Examples with BytesColumnVector org.apache.hadoop.hive.ql.exec.vector.BytesColumnVector used on opensource projects

Example 21 with BytesColumnVector

use of org.apache.hadoop.hive.ql.exec.vector.BytesColumnVector in project hive by apache.

the class TestVectorStringExpressions method testAssignBytesColumnVector.

@Test
public // Test basic assign to vector.
void testAssignBytesColumnVector() {
    BytesColumnVector outV = new BytesColumnVector(VectorizedRowBatch.DEFAULT_SIZE);
    // initialize with estimated element size 35
    outV.initBuffer(35);
    int i = 0;
    int expectedResultLen;
    Assert.assertTrue(StringExpr.characterCount(blue, 0, blue.length) == 4);
    StringExpr.assign(outV, i, blue, 0, blue.length);
    expectedResultLen = blue.length;
    Assert.assertTrue(vectorEqual(outV, i, blue, 0, expectedResultLen));
    i++;
    Assert.assertTrue(StringExpr.characterCount(redgreen, 0, redgreen.length) == 8);
    StringExpr.assign(outV, i, redgreen, 0, redgreen.length);
    expectedResultLen = redgreen.length;
    Assert.assertTrue(vectorEqual(outV, i, redgreen, 0, expectedResultLen));
    i++;
    Assert.assertTrue(StringExpr.characterCount(ascii_sentence, 0, ascii_sentence.length) == 31);
    StringExpr.assign(outV, i, ascii_sentence, 0, ascii_sentence.length);
    expectedResultLen = ascii_sentence.length;
    Assert.assertTrue(vectorEqual(outV, i, ascii_sentence, 0, expectedResultLen));
    i++;
    Assert.assertTrue(StringExpr.characterCount(blanksLeft, 0, blanksLeft.length) == 5);
    StringExpr.assign(outV, i, blanksLeft, 0, blanksLeft.length);
    expectedResultLen = blanksLeft.length;
    Assert.assertTrue(vectorEqual(outV, i, blanksLeft, 0, expectedResultLen));
    i++;
    // Multi-byte characters with blank ranges.
    byte[] sentenceBlankRanges = new byte[100];
    int sentenceBlankRangesLen = addMultiByteCharSentenceBlankRanges(sentenceBlankRanges, 0);
    Assert.assertTrue(StringExpr.characterCount(sentenceBlankRanges, 0, sentenceBlankRangesLen) == 17);
    StringExpr.assign(outV, i, sentenceBlankRanges, 0, sentenceBlankRangesLen);
    expectedResultLen = sentenceBlankRangesLen;
    Assert.assertTrue(vectorEqual(outV, i, sentenceBlankRanges, 0, expectedResultLen));
    i++;
    Assert.assertTrue(StringExpr.characterCount(sentenceBlankRanges, 0, sentenceBlankRangesLen - 3) == 16);
    StringExpr.assign(outV, i, sentenceBlankRanges, 0, sentenceBlankRangesLen - 3);
    expectedResultLen = sentenceBlankRangesLen - 3;
    Assert.assertTrue(vectorEqual(outV, i, sentenceBlankRanges, 0, expectedResultLen));
    i++;
    // Some non-zero offsets.
    Assert.assertTrue(StringExpr.characterCount(sentenceBlankRanges, 4, sentenceBlankRangesLen - 4) == 16);
    StringExpr.assign(outV, i, sentenceBlankRanges, 4, sentenceBlankRangesLen - 4);
    expectedResultLen = sentenceBlankRangesLen - 4;
    Assert.assertTrue(vectorEqual(outV, i, sentenceBlankRanges, 4, expectedResultLen));
    Assert.assertTrue(vectorCharacterCount(outV, i) == 16);
    i++;
    Assert.assertTrue(StringExpr.characterCount(sentenceBlankRanges, 7, 17) == 13);
    StringExpr.assign(outV, i, sentenceBlankRanges, 7, 17);
    expectedResultLen = 17;
    Assert.assertTrue(vectorEqual(outV, i, sentenceBlankRanges, 7, expectedResultLen));
    Assert.assertTrue(vectorCharacterCount(outV, i) == 13);
    i++;
}

Example 22 with BytesColumnVector

use of org.apache.hadoop.hive.ql.exec.vector.BytesColumnVector in project hive by apache.

the class TestVectorStringExpressions method makeTrimBatch.

