVal / 1000000 * 1000000;
        } else if (tsDigits == TSDB_TIME_PRECISION_NANO_DIGITS) {
          timeVal = timeVal / 1000000000 * 1000000000;
        } else if (tsDigits <= TSDB_TIME_PRECISION_SEC_DIGITS) {
          timeVal = timeVal * factor;
        } else {
          assert(0);
        }
        break;
      }
      case 60000: { /* 1m */
        if (tsDigits == TSDB_TIME_PRECISION_MILLI_DIGITS) {
          timeVal = timeVal / 1000 / 60 * 60 * 1000;
        } else if (tsDigits == TSDB_TIME_PRECISION_MICRO_DIGITS) {
          timeVal = timeVal / 1000000 / 60 * 60 * 1000000;
        } else if (tsDigits == TSDB_TIME_PRECISION_NANO_DIGITS) {
          timeVal = timeVal / 1000000000 / 60 * 60 * 1000000000;
        } else if (tsDigits <= TSDB_TIME_PRECISION_SEC_DIGITS) {
          timeVal = timeVal * factor / factor / 60 * 60 * factor;
        } else {
          assert(0);
        }
        break;
      }
      case 3600000: { /* 1h */
        if (tsDigits == TSDB_TIME_PRECISION_MILLI_DIGITS) {
          timeVal = timeVal / 1000 / 3600 * 3600 * 1000;
        } else if (tsDigits == TSDB_TIME_PRECISION_MICRO_DIGITS) {
          timeVal = timeVal / 1000000 / 3600 * 3600 * 1000000;
        } else if (tsDigits == TSDB_TIME_PRECISION_NANO_DIGITS) {
          timeVal = timeVal / 1000000000 / 3600 * 3600 * 1000000000;
        } else if (tsDigits <= TSDB_TIME_PRECISION_SEC_DIGITS) {
          timeVal = timeVal * factor / factor / 3600 * 3600 * factor;
        } else {
          assert(0);
        }
        break;
      }
      case 86400000: { /* 1d */
        if (tsDigits == TSDB_TIME_PRECISION_MILLI_DIGITS) {
          timeVal = timeVal / 1000 / 86400 * 86400 * 1000;
        } else if (tsDigits == TSDB_TIME_PRECISION_MICRO_DIGITS) {
          timeVal = timeVal / 1000000 / 86400 * 86400 * 1000000;
        } else if (tsDigits == TSDB_TIME_PRECISION_NANO_DIGITS) {
          timeVal = timeVal / 1000000000 / 86400 * 86400 * 1000000000;
        } else if (tsDigits <= TSDB_TIME_PRECISION_SEC_DIGITS) {
          timeVal = timeVal * factor / factor / 86400* 86400 * factor;
        } else {
          assert(0);
        }
        break;
      }
      case 604800000: { /* 1w */
        if (tsDigits == TSDB_TIME_PRECISION_MILLI_DIGITS) {
          timeVal = timeVal / 1000 / 604800 * 604800 * 1000;
        } else if (tsDigits == TSDB_TIME_PRECISION_MICRO_DIGITS) {
          timeVal = timeVal / 1000000 / 604800 * 604800 * 1000000;
        } else if (tsDigits == TSDB_TIME_PRECISION_NANO_DIGITS) {
          timeVal = timeVal / 1000000000 / 604800 * 604800 * 1000000000;
        } else if (tsDigits <= TSDB_TIME_PRECISION_SEC_DIGITS) {
          timeVal = timeVal * factor / factor / 604800 * 604800* factor;
        } else {
          assert(0);
        }
        break;
      }
      default: {
        timeVal = timeVal * 1;
        break;
      }
    }

    //truncate the timestamp to db precision
    switch (timePrec) {
      case TSDB_TIME_PRECISION_MILLI: {
        if (tsDigits == TSDB_TIME_PRECISION_MICRO_DIGITS) {
          timeVal = timeVal / 1000;
        } else if (tsDigits == TSDB_TIME_PRECISION_NANO_DIGITS) {
          timeVal = timeVal / 1000000;
        }
        break;
      }
      case TSDB_TIME_PRECISION_MICRO: {
        if (tsDigits == TSDB_TIME_PRECISION_NANO_DIGITS) {
          timeVal = timeVal / 1000;
        } else if (tsDigits == TSDB_TIME_PRECISION_MILLI_DIGITS) {
          timeVal = timeVal * 1000;
        }
        break;
      }
      case TSDB_TIME_PRECISION_NANO: {
        if (tsDigits == TSDB_TIME_PRECISION_MICRO_DIGITS) {
          timeVal = timeVal * 1000;
        } else if (tsDigits == TSDB_TIME_PRECISION_MILLI_DIGITS) {
          timeVal = timeVal * 1000000;
        }
        break;
      }
    }

