public static MessageType convert(StructType struct, FieldProjectionFilter filter) {
State state = new State(new FieldsPath(), REPEATED, "ParquetSchema");
ConvertedField converted = struct.accept(new ThriftSchemaConvertVisitor(filter, true), state);
if (!converted.isKeep()) {
throw new ThriftProjectionException("No columns have been selected");
}
return new MessageType(state.name, converted.asKeep().getType().asGroupType().getFields());
}
@Override
public ConvertedField visit(StructType structType, State state) {
// special care is taken when converting unions,
// because we are actually both converting + projecting in
// one pass, and unions need special handling when projecting.
final boolean isUnion = isUnion(structType.getStructOrUnionType());
boolean hasSentinelUnionColumns = false;
boolean hasNonSentinelUnionColumns = false;
List<Type> convertedChildren = new ArrayList<Type>();
for (ThriftField child : structType.getChildren()) {
State childState = new State(state.path.push(child), getRepetition(child), child.getName());
ConvertedField converted = child.getType().accept(this, childState);
if (isUnion && !converted.isKeep()) {
// user is not keeping this "kind" of union, but we still need
// to keep at least one of the primitives of this union around.
// in order to know what "kind" of union each record is.
// TODO: in the future, we should just filter these records out instead
// re-do the recursion, with a new projection filter that keeps only
// the first primitive it encounters
ConvertedField firstPrimitive =
child
.getType()
.accept(
new ThriftSchemaConvertVisitor(new KeepOnlyFirstPrimitiveFilter(), true),
childState);
convertedChildren.add(firstPrimitive.asKeep().getType().withId(child.getFieldId()));
hasSentinelUnionColumns = true;
}
if (converted.isSentinelUnion()) {
// child field is a sentinel union that we should drop if possible
if (childState.repetition == REQUIRED) {
// but this field is required, so we may still need it
convertedChildren.add(converted.asSentinelUnion().getType().withId(child.getFieldId()));
hasSentinelUnionColumns = true;
}
} else if (converted.isKeep()) {
// user has selected this column, so we keep it.
convertedChildren.add(converted.asKeep().getType().withId(child.getFieldId()));
hasNonSentinelUnionColumns = true;
}
}
if (!hasNonSentinelUnionColumns && hasSentinelUnionColumns) {
// this is a union, and user has not requested any of the children
// of this union. We should drop this union, if possible, but
// we may not be able to, so tag this as a sentinel.
return new SentinelUnion(
state.path, new GroupType(state.repetition, state.name, convertedChildren));
}
if (hasNonSentinelUnionColumns) {
// user requested some of the fields of this struct, so we keep the struct
return new Keep(state.path, new GroupType(state.repetition, state.name, convertedChildren));
} else {
// user requested none of the fields of this struct, so we drop it
return new Drop(state.path);
}
}