Skip to content
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

Jump to bottom

fix: arrow parsing for nested and nullable fields #519

Open
wants to merge 3 commits into
base: main
Choose a base branch
from
+216 −19
Open

fix: arrow parsing for nested and nullable fields #519

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
05c7f1c
fix: arrow parsing for nested and nullable fields
alvarowolfx Oct 15, 2024
0405681
fix: address pr comments
alvarowolfx Oct 17, 2024
fc0ae4a
docs: add comment
alvarowolfx Oct 17, 2024
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
44 changes: 30 additions & 14 deletions src/reader/arrow_transform.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -17,7 +17,7 @@
RecordBatchReader,
RecordBatch,
RecordBatchStreamReader,
Vector,
DataType,
} from 'apache-arrow';
import * as protos from '../../protos/protos';

Expand All @@ -26,7 +26,7 @@
type ReadSession = protos.google.cloud.bigquery.storage.v1.IReadSession;

interface TableCell {
v?: any;

Check warning on line 29 in src/reader/arrow_transform.ts

View workflow job for this annotation

GitHub Actions / lint 

Unexpected any. Specify a different type
}
interface TableRow {
f?: Array<TableCell>;
Expand Down Expand Up @@ -140,12 +140,13 @@
}
for (let j = 0; j < batch.numCols; j++) {
const column = batch.selectAt([j]);
const columnName = column.schema.fields[0].name;
const field = column.schema.fields[0];
const columnName = field.name;
for (let i = 0; i < batch.numRows; i++) {
const fieldData = column.get(i);
const fieldValue = fieldData?.toJSON()[columnName];
rows[i].f[j] = {
v: convertArrowValue(fieldValue),
v: convertArrowValue(fieldValue, field.type as DataType),
};
}
}
Expand All @@ -156,21 +157,36 @@
}
}

function convertArrowValue(fieldValue: any): any {
if (typeof fieldValue === 'object') {
if (fieldValue instanceof Vector) {
const arr = fieldValue.toJSON();
return arr.map((v: any) => {
return {v: convertArrowValue(v)};
});
}
const tableRow: TableRow = {f: []};
function convertArrowValue(fieldValue: any, type: DataType): any {

Check warning on line 160 in src/reader/arrow_transform.ts

View workflow job for this annotation

GitHub Actions / lint 

Unexpected any. Specify a different type

Check warning on line 160 in src/reader/arrow_transform.ts

View workflow job for this annotation

GitHub Actions / lint 

Unexpected any. Specify a different type
if (fieldValue === null) {
return null;
alvarowolfx marked this conversation as resolved.
Show resolved Hide resolved
}
if (DataType.isList(type)) {
const arr = fieldValue.toJSON();
return arr.map((v: any) => {

Check warning on line 166 in src/reader/arrow_transform.ts

View workflow job for this annotation

GitHub Actions / lint 

Unexpected any. Specify a different type
// Arrays/lists in BigQuery have the same datatype for every element
// so getting the first one is all we need
const elemType = type.children[0].type;
alvarowolfx marked this conversation as resolved.
Show resolved Hide resolved
return {v: convertArrowValue(v, elemType)};
});
}
if (DataType.isStruct(type)) {
const tableRow: TableRow = {};
Object.keys(fieldValue).forEach(key => {
tableRow.f?.push({
v: convertArrowValue(fieldValue[key]),
const elemType = type.children.find(f => f.name === key);
if (!tableRow.f) {
tableRow.f = [];
}
tableRow.f.push({
v: convertArrowValue(fieldValue[key], elemType?.type as DataType),
});
});
return tableRow;
}
if (DataType.isTimestamp(type)) {
// timestamp comes in microsecond, convert to nanoseconds
// to make it compatible with BigQuery.timestamp.
return fieldValue * 1000;
}
return fieldValue;
}
191 changes: 186 additions & 5 deletions system-test/reader_client_test.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -72,6 +72,33 @@ describe('reader.ReaderClient', () => {
type: 'INTEGER',
mode: 'REQUIRED',
},
{
name: 'optional',
type: 'STRING',
mode: 'NULLABLE',
},
{
name: 'list',
type: 'INT64',
mode: 'REPEATED',
},
{
name: 'metadata',
type: 'RECORD',
mode: 'NULLABLE',
fields: [
{
name: 'created_at',
type: 'TIMESTAMP',
mode: 'REQUIRED',
},
{
name: 'updated_at',
type: 'TIMESTAMP',
mode: 'NULLABLE',
},
],
},
],
};

