Skip to content

ARROW-2650: [JS] Implement Unions #2092

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
trxcllnt wants to merge 2 commits into from
Closed

ARROW-2650: [JS] Implement Unions #2092

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
953232c
write schema metadata
trxcllnt May 27, 2018
9b199c3
finish implementing unions in the JS reader
trxcllnt May 31, 2018
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 4 additions & 0 deletions js/src/data.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -216,9 +216,13 @@ export class ListData<T extends ListType> extends SingleNestedData<T> {
export class UnionData<T extends (DenseUnion | SparseUnion) = any> extends NestedData<T> {
public /* [VectorType.TYPE]:*/ 3: T['TArray'];
public get typeIds() { return this[VectorType.TYPE]; }
public readonly typeIdToChildIndex: { [key: number]: number };
constructor(type: T, length: number, nullBitmap: Uint8Array | null | undefined, typeIds: Iterable<number>, childData: Data<any>[], offset?: number, nullCount?: number) {
super(type, length, nullBitmap, childData, offset, nullCount);
this[VectorType.TYPE] = toTypedArray(Int8Array, typeIds);
this.typeIdToChildIndex = type.typeIds.reduce((typeIdToChildIndex, typeId, idx) => {
return (typeIdToChildIndex[typeId] = idx) && typeIdToChildIndex || typeIdToChildIndex;
}, Object.create(null) as { [key: number]: number });
}
public clone<R extends T>(type: R, length = this.length, offset = this.offset, nullCount = this._nullCount) {
return new UnionData<R>(type, length, this[VectorType.VALIDITY], this[VectorType.TYPE], this.childData, offset, nullCount);
Expand Down
37 changes: 20 additions & 17 deletions js/src/ipc/reader/json.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -76,6 +76,7 @@ function flattenDataSources(xs: any[]): any[][] {
...buffers,
...(column['VALIDITY'] && [column['VALIDITY']] || []),
...(column['OFFSET'] && [column['OFFSET']] || []),
...(column['TYPE'] && [column['TYPE']] || []),
...(column['DATA'] && [column['DATA']] || []),
...flattenDataSources(column['children'])
], [] as any[][]);
Expand Down Expand Up @@ -156,6 +157,7 @@ import * as Schema_ from '../../fb/Schema';
import Type = Schema_.org.apache.arrow.flatbuf.Type;
import DateUnit = Schema_.org.apache.arrow.flatbuf.DateUnit;
import TimeUnit = Schema_.org.apache.arrow.flatbuf.TimeUnit;
import UnionMode = Schema_.org.apache.arrow.flatbuf.UnionMode;
import Precision = Schema_.org.apache.arrow.flatbuf.Precision;
import IntervalUnit = Schema_.org.apache.arrow.flatbuf.IntervalUnit;
import MetadataVersion = Schema_.org.apache.arrow.flatbuf.MetadataVersion;
Expand Down Expand Up @@ -209,6 +211,7 @@ function buffersFromJSON(xs: any[], buffers: BufferMetadata[] = []): BufferMetad
const column = xs[i];
column['VALIDITY'] && buffers.push(new BufferMetadata(new Long(buffers.length, 0), new Long(column['VALIDITY'].length, 0)));
column['OFFSET'] && buffers.push(new BufferMetadata(new Long(buffers.length, 0), new Long(column['OFFSET'].length, 0)));
column['TYPE'] && buffers.push(new BufferMetadata(new Long(buffers.length, 0), new Long(column['TYPE'].length, 0)));
column['DATA'] && buffers.push(new BufferMetadata(new Long(buffers.length, 0), new Long(column['DATA'].length, 0)));
buffers = buffersFromJSON(column['children'], buffers);
}
Expand Down Expand Up @@ -293,31 +296,31 @@ function typeFromJSON(t: any, children?: Field[]) {
throw new Error(`Unrecognized type ${t['name']}`);
}

