/** Unit test for TupleDesc.combine() */
@Test
public void combine() {
TupleDesc td1, td2, td3;
td1 = Utility.getTupleDesc(1, "td1");
td2 = Utility.getTupleDesc(2, "td2");
// test td1.combine(td2)
td3 = TupleDesc.merge(td1, td2);
assertEquals(3, td3.numFields());
assertEquals(3 * Type.INT_TYPE.getLen(), td3.getSize());
for (int i = 0; i < 3; ++i) assertEquals(Type.INT_TYPE, td3.getFieldType(i));
assertEquals(combinedStringArrays(td1, td2, td3), true);
// test td2.combine(td1)
td3 = TupleDesc.merge(td2, td1);
assertEquals(3, td3.numFields());
assertEquals(3 * Type.INT_TYPE.getLen(), td3.getSize());
for (int i = 0; i < 3; ++i) assertEquals(Type.INT_TYPE, td3.getFieldType(i));
assertEquals(combinedStringArrays(td2, td1, td3), true);
// test td2.combine(td2)
td3 = TupleDesc.merge(td2, td2);
assertEquals(4, td3.numFields());
assertEquals(4 * Type.INT_TYPE.getLen(), td3.getSize());
for (int i = 0; i < 4; ++i) assertEquals(Type.INT_TYPE, td3.getFieldType(i));
assertEquals(combinedStringArrays(td2, td2, td3), true);
}
@Test
public void new_getFieldTypeTest() {
int length = 100;
String name = "td";
TupleDesc td = Utility.getTupleDesc(length, name);
// Lower than index bound.
try {
td.getFieldType(length + 1);
fail("expected exception");
} catch (NoSuchElementException e) {
}
// Higher than index bound.
try {
td.getFieldType(-1);
fail("expected exception");
} catch (NoSuchElementException e) {
}
// Check each name.
for (int i = 0; i < length; i++) {
assertEquals(Type.INT_TYPE, td.getFieldType(i));
}
}
/**
* Create a new TableStats object, that keeps track of statistics on each column of a table
*
* @param tableid The table over which to compute statistics
* @param ioCostPerPage The cost per page of IO. This doesn't differentiate between
* sequential-scan IO and disk seeks.
*/
public TableStats(int tableid, int ioCostPerPage) {
// For this function, we use the DbFile for the table in question,
// then scan through its tuples and calculate the values that you
// to build the histograms.
// TODO: Fill out the rest of the constructor.
// Feel free to change anything already written, it's only a guideline
this.ioCostPerPage = ioCostPerPage;
DbFile file = Database.getCatalog().getDbFile(tableid);
tupleDesc = file.getTupleDesc();
numPages = ((HeapFile) file).numPages();
numTuples = 0;
int numFields = tupleDesc.numFields();
// TODO: what goes here?
statistics = new ArrayList<Object>();
for (int i = 0; i < numFields; i++) {
if (Type.INT_TYPE.equals(tupleDesc.getFieldType(i))) {
statistics.add(new IntStatistics(NUM_HIST_BINS));
} else {
statistics.add(new StringHistogram(NUM_HIST_BINS));
}
}
final DbFileIterator iter = file.iterator(null);
try {
iter.open();
while (iter.hasNext()) {
Tuple t = iter.next();
numTuples++;
// TODO: and here?
for (int i = 0; i < numFields; i++) {
if (Type.INT_TYPE.equals(tupleDesc.getFieldType(i))) {
((IntStatistics) statistics.get(i)).addValue(((IntField) t.getField(i)).getValue());
} else {
((StringHistogram) statistics.get(i))
.addValue(((StringField) t.getField(i)).getValue());
}
}
}
iter.close();
} catch (DbException e) {
e.printStackTrace();
} catch (TransactionAbortedException e) {
e.printStackTrace();
}
}
/** Suck up tuples from the source file. */
private Tuple readNextTuple(DataInputStream dis, int slotId) throws NoSuchElementException {
// if associated bit is not set, read forward to the next tuple, and
// return null.
if (!isSlotUsed(slotId)) {
for (int i = 0; i < td.getSize(); i++) {
try {
dis.readByte();
} catch (IOException e) {
throw new NoSuchElementException("error reading empty tuple");
}
}
return null;
}
// read fields in the tuple
Tuple t = new Tuple(td);
RecordId rid = new RecordId(pid, slotId);
t.setRecordId(rid);
try {
for (int j = 0; j < td.numFields(); j++) {
Field f = td.getFieldType(j).parse(dis);
t.setField(j, f);
}
} catch (java.text.ParseException e) {
e.printStackTrace();
throw new NoSuchElementException("parsing error!");
}
return t;
}
/** Unit test for TupleDesc.getType() */
@Test
public void getType() {
int[] lengths = new int[] {1, 2, 1000};
for (int len : lengths) {
TupleDesc td = Utility.getTupleDesc(len);
for (int i = 0; i < len; ++i) assertEquals(Type.INT_TYPE, td.getFieldType(i));
}
}
/**
* Estimate the selectivity of predicate <tt>field op constant</tt> on the table.
*
* @param field The field over which the predicate ranges
* @param op The logical operation in the predicate
* @param constant The value against which the field is compared
* @return The estimated selectivity (fraction of tuples that satisfy) the predicate
*/
public double estimateSelectivity(int field, Predicate.Op op, Field constant) {
// TODO: some code goes here
if (Type.INT_TYPE.equals(tupleDesc.getFieldType(field))) {
return ((IntStatistics) statistics.get(field))
.estimateSelectivity(op, ((IntField) constant).getValue());
} else {
return ((StringHistogram) statistics.get(field))
.estimateSelectivity(op, ((StringField) constant).getValue());
}
}
/**
* Returns the TupleDesc with field names from the underlying HeapFile, prefixed with the
* tableAlias string from the constructor. This prefix becomes useful when joining tables
* containing a field(s) with the same name.
*
* @return the TupleDesc with field names from the underlying HeapFile, prefixed with the
* tableAlias string from the constructor.
*/
public TupleDesc getTupleDesc() {
TupleDesc tup = Database.getCatalog().getTupleDesc(id);
int length = tup.numFields();
String[] field = new String[length];
Type[] types = new Type[length];
for (int i = 0; i < length; i++) {
types[i] = tup.getFieldType(i);
field[i] = alias + "." + tup.getFieldName(i);
}
return new TupleDesc(types, field);
}
private void createAliasedTd() {
Catalog gc = Database.getCatalog();
TupleDesc old_td = gc.getTupleDesc(tableid);
String[] newFieldAr = new String[old_td.numFields()];
Type[] typeAr = new Type[old_td.numFields()];
String field = null;
for (int i = 0; i < newFieldAr.length; i++) {
field = old_td.getFieldName(i);
if (alias == null) {
alias = "null";
} else if (field == null) {
field = "null";
}
newFieldAr[i] = alias + "." + field;
typeAr[i] = old_td.getFieldType(i);
}
td = new TupleDesc(typeAr, newFieldAr);
}