    colDataAppend(pOutput->columnData, i, (char *)&timeVal, false);
    if (IS_VAR_DATA_TYPE(type)) {
      input += varDataTLen(input);
    } else {
      input += tDataTypes[type].bytes;
    }
  }

  pOutput->numOfRows = pInput->numOfRows;

  return TSDB_CODE_SUCCESS;
}

G
Ganlin Zhao 已提交
1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187
int32_t timeDiffFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
  if (inputNum != 2 && inputNum != 3) {
    return TSDB_CODE_FAILED;
  }

  int32_t timePrec = GET_PARAM_PRECISON(&pInput[0]);
  int64_t timeUnit = -1, timeVal[2] = {0};
  if (inputNum == 3) {
    if (GET_PARAM_TYPE(&pInput[2]) != TSDB_DATA_TYPE_BIGINT) {
      return TSDB_CODE_FAILED;
    }
    GET_TYPED_DATA(timeUnit, int64_t, GET_PARAM_TYPE(&pInput[2]), pInput[2].columnData->pData);
  }

  char *input[2];
  for (int32_t k = 0; k < 2; ++k) {
    int32_t type = GET_PARAM_TYPE(&pInput[k]);
    if (type != TSDB_DATA_TYPE_BIGINT && type != TSDB_DATA_TYPE_TIMESTAMP &&
        type != TSDB_DATA_TYPE_BINARY && type != TSDB_DATA_TYPE_NCHAR) {
      return TSDB_CODE_FAILED;
    }

    if (IS_VAR_DATA_TYPE(type)) {
      input[k] = pInput[k].columnData->pData + pInput[k].columnData->varmeta.offset[0];
    } else {
      input[k] = pInput[k].columnData->pData;
    }
  }

  for (int32_t i = 0; i < pInput[0].numOfRows; ++i) {
    for (int32_t k = 0; k < 2; ++k) {
      if (colDataIsNull_s(pInput[0].columnData, i)) {
        colDataAppendNULL(pOutput->columnData, i);
        continue;
      }

      int32_t type = GET_PARAM_TYPE(&pInput[k]);
      if (IS_VAR_DATA_TYPE(type)) { /* datetime format strings */
        convertStringToTimestamp(type, input[k], TSDB_TIME_PRECISION_NANO, &timeVal[k]);
      } else if (type == TSDB_DATA_TYPE_BIGINT || type == TSDB_DATA_TYPE_TIMESTAMP) { /* unix timestamp or ts column*/
        GET_TYPED_DATA(timeVal[k], int64_t, type, input[k]);
        if (type == TSDB_DATA_TYPE_TIMESTAMP) {
          int64_t factor = (timePrec == TSDB_TIME_PRECISION_MILLI) ? 1000 :
                           (timePrec == TSDB_TIME_PRECISION_MICRO ? 1000000 : 1000000000);
          int64_t timeValSec = timeVal[k] / factor;
          if (timeValSec < 1000000000) {
            timeVal[k] = timeValSec;
          }
        }

        char buf[20] = {0};
        NUM_TO_STRING(TSDB_DATA_TYPE_BIGINT, &timeVal[k], sizeof(buf), buf);
        int32_t tsDigits = (int32_t)strlen(buf);
        if (tsDigits <= TSDB_TIME_PRECISION_SEC_DIGITS) {
          timeVal[k] = timeVal[k] * 1000000000;
        } else if (tsDigits == TSDB_TIME_PRECISION_MILLI_DIGITS) {
          timeVal[k] = timeVal[k] * 1000000;
        } else if (tsDigits == TSDB_TIME_PRECISION_MICRO_DIGITS) {
          timeVal[k] = timeVal[k] * 1000;
        } else if (tsDigits == TSDB_TIME_PRECISION_NANO_DIGITS) {
          timeVal[k] = timeVal[k];
        }
      }

