Skip to content

normalize Spanner CDC #40040

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

Merged
Abacn merged 3 commits into from
Sep 16, 2026
Merged

normalize Spanner CDC #40040

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
df05227
normalize Spanner CDC
aIbrahiim Sep 7, 2026
bda4c79
use hashset for spanner table filter
aIbrahiim Sep 15, 2026
25a528f
Merge master into normalize-spanner-cdc
aIbrahiim Sep 15, 2026
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
2 changes: 1 addition & 1 deletion .github/trigger_files/beam_PostCommit_Yaml_Xlang_Direct.json
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,4 +1,4 @@
{
"comment": "Modify this file in a trivial way to cause this test suite to run",
"revision": 3
"revision": 9
}
73 changes: 53 additions & 20 deletions ...le-cloud-platform/src/main/java/org/apache/beam/sdk/io/gcp/spanner/ReadSpannerSchema.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -23,8 +23,11 @@
import com.google.cloud.spanner.ResultSet;
import com.google.cloud.spanner.Statement;
import io.opentelemetry.api.OpenTelemetry;
import java.util.Collections;
import java.util.HashSet;
import java.util.Locale;
import java.util.Set;
import java.util.stream.Collectors;
import org.apache.beam.sdk.options.PipelineOptions;
import org.apache.beam.sdk.options.SdkHarnessOptions;
import org.apache.beam.sdk.transforms.DoFn;
Expand Down Expand Up @@ -77,22 +80,29 @@ public ReadSpannerSchema(
this.allowedTableNames = allowedTableNames == null ? new HashSet<>() : allowedTableNames;
}

@Setup
public void setup(PipelineOptions options) throws Exception {
OpenTelemetry otel = options.as(SdkHarnessOptions.class).getOpenTelemetry();
spannerAccessor = SpannerAccessor.getOrCreate(config, otel);
}

@Teardown
public void teardown() throws Exception {
spannerAccessor.close();
/**
* Reads Spanner schema information without running a Beam pipeline.
*
* <p>Used by SchemaTransforms during expansion (including cross-language expansion services that
* do not ship DirectRunner).
*/
public static SpannerSchema getSpannerSchema(
SpannerConfig config, Dialect dialect, Set<String> allowedTableNames) {
try (SpannerAccessor spannerAccessor = SpannerAccessor.getOrCreate(config)) {
return getSpannerSchema(spannerAccessor.getDatabaseClient(), dialect, allowedTableNames);
}
}

@ProcessElement
public void processElement(ProcessContext c) throws Exception {
Dialect dialect = c.sideInput(dialectView);
static SpannerSchema getSpannerSchema(
DatabaseClient databaseClient, Dialect dialect, Set<String> allowedTableNames) {
// Case insensitive match via lower cased HashSet
Set<String> allowedLower =
allowedTableNames == null || allowedTableNames.isEmpty()
? Collections.emptySet()
: allowedTableNames.stream()
.map(name -> name.toLowerCase(Locale.ROOT))
.collect(Collectors.toSet());
SpannerSchema.Builder builder = SpannerSchema.builder(dialect);
DatabaseClient databaseClient = spannerAccessor.getDatabaseClient();
try (ReadOnlyTransaction tx = databaseClient.readOnlyTransaction()) {
ResultSet resultSet = readTableInfo(tx, dialect);

Expand All @@ -101,9 +111,7 @@ public void processElement(ProcessContext c) throws Exception {
String columnName = resultSet.getString(1);
String type = resultSet.getString(2);
long cellsMutated = resultSet.getLong(3);
if (allowedTableNames.size() > 0 && !allowedTableNames.contains(tableName)) {
// If we want to filter out table names, and the current table name is not part
// of the allowed names, we exclude it.
if (!isTableAllowed(allowedLower, tableName)) {
continue;
}
builder.addColumn(tableName, columnName, type, cellsMutated);
Expand All @@ -114,14 +122,39 @@ public void processElement(ProcessContext c) throws Exception {
String tableName = resultSet.getString(0);
String columnName = resultSet.getString(1);
String ordering = resultSet.getString(2);

if (!isTableAllowed(allowedLower, tableName)) {
continue;
}
builder.addKeyPart(tableName, columnName, "DESC".equalsIgnoreCase(ordering));
}
}
c.output(builder.build());
return builder.build();
}

private static boolean isTableAllowed(Set<String> allowedLowerTableNames, String tableName) {
return allowedLowerTableNames.isEmpty()
|| allowedLowerTableNames.contains(tableName.toLowerCase(Locale.ROOT));
}

@Setup
public void setup(PipelineOptions options) throws Exception {
OpenTelemetry otel = options.as(SdkHarnessOptions.class).getOpenTelemetry();
spannerAccessor = SpannerAccessor.getOrCreate(config, otel);
}

@Teardown
public void teardown() throws Exception {
spannerAccessor.close();
}

@ProcessElement
public void processElement(ProcessContext c) throws Exception {
c.output(
getSpannerSchema(
spannerAccessor.getDatabaseClient(), c.sideInput(dialectView), allowedTableNames));
}

private ResultSet readTableInfo(ReadOnlyTransaction tx, Dialect dialect) {
private static ResultSet readTableInfo(ReadOnlyTransaction tx, Dialect dialect) {
// retrieve schema information for all tables, as well as aggregating the
// number of indexes that cover each column. this will be used to estimate
// the number of cells (table column plus indexes) mutated in an upsert operation
Expand Down Expand Up @@ -174,7 +207,7 @@ private ResultSet readTableInfo(ReadOnlyTransaction tx, Dialect dialect) {
return tx.executeQuery(Statement.of(statement));
}