function nullFromJSON (_type: any) { return new Null(); }
function nullFromJSON (_type: any) { return new Null(); }
function intFromJSON (_type: any) { switch (_type['bitWidth']) {
case 8: return _type['isSigned'] ? new Int8() : new Uint8();
case 16: return _type['isSigned'] ? new Int16() : new Uint16();
case 32: return _type['isSigned'] ? new Int32() : new Uint32();
case 64: return _type['isSigned'] ? new Int64() : new Uint64();
}
return null; }
return null; }
function floatingPointFromJSON (_type: any) { switch (Precision[_type['precision']] as any) {
case Precision.HALF: return new Float16();
case Precision.SINGLE: return new Float32();
case Precision.DOUBLE: return new Float64();
}
return null; }
function binaryFromJSON (_type: any) { return new Binary(); }
function utf8FromJSON (_type: any) { return new Utf8(); }
function boolFromJSON (_type: any) { return new Bool(); }
function decimalFromJSON (_type: any) { return new Decimal(_type['scale'], _type['precision']); }
function dateFromJSON (_type: any) { return new Date_(DateUnit[_type['unit']] as any); }
function timeFromJSON (_type: any) { return new Time(TimeUnit[_type['unit']] as any, _type['bitWidth'] as TimeBitWidth); }
function timestampFromJSON (_type: any) { return new Timestamp(TimeUnit[_type['unit']] as any, _type['timezone']); }
function intervalFromJSON (_type: any) { return new Interval(IntervalUnit[_type['unit']] as any); }
function listFromJSON (_type: any, children: Field[]) { return new List(children); }
function structFromJSON (_type: any, children: Field[]) { return new Struct(children); }
function unionFromJSON (_type: any, children: Field[]) { return new Union(_type['mode'], (_type['typeIdsArray'] || []) as Type[], children); }
function fixedSizeBinaryFromJSON(_type: any) { return new FixedSizeBinary(_type['byteWidth']); }
function fixedSizeListFromJSON (_type: any, children: Field[]) { return new FixedSizeList(_type['listSize'], children); }
function mapFromJSON (_type: any, children: Field[]) { return new Map_(_type['keysSorted'], children); }
return null; }
function binaryFromJSON (_type: any) { return new Binary(); }
function utf8FromJSON (_type: any) { return new Utf8(); }
function boolFromJSON (_type: any) { return new Bool(); }
function decimalFromJSON (_type: any) { return new Decimal(_type['scale'], _type['precision']); }
function dateFromJSON (_type: any) { return new Date_(DateUnit[_type['unit']] as any); }
function timeFromJSON (_type: any) { return new Time(TimeUnit[_type['unit']] as any, _type['bitWidth'] as TimeBitWidth); }
function timestampFromJSON (_type: any) { return new Timestamp(TimeUnit[_type['unit']] as any, _type['timezone']); }
function intervalFromJSON (_type: any) { return new Interval(IntervalUnit[_type['unit']] as any); }
function listFromJSON (_type: any, children: Field[]) { return new List(children); }
function structFromJSON (_type: any, children: Field[]) { return new Struct(children); }
function unionFromJSON (_type: any, children: Field[]) { return new Union(UnionMode[_type['mode']] as any, (_type['typeIds'] || []) as Type[], children); }
function fixedSizeBinaryFromJSON(_type: any) { return new FixedSizeBinary(_type['byteWidth']); }
function fixedSizeListFromJSON (_type: any, children: Field[]) { return new FixedSizeList(_type['listSize'], children); }
function mapFromJSON (_type: any, children: Field[]) { return new Map_(_type['keysSorted'], children); }
24 changes: 22 additions & 2 deletions js/src/ipc/writer/binary.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -588,14 +588,34 @@ function writeMessage(b: Builder, node: Message) {
}

function writeSchema(b: Builder, node: Schema) {

const fieldOffsets = node.fields.map((f) => writeField(b, f));
const fieldsOffset =
_Schema.startFieldsVector(b, fieldOffsets.length) ||
_Schema.createFieldsVector(b, fieldOffsets);

let metadata: number | undefined = undefined;
if (node.metadata && node.metadata.size > 0) {
metadata = _Schema.createCustomMetadataVector(
b,
[...node.metadata].map(([k, v]) => {
const key = b.createString(`${k}`);
const val = b.createString(`${v}`);
return (
_KeyValue.startKeyValue(b) ||
_KeyValue.addKey(b, key) ||
_KeyValue.addValue(b, val) ||
_KeyValue.endKeyValue(b)
);
})
);
}