// Make a batch to test the trim functions.
private VectorizedRowBatch makeTrimBatch() {
    VectorizedRowBatch b = new VectorizedRowBatch(2);
    BytesColumnVector inV = new BytesColumnVector();
    BytesColumnVector outV = new BytesColumnVector();
    b.cols[0] = inV;
    b.cols[1] = outV;
    inV.setRef(0, emptyString, 0, 0);
    inV.setRef(1, blanksLeft, 0, blanksLeft.length);
    inV.setRef(2, blanksRight, 0, blanksRight.length);
    inV.setRef(3, blanksBoth, 0, blanksBoth.length);
    inV.setRef(4, red, 0, red.length);
    inV.setRef(5, blankString, 0, blankString.length);
    b.size = 5;
    return b;
}

Example 23 with BytesColumnVector

use of org.apache.hadoop.hive.ql.exec.vector.BytesColumnVector in project hive by apache.

the class TestVectorStringExpressions method testLoadBytesColumnVectorByRef.

@Test
public // set values by reference, copy the data out, and verify equality
void testLoadBytesColumnVectorByRef() {
    BytesColumnVector bcv = new BytesColumnVector(VectorizedRowBatch.DEFAULT_SIZE);
    String s = "red";
    byte[] b = null;
    try {
        b = s.getBytes("UTF-8");
    } catch (UnsupportedEncodingException e) {
        e.printStackTrace();
    }
    for (int i = 0; i != VectorizedRowBatch.DEFAULT_SIZE; i++) {
        bcv.setRef(i, b, 0, b.length);
    }
    // verify
    byte[] v = new byte[b.length];
    for (int i = 0; i != VectorizedRowBatch.DEFAULT_SIZE; i++) {
        Assert.assertTrue(bcv.length[i] == b.length);
        System.arraycopy(bcv.vector[i], bcv.start[i], v, 0, b.length);
        Assert.assertTrue(Arrays.equals(b, v));
    }
}

Example 24 with BytesColumnVector

use of org.apache.hadoop.hive.ql.exec.vector.BytesColumnVector in project hive by apache.

the class TestVectorStringExpressions method testLoadBytesColumnVectorByValueLargeData.

@Test
public // the buffer to expand.
void testLoadBytesColumnVectorByValueLargeData() {
    BytesColumnVector bcv = new BytesColumnVector(VectorizedRowBatch.DEFAULT_SIZE);
    // initialize with estimated element size 10
    bcv.initBuffer(10);
    // Record initial buffer size
    int initialBufferSize = bcv.bufferSize();
    String s = "0123456789";
    while (s.length() < 500) {
        s += s;
    }
    byte[] b = null;
    try {
        b = s.getBytes("UTF-8");
    } catch (UnsupportedEncodingException e) {
        e.printStackTrace();
    }
    for (int i = 0; i != VectorizedRowBatch.DEFAULT_SIZE; i++) {
        bcv.setVal(i, b, 0, b.length);
    }
    // Current buffer size should be larger than initial size
    Assert.assertTrue(bcv.bufferSize() > initialBufferSize);
}

Example 25 with BytesColumnVector

use of org.apache.hadoop.hive.ql.exec.vector.BytesColumnVector in project hive by apache.

the class TestVectorStringExpressions method testStringLikePatternType.

@Test
public void testStringLikePatternType() throws UnsupportedEncodingException, HiveException {
    FilterStringColLikeStringScalar expr;
    VectorizedRowBatch vrb = VectorizedRowGroupGenUtil.getVectorizedRowBatch(1, 1, 1);
    vrb.cols[0] = new BytesColumnVector(1);
    BytesColumnVector bcv = (BytesColumnVector) vrb.cols[0];
    vrb.size = 0;
    // BEGIN pattern
    expr = new FilterStringColLikeStringScalar(0, "abc%".getBytes());
    expr.evaluate(vrb);
    Assert.assertEquals(FilterStringColLikeStringScalar.BeginChecker.class, expr.checker.getClass());
    // END pattern
    expr = new FilterStringColLikeStringScalar(0, "%abc".getBytes("UTF-8"));
    expr.evaluate(vrb);
    Assert.assertEquals(FilterStringColLikeStringScalar.EndChecker.class, expr.checker.getClass());
    // MIDDLE pattern
    expr = new FilterStringColLikeStringScalar(0, "%abc%".getBytes());
    expr.evaluate(vrb);
    Assert.assertEquals(FilterStringColLikeStringScalar.MiddleChecker.class, expr.checker.getClass());
    // CHAIN pattern
    expr = new FilterStringColLikeStringScalar(0, "%abc%de".getBytes());
    expr.evaluate(vrb);
    Assert.assertEquals(FilterStringColLikeStringScalar.ChainedChecker.class, expr.checker.getClass());
    // COMPLEX pattern
    expr = new FilterStringColLikeStringScalar(0, "%abc_%de".getBytes());
    expr.evaluate(vrb);
    Assert.assertEquals(FilterStringColLikeStringScalar.ComplexChecker.class, expr.checker.getClass());
    // NONE pattern
    expr = new FilterStringColLikeStringScalar(0, "abc".getBytes());
    expr.evaluate(vrb);
    Assert.assertEquals(FilterStringColLikeStringScalar.NoneChecker.class, expr.checker.getClass());
}