private ResultSet readPrimaryKeyInfo(ReadOnlyTransaction tx, Dialect dialect) {
private static ResultSet readPrimaryKeyInfo(ReadOnlyTransaction tx, Dialect dialect) {
String statement = "";
switch (dialect) {
case GOOGLE_STANDARD_SQL:
Expand Down
50 changes: 10 additions & 40 deletions ...eam/sdk/io/gcp/spanner/changestreams/SpannerChangestreamsReadSchemaTransformProvider.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -27,7 +27,6 @@
import java.nio.charset.StandardCharsets;
import java.util.Arrays;
import java.util.Collections;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
import java.util.Objects;
Expand All @@ -36,7 +35,6 @@
import java.util.stream.Collectors;
import java.util.stream.IntStream;
import org.apache.beam.sdk.Pipeline;
import org.apache.beam.sdk.coders.StringUtf8Coder;
import org.apache.beam.sdk.io.gcp.spanner.ReadSpannerSchema;
import org.apache.beam.sdk.io.gcp.spanner.SpannerConfig;
import org.apache.beam.sdk.io.gcp.spanner.SpannerIO;
Expand All @@ -52,14 +50,11 @@
import org.apache.beam.sdk.schemas.transforms.SchemaTransform;
import org.apache.beam.sdk.schemas.transforms.SchemaTransformProvider;
import org.apache.beam.sdk.schemas.transforms.TypedSchemaTransformProvider;
import org.apache.beam.sdk.transforms.Create;
import org.apache.beam.sdk.transforms.DoFn;
import org.apache.beam.sdk.transforms.DoFn.FinishBundle;
import org.apache.beam.sdk.transforms.ParDo;
import org.apache.beam.sdk.transforms.View;
import org.apache.beam.sdk.values.PCollectionRowTuple;
import org.apache.beam.sdk.values.PCollectionTuple;
import org.apache.beam.sdk.values.PCollectionView;
import org.apache.beam.sdk.values.Row;
import org.apache.beam.sdk.values.TupleTag;
import org.apache.beam.sdk.values.TupleTagList;
Expand Down Expand Up @@ -304,42 +299,17 @@ public void finish(FinishBundleContext c) {
}
}

private static final HashMap<String, SpannerSchema> TABLE_SCHEMAS = new HashMap<>();

private static Schema getTableSchema(SpannerChangestreamsReadConfiguration config) {
Pipeline miniPipeline = Pipeline.create();

@Abacn Abacn Sep 15, 2026

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Either way, this requires pipeline submission time access to the Spanner instance. As a follow up, to allow pass a schema so that it won't need IO resource access at pipeline submission time, or create a Issue to track

@aIbrahiim aIbrahiim Sep 15, 2026

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

ok and yes both still need Spanner access at expansion and submission time

I'll create a follow up issue to allow an optional schema config on the Spanner CDC SchemaTransform so expansion can skip live information_schema access when the schema is provided

PCollectionView<Dialect> sqlDialectView =
miniPipeline
.apply("Create Dialect", Create.of(Dialect.GOOGLE_STANDARD_SQL))
.apply("Dialect to View", View.asSingleton());
miniPipeline
.apply(Create.of((Void) null))
.apply(
ParDo.of(
new ReadSpannerSchema(
SpannerConfig.create()
.withDatabaseId(config.getDatabaseId())
.withInstanceId(config.getInstanceId())
.withProjectId(config.getProjectId()),
sqlDialectView,
Sets.newHashSet(config.getTable())))
.withSideInput("dialect", sqlDialectView))
.apply(
ParDo.of(
new DoFn<SpannerSchema, String>() {
@ProcessElement
public void process(@DoFn.Element SpannerSchema schema) {
TABLE_SCHEMAS.put(config.getTable(), schema);
}
}))
.setCoder(StringUtf8Coder.of());
miniPipeline.run().waitUntilFinish();
// Clean up the static map from the object.
SpannerSchema finalSchemaObj = TABLE_SCHEMAS.remove(config.getTable());
if (finalSchemaObj == null) {
throw new RuntimeException(
String.format("Could not get schema for configuration %s", config));
}
// Query information_schema directly. A nested Pipeline would require DirectRunner,
// which is not on the GCP expansion-service classpath used by cross-language YAML.
SpannerSchema finalSchemaObj =
ReadSpannerSchema.getSpannerSchema(
SpannerConfig.create()
.withDatabaseId(config.getDatabaseId())
.withInstanceId(config.getInstanceId())
.withProjectId(config.getProjectId()),
Dialect.GOOGLE_STANDARD_SQL,
Sets.newHashSet(config.getTable()));
return spannerSchemaToBeamSchema(finalSchemaObj, config.getTable());
}

Expand Down
21 changes: 21 additions & 0 deletions sdks/python/apache_beam/yaml/standard_io.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -515,6 +515,27 @@
config:
gradle_target: 'sdks:java:io:google-cloud-platform:expansion-service:shadowJar'

# Spanner CDC
- type: renaming
transforms:
'ReadFromSpannerCDC': 'ReadFromSpannerCDC'
config:
mappings:
'ReadFromSpannerCDC':
project: 'project_id'
instance: 'instance_id'
database: 'database_id'
table: 'table'
change_stream: 'change_stream_name'
start_at: 'start_at_timestamp'
end_at: 'end_at_timestamp'
underlying_provider:
type: beamJar
transforms:
'ReadFromSpannerCDC': 'beam:schematransform:org.apache.beam:spanner_cdc_read:v1'
config:
gradle_target: 'sdks:java:io:google-cloud-platform:expansion-service:shadowJar'

# Firestore
- type: renaming
transforms:
Expand Down
Loading