G
Ganlin Zhao 已提交
1188 1189 1190 1191 1192 1193
      if (pInput[k].numOfRows != 1) {
        if (IS_VAR_DATA_TYPE(type)) {
          input[k] += varDataTLen(input[k]);
        } else {
          input[k] += tDataTypes[type].bytes;
        }
G
Ganlin Zhao 已提交
1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261
      }
    }

    int64_t result = (timeVal[0] >= timeVal[1]) ? (timeVal[0] - timeVal[1]) :
                                                  (timeVal[1] - timeVal[0]);

    if (timeUnit < 0) { // if no time unit given use db precision
      switch(timePrec) {
        case TSDB_TIME_PRECISION_MILLI: {
          result = result / 1000000;
          break;
        }
        case TSDB_TIME_PRECISION_MICRO: {
          result = result / 1000;
          break;
        }
        case TSDB_TIME_PRECISION_NANO: {
          result = result / 1;
          break;
        }
      }
    } else {
      int64_t factor = (timePrec == TSDB_TIME_PRECISION_MILLI) ? 1000 :
                       (timePrec == TSDB_TIME_PRECISION_MICRO ? 1000000 : 1000000000);
      timeUnit = timeUnit * 1000 / factor;
      switch(timeUnit) {
        case 0: { /* 1u */
          result = result / 1000;
          break;
        }
        case 1: { /* 1a */
          result = result / 1000000;
          break;
        }
        case 1000: { /* 1s */
          result = result / 1000000000;
          break;
        }
        case 60000: { /* 1m */
          result = result / 1000000000 / 60;
          break;
        }
        case 3600000: { /* 1h */
          result = result / 1000000000 / 3600;
          break;
        }
        case 86400000: { /* 1d */
          result = result / 1000000000 / 86400;
          break;
        }
        case 604800000: { /* 1w */
          result = result / 1000000000 / 604800;
          break;
        }
        default: {
          break;
        }
      }
    }

    colDataAppend(pOutput->columnData, i, (char *)&result, false);
  }

  pOutput->numOfRows = pInput->numOfRows;

  return TSDB_CODE_SUCCESS;
}

1262 1263 1264
int32_t atanFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
  return doScalarFunctionUnique(pInput, inputNum, pOutput, atan);
}
1265

1266 1267 1268
int32_t sinFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
  return doScalarFunctionUnique(pInput, inputNum, pOutput, sin);
}
1269

1270 1271 1272
int32_t cosFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
  return doScalarFunctionUnique(pInput, inputNum, pOutput, cos);
}
1273

1274 1275 1276
int32_t tanFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
  return doScalarFunctionUnique(pInput, inputNum, pOutput, tan);
}
1277

1278 1279 1280
int32_t asinFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
  return doScalarFunctionUnique(pInput, inputNum, pOutput, asin);
}
1281

1282 1283 1284
int32_t acosFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
  return doScalarFunctionUnique(pInput, inputNum, pOutput, acos);
}
H
Haojun Liao 已提交
1285

G
Ganlin Zhao 已提交
1286 1287 1288 1289 1290 1291 1292 1293
int32_t powFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
  return doScalarFunctionUnique2(pInput, inputNum, pOutput, pow);
}

int32_t logFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
  return doScalarFunctionUnique2(pInput, inputNum, pOutput, tlog);
}

1294 1295 1296
int32_t sqrtFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
  return doScalarFunctionUnique(pInput, inputNum, pOutput, sqrt);
}
1297

1298 1299 1300 1301 1302 1303 1304 1305 1306 1307
int32_t ceilFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
  return doScalarFunction(pInput, inputNum, pOutput, ceilf, ceil);
}

int32_t floorFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
  return doScalarFunction(pInput, inputNum, pOutput, floorf, floor);
}

int32_t roundFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
  return doScalarFunction(pInput, inputNum, pOutput, roundf, round);
1308 1309
}

G
Ganlin Zhao 已提交
1310 1311
int32_t lowerFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
  return doCaseConvFunction(pInput, inputNum, pOutput, tolower);
1312 1313
}

G
Ganlin Zhao 已提交
1314 1315
int32_t upperFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
  return doCaseConvFunction(pInput, inputNum, pOutput, toupper);
1316 1317
}

G
Ganlin Zhao 已提交
1318 1319
int32_t ltrimFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
  return doTrimFunction(pInput, inputNum, pOutput, tltrim);
1320 1321
}

G
Ganlin Zhao 已提交
1322 1323
int32_t rtrimFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
  return doTrimFunction(pInput, inputNum, pOutput, trtrim);
1324 1325
}