return (
_Schema.startSchema(b) ||
_Schema.addFields(b, fieldsOffset) ||
_Schema.addEndianness(b, platformIsLittleEndian ? _Endianness.Little : _Endianness.Big) ||
(metadata !== undefined && _Schema.addCustomMetadata(b, metadata)) ||
_Schema.endSchema(b)
);
}
Expand Down Expand Up @@ -662,8 +682,8 @@ function writeField(b: Builder, node: Field) {
metadata = _Field.createCustomMetadataVector(
b,
[...node.metadata].map(([k, v]) => {
const key = b.createString(k);
const val = b.createString(v);
const key = b.createString(`${k}`);
const val = b.createString(`${v}`);
return (
_KeyValue.startKeyValue(b) ||
_KeyValue.addKey(b, key) ||
Expand Down
6 changes: 4 additions & 2 deletions js/src/type.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -372,9 +372,11 @@ export class Union<TType extends Type = any> extends DataType<TType> {
constructor(public readonly mode: UnionMode,
public readonly typeIds: ArrowType[],
public readonly children: Field[]) {
super(<TType> (mode === UnionMode.Sparse ? Type.SparseUnion : Type.DenseUnion), children);
super(<TType> Type.Union, children);
}
public toString() { return `${this[Symbol.toStringTag]}<${this.typeIds.map((x) => Type[x]).join(` | `)}>`; }
public toString() { return `${this[Symbol.toStringTag]}<${
this.children.map((x) => `${x.type}`).join(` | `)
}>`; }
protected static [Symbol.toStringTag] = ((proto: Union) => {
(<any> proto).ArrayType = Int8Array;
return proto[Symbol.toStringTag] = 'Union';
Expand Down
28 changes: 16 additions & 12 deletions js/src/vector/nested.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -76,31 +76,35 @@ export class UnionView<T extends (DenseUnion | SparseUnion) = SparseUnion> exten
public typeIds: Int8Array;
// @ts-ignore
public valueOffsets?: Int32Array;
// @ts-ignore
protected typeIdToChildIndex: { [key: number]: number };
constructor(data: Data<T>, children?: Vector<any>[]) {
super(data, children);
this.length = data.length;
this.typeIds = data.typeIds;
this.typeIdToChildIndex = data.typeIdToChildIndex;
}
protected getNested(self: UnionView<T>, index: number): T['TValue'] {
return self.getChildValue(self, index, self.typeIds, self.valueOffsets);
return self.getChildValue(self, index, self.typeIds, self.valueOffsets, self.typeIdToChildIndex);
}
protected setNested(self: UnionView<T>, index: number, value: T['TValue']): void {
return self.setChildValue(self, index, value, self.typeIds, self.valueOffsets);
return self.setChildValue(self, index, value, self.typeIds, self.valueOffsets, self.typeIdToChildIndex);
}
protected getChildValue(self: NestedView<T>, index: number, typeIds: Int8Array, _valueOffsets?: any): any | null {
const child = self.getChildAt(typeIds[index]);
protected getChildValue(self: NestedView<T>, index: number, typeIds: Int8Array, _valueOffsets: any, typeIdToChildIndex: { [key: number]: number }): any | null {
const child = self.getChildAt(typeIdToChildIndex[typeIds[index]]);
return child ? child.get(index) : null;
}
protected setChildValue(self: NestedView<T>, index: number, value: T['TValue'], typeIds: Int8Array, _valueOffsets?: any): any | null {
const child = self.getChildAt(typeIds[index]);
protected setChildValue(self: NestedView<T>, index: number, value: T['TValue'], typeIds: Int8Array, _valueOffsets: any, typeIdToChildIndex: { [key: number]: number }): any | null {
const child = self.getChildAt(typeIdToChildIndex[typeIds[index]]);
return child ? child.set(index, value) : null;
}
public *[Symbol.iterator](): IterableIterator<T['TValue']> {
const length = this.length;
const get = this.getChildValue;
const { typeIdToChildIndex } = this;
const { typeIds, valueOffsets } = this;
for (let index = -1; ++index < length;) {
yield get(this, index, typeIds, valueOffsets);
yield get(this, index, typeIds, valueOffsets, typeIdToChildIndex);
}
}
}
Expand All @@ -112,14 +116,14 @@ export class DenseUnionView extends UnionView<DenseUnion> {
this.valueOffsets = data.valueOffsets;
}
protected getNested(self: DenseUnionView, index: number): any | null {
return self.getChildValue(self, index, self.typeIds, self.valueOffsets);
return self.getChildValue(self, index, self.typeIds, self.valueOffsets, self.typeIdToChildIndex);
}
protected getChildValue(self: NestedView<DenseUnion>, index: number, typeIds: Int8Array, valueOffsets: any): any | null {
const child = self.getChildAt(typeIds[index]);
protected getChildValue(self: NestedView<DenseUnion>, index: number, typeIds: Int8Array, valueOffsets: any, typeIdToChildIndex: { [key: number]: number }): any | null {
const child = self.getChildAt(typeIdToChildIndex[typeIds[index]]);
return child ? child.get(valueOffsets[index]) : null;
}
protected setChildValue(self: NestedView<DenseUnion>, index: number, value: any, typeIds: Int8Array, valueOffsets?: any): any | null {
const child = self.getChildAt(typeIds[index]);
protected setChildValue(self: NestedView<DenseUnion>, index: number, value: any, typeIds: Int8Array, valueOffsets: any, typeIdToChildIndex: { [key: number]: number }): any | null {
const child = self.getChildAt(typeIdToChildIndex[typeIds[index]]);
return child ? child.set(valueOffsets[index], value) : null;
}
}
Expand Down