qingzhu521 opened a new issue, #2089:
URL: https://github.com/apache/orc/issues/2089
#include "orc/Exceptions.hh"
#include "orc/OrcFile.hh"
#include "orc/Type.hh"
#include "Timezone.hh"
#include <algorithm>
#include <fstream>
#include <iostream>
#include <list>
#include <memory>
#include <getopt.h>
#include <string>
#include <sys/time.h>
#include <time.h>
using namespace std;
int main() {
string schema =
"struct<c0:smallint,c1:timestamp,c2:smallint,c3:double,c4:double,c5:varchar(500),date_part:varchar(20)>";
ORC_UNIQUE_PTR<orc::Type> fileType =
orc::Type::buildTypeFromString(schema);
ORC_UNIQUE_PTR<orc::OutputStream> outStream =
orc::writeLocalFile("test.orc");
orc::WriterOptions options;
options.setStripeSize(0);
ORC_UNIQUE_PTR<orc::Writer> writer =
orc::createWriter(*fileType, outStream.get(), options);
ORC_UNIQUE_PTR<orc::ColumnVectorBatch> rowBatch =
writer->createRowBatch(2);
/*
| 2629 | NULL | -1984 | 38270028153.095 | -3.337680 |
7<pw4T8Cj*_VmB4sDEFN3'O\\ilL]2m*L[Q+I6+K8.Xta!H?L!L_e-+8A7Ly#(oB.J9D8aWr?*ydB'[r3&STBt'+THNsq
| 2024121003 |
| -8130 | NULL | -22563 | 5399.785 | -1.200099 | rY<IWC+cgsgMt7
b[6('t1wPZ2t
| 2024121005 |
*/
orc::StructVectorBatch* root =
static_cast<orc::StructVectorBatch*>(rowBatch.get());
orc::LongVectorBatch* longBatch =
dynamic_cast<orc::LongVectorBatch*>(root->fields[0]);
longBatch->notNull[0] = true;
longBatch->data[0] = 2629;
longBatch->notNull[1] = true;
longBatch->data[1] = -8130;
longBatch->hasNulls = false;
longBatch->numElements = 2;
orc::TimestampVectorBatch* timestampBatch =
dynamic_cast<orc::TimestampVectorBatch*>(root->fields[1]);
timestampBatch->notNull[0] = false;
timestampBatch->notNull[1] = false;
timestampBatch->hasNulls = true;
timestampBatch->numElements = 2;
orc::LongVectorBatch* longBatch2 =
dynamic_cast<orc::LongVectorBatch*>(root->fields[2]);
longBatch2->notNull[0] = true;
longBatch2->data[0] = -1984;
longBatch2->notNull[1] = true;
longBatch2->data[1] = -22563;
longBatch2->hasNulls = false;
longBatch2->numElements = 2;
orc::DoubleVectorBatch* doubleBatch =
dynamic_cast<orc::DoubleVectorBatch*>(root->fields[3]);
doubleBatch->notNull[0] = true;
doubleBatch->data[0] = 38270028153.095;
doubleBatch->notNull[1] = true;
doubleBatch->data[1] = 5399.785;
doubleBatch->hasNulls = false;
doubleBatch->numElements = 2;
orc::DoubleVectorBatch* doubleBatch2 =
dynamic_cast<orc::DoubleVectorBatch*>(root->fields[4]);
doubleBatch2->notNull[0] = true;
doubleBatch2->data[0] = -3.337680;
doubleBatch2->notNull[1] = true;
doubleBatch2->data[1] = -1.200099;
doubleBatch2->hasNulls = false;
doubleBatch2->numElements = 2;
orc::StringVectorBatch* stringBatch =
dynamic_cast<orc::StringVectorBatch*>(root->fields[5]);
stringBatch->notNull[0] = true;
stringBatch->length[0] = 93;
stringBatch->data[0] = (char*)
"<pw4T8Cj*_VmB4sDEFN3'O\\ilL]2m*L[Q+I6+K8.Xta!H?L!L_e-+8A7Ly#(oB.J9D8aWr?*ydB'[r3&STBt'+THNsq";
stringBatch->notNull[1] = true;
stringBatch->length[1] = 28;
stringBatch->data[1] = (char*)"rY<IWC+cgsgMt7 b[6('t1wPZ2t";
stringBatch->hasNulls = false;
stringBatch->numElements = 2;
orc::StringVectorBatch* stringBatch2
=dynamic_cast<orc::StringVectorBatch*>(root->fields[6]);
stringBatch2->notNull[0] = true;
stringBatch2->data[0] = (char*)"2024121003";
stringBatch2->length[0] = 10;
stringBatch2->notNull[1] = true;
stringBatch2->data[1] = (char*)"2024121005";
stringBatch2->hasNulls = false;
stringBatch2->length[1] = 10;
stringBatch2->numElements = 2;
root->numElements = 2;
// writer->add(*root);
/*
| -26701 | NULL | -22098 | 646836227.275 | 6.786988 | z
| 2024121001 |
| -15326 | NULL | -26701 | -28089588.567 | -2.322138 | NULL
| 2024121002 |
*/
longBatch->notNull[0] = true;
longBatch->data[0] = -26701;
longBatch->notNull[1] = true;
longBatch->data[1] = -15326;
longBatch->hasNulls = false;
longBatch->numElements = 2;
timestampBatch->notNull[0] = false;
timestampBatch->notNull[1] = false;
timestampBatch->hasNulls = true;
timestampBatch->numElements = 2;
longBatch2->notNull[0] = true;
longBatch2->data[0] = -22098;
longBatch2->notNull[1] = true;
longBatch2->data[1] = -26701;
longBatch2->hasNulls = false;
longBatch2->numElements = 2;
doubleBatch->notNull[0] = true;
doubleBatch->data[0] = 646836227.275;
doubleBatch->notNull[1] = true;
doubleBatch->data[1] = -28089588.567;
doubleBatch->hasNulls = false;
doubleBatch->numElements = 2;
doubleBatch2->notNull[0] = true;
doubleBatch2->data[0] = 6.786988;
doubleBatch2->notNull[1] = true;
doubleBatch2->data[1] = -2.322138;
doubleBatch2->hasNulls = false;
doubleBatch2->numElements = 2;
stringBatch->notNull[0] = true;
stringBatch->data[0] = (char*)"a";
stringBatch->length[0] = 1;
stringBatch->notNull[1] = false;
// stringBatch->data[1] = NULL;
// stringBatch->length[1] = 0;
stringBatch->hasNulls = true;
stringBatch->numElements = 1;
stringBatch2->notNull[0] = true;
stringBatch2->data[0] = (char*)"2024121001";
stringBatch2->length[0] = 10;
stringBatch2->notNull[1] = true;
stringBatch2->data[1] = (char*)"2024121002";
stringBatch2->length[1] = 10;
stringBatch2->hasNulls = false;
stringBatch2->numElements = 2;
root->numElements = 2;
writer->add(*root);
writer->close();
return 0;
}
produce a file. If we read will report Caught exception in test.orc: bad
read in RleDecoderV2::readByte
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