G
Ganlin Zhao 已提交
1326 1327 1328
int32_t lengthFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
  return doLengthFunction(pInput, inputNum, pOutput, tlength);
}
1329

G
Ganlin Zhao 已提交
1330 1331
int32_t charLengthFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
  return doLengthFunction(pInput, inputNum, pOutput, tcharlength);
1332 1333
}

H
Haojun Liao 已提交
1334
#if 0
1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398
static void reverseCopy(char* dest, const char* src, int16_t type, int32_t numOfRows) {
  switch(type) {
    case TSDB_DATA_TYPE_TINYINT:
    case TSDB_DATA_TYPE_UTINYINT:{
      int8_t* p = (int8_t*) dest;
      int8_t* pSrc = (int8_t*) src;

      for(int32_t i = 0; i < numOfRows; ++i) {
        p[i] = pSrc[numOfRows - i - 1];
      }
      return;
    }

    case TSDB_DATA_TYPE_SMALLINT:
    case TSDB_DATA_TYPE_USMALLINT:{
      int16_t* p = (int16_t*) dest;
      int16_t* pSrc = (int16_t*) src;

      for(int32_t i = 0; i < numOfRows; ++i) {
        p[i] = pSrc[numOfRows - i - 1];
      }
      return;
    }
    case TSDB_DATA_TYPE_INT:
    case TSDB_DATA_TYPE_UINT: {
      int32_t* p = (int32_t*) dest;
      int32_t* pSrc = (int32_t*) src;

      for(int32_t i = 0; i < numOfRows; ++i) {
        p[i] = pSrc[numOfRows - i - 1];
      }
      return;
    }
    case TSDB_DATA_TYPE_BIGINT:
    case TSDB_DATA_TYPE_UBIGINT: {
      int64_t* p = (int64_t*) dest;
      int64_t* pSrc = (int64_t*) src;

      for(int32_t i = 0; i < numOfRows; ++i) {
        p[i] = pSrc[numOfRows - i - 1];
      }
      return;
    }
    case TSDB_DATA_TYPE_FLOAT: {
      float* p = (float*) dest;
      float* pSrc = (float*) src;

      for(int32_t i = 0; i < numOfRows; ++i) {
        p[i] = pSrc[numOfRows - i - 1];
      }
      return;
    }
    case TSDB_DATA_TYPE_DOUBLE: {
      double* p = (double*) dest;
      double* pSrc = (double*) src;

      for(int32_t i = 0; i < numOfRows; ++i) {
        p[i] = pSrc[numOfRows - i - 1];
      }
      return;
    }
    default: assert(0);
  }
}
H
Haojun Liao 已提交
1399
#endif
1400

1401 1402 1403 1404 1405 1406 1407 1408
bool getTimePseudoFuncEnv(SFunctionNode* UNUSED_PARAM(pFunc), SFuncExecEnv* pEnv) {
  pEnv->calcMemSize = sizeof(int64_t);
  return true;
}

int32_t qStartTsFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
  ASSERT(inputNum == 1);
  colDataAppendInt64(pOutput->columnData, pOutput->numOfRows, (int64_t *)colDataGetData(pInput->columnData, 0));
1409
  return TSDB_CODE_SUCCESS;
1410 1411 1412 1413 1414
}

int32_t qEndTsFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
  ASSERT(inputNum == 1);
  colDataAppendInt64(pOutput->columnData, pOutput->numOfRows, (int64_t *)colDataGetData(pInput->columnData, 1));
1415
  return TSDB_CODE_SUCCESS;
1416 1417 1418 1419 1420
}

int32_t winDurFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
  ASSERT(inputNum == 1);
  colDataAppendInt64(pOutput->columnData, pOutput->numOfRows, (int64_t *)colDataGetData(pInput->columnData, 2));
1421
  return TSDB_CODE_SUCCESS;
1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433
}

int32_t winStartTsFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
  ASSERT(inputNum == 1);
  colDataAppendInt64(pOutput->columnData, pOutput->numOfRows, (int64_t*) colDataGetData(pInput->columnData, 3));
  return TSDB_CODE_SUCCESS;
}

int32_t winEndTsFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
  ASSERT(inputNum == 1);
  colDataAppendInt64(pOutput->columnData, pOutput->numOfRows, (int64_t*) colDataGetData(pInput->columnData, 4));
  return TSDB_CODE_SUCCESS;
1434
}