Expand All @@ -97,9 +124,26 @@ describe('reader.ReaderClient', () => {
.dataset(datasetId)
.table(tableId)
.insert([
{name: 'Ada Lovelace', row_num: 1},
{name: 'Alan Turing', row_num: 2},
{name: 'Bell', row_num: 3},
{
name: 'Ada Lovelace',
row_num: 1,
optional: 'Some data',
list: [1],
metadata: {
created_at: bigquery.timestamp('2020-04-27T18:07:25.356Z'),
updated_at: bigquery.timestamp('2020-04-27T20:07:25.356Z'),
alvarowolfx marked this conversation as resolved.
Show resolved Hide resolved
},
},
{
name: 'Alan Turing',
row_num: 2,
optional: 'Some other data',
list: [1, 2],
metadata: {
created_at: bigquery.timestamp('2020-04-27T18:07:25.356Z'),
},
},
{name: 'Bell', row_num: 3, list: [1, 2, 3]},
]);
});

Expand Down Expand Up @@ -218,7 +262,7 @@ describe('reader.ReaderClient', () => {
const table = await tableFromIPC(content);

assert.equal(table.numRows, 3);
assert.equal(table.numCols, 2);
assert.equal(table.numCols, 5);

reader.close();
} finally {
Expand Down Expand Up @@ -253,7 +297,7 @@ describe('reader.ReaderClient', () => {
const table = new Table(batches);

assert.equal(table.numRows, 3);
assert.equal(table.numCols, 2);
assert.equal(table.numCols, 5);

reader.close();
} finally {
Expand Down Expand Up @@ -295,6 +339,143 @@ describe('reader.ReaderClient', () => {

assert.equal(rows.length, 3);

assert.deepEqual(rows, [
{
f: [
{
v: 'Ada Lovelace',
},
{
v: '1',
},
{
v: 'Some data',
},
{
v: [
{
v: '1',
},
],
},
{
v: {
f: [
{
v: 1588010845356000,
},
{
v: 1588018045356000,
},
],
},
},
],
},
{
f: [
{
v: 'Alan Turing',
},
{
v: '2',
},
{
v: 'Some other data',
},
{
v: [
{
v: '1',
},
{
v: '2',
},
],
},
{
v: {
f: [
{
v: 1588010845356000,
},
{
v: null,
},
],
},
},
],
},
{
f: [
{
v: 'Bell',
},
{
v: '3',
},
{
v: null,
},
{
v: [
{
v: '1',
},
{
v: '2',
},
{
v: '3',
},
],
},
{
v: null,
},
],
},
]);
const mergedRows = BigQuery.mergeSchemaWithRows_(schema, rows, {
wrapIntegers: false,
});
assert.deepEqual(mergedRows, [
{
name: 'Ada Lovelace',
row_num: 1,
optional: 'Some data',
list: [1],
metadata: {
created_at: {
value: '2020-04-27T18:07:25.356Z',
},
updated_at: {
value: '2020-04-27T20:07:25.356Z',
},
},
},
{
name: 'Alan Turing',
row_num: 2,
optional: 'Some other data',
list: [1, 2],
metadata: {
created_at: {
value: '2020-04-27T18:07:25.356Z',
},
updated_at: null,
},
},
{
name: 'Bell',
row_num: 3,
list: [1, 2, 3],
optional: null,
metadata: null,
},
]);

reader.close();
} finally {
client.close();
Expand Down
Loading