Use statistics in Faker CTAS

plugin/trino-faker/src/main/java/io/trino/plugin/faker/ColumnInfo.java

@@ -13,9 +13,11 @@
  */
 package io.trino.plugin.faker;
 
+import com.google.common.collect.ImmutableMap;
 import io.trino.spi.connector.ColumnMetadata;
 import io.trino.spi.type.Type;
 
+import java.util.Map;
 import java.util.Optional;
 
 import static java.util.Objects.requireNonNull;
@@ -57,4 +59,16 @@ public ColumnInfo withComment(Optional<String> comment)
                 .setComment(comment)
                 .build());
     }
+
+    public ColumnInfo withHandle(FakerColumnHandle handle)
+    {
+        return new ColumnInfo(handle, metadata);
+    }
+
+    public ColumnInfo withProperties(Map<String, Object> properties)
+    {
+        return new ColumnInfo(handle, ColumnMetadata.builderFrom(metadata)
+                .setProperties(ImmutableMap.copyOf(properties))
+                .build());
+    }
 }

plugin/trino-faker/src/main/java/io/trino/plugin/faker/FakerColumnHandle.java

@@ -174,4 +174,19 @@ private static List<String> strings(Collection<?> values)
                 .map(String.class::cast)
                 .collect(toImmutableList());
     }
+
+    public FakerColumnHandle withNullProbability(double nullProbability)
+    {
+        return new FakerColumnHandle(columnIndex, name, type, nullProbability, generator, domain, step);
+    }
+
+    public FakerColumnHandle withDomain(Domain domain)
+    {
+        return new FakerColumnHandle(columnIndex, name, type, nullProbability, generator, domain, step);
+    }
+
+    public FakerColumnHandle withStep(ValueSet step)
+    {
+        return new FakerColumnHandle(columnIndex, name, type, nullProbability, generator, domain, step);
+    }
 }

plugin/trino-faker/src/main/java/io/trino/plugin/faker/FakerConfig.java

@@ -26,6 +26,8 @@ public class FakerConfig
     private double nullProbability = 0.5;
     private long defaultLimit = 1000L;
     private Locale locale = Locale.ENGLISH;
+    private boolean sequenceDetectionEnabled = true;
+    private boolean dictionaryDetectionEnabled = true;
 
     @Max(1)
     @Min(0)
@@ -68,4 +70,36 @@ public FakerConfig setLocale(String value)
         this.locale = new Locale.Builder().setLanguageTag(value).build();
         return this;
     }
+
+    public boolean isSequenceDetectionEnabled()
+    {
+        return sequenceDetectionEnabled;
+    }
+
+    @Config("faker.sequence-detection-enabled")
+    @ConfigDescription(
+            """
+            If true, when creating a table using existing data, columns with the number of distinct values close to
+            the number of rows will be treated as sequences""")
+    public FakerConfig setSequenceDetectionEnabled(boolean value)
+    {
+        this.sequenceDetectionEnabled = value;
+        return this;
+    }
+
+    public boolean isDictionaryDetectionEnabled()
+    {
+        return dictionaryDetectionEnabled;
+    }
+
+    @Config("faker.dictionary-detection-enabled")
+    @ConfigDescription(
+            """
+            If true, when creating a table using existing data, columns with a low number of distinct values
+            will have the allowed_values column property populated with random values""")
+    public FakerConfig setDictionaryDetectionEnabled(boolean value)
+    {
+        this.dictionaryDetectionEnabled = value;
+        return this;
+    }
 }

plugin/trino-faker/src/main/java/io/trino/plugin/faker/FakerConnector.java

@@ -44,6 +44,7 @@
 import static io.trino.spi.StandardErrorCode.INVALID_SCHEMA_PROPERTY;
 import static io.trino.spi.StandardErrorCode.INVALID_TABLE_PROPERTY;
 import static io.trino.spi.connector.ConnectorCapabilities.NOT_NULL_COLUMN_CONSTRAINT;
+import static io.trino.spi.session.PropertyMetadata.booleanProperty;
 import static io.trino.spi.session.PropertyMetadata.doubleProperty;
 import static io.trino.spi.session.PropertyMetadata.longProperty;
 import static io.trino.spi.session.PropertyMetadata.stringProperty;
@@ -125,6 +126,20 @@ public List<PropertyMetadata<?>> getSchemaProperties()
                         "Default limit of rows returned from any table in this schema, if not specified in the query",
                         null,
                         defaultLimit -> checkProperty(1 <= defaultLimit, INVALID_SCHEMA_PROPERTY, "default_limit value must be equal or greater than 1"),
+                        false),
+                booleanProperty(
+                        SchemaInfo.SEQUENCE_DETECTION_ENABLED,
+                        """
+                        If true, when creating a table using existing data, columns with the number of distinct values close to
+                        the number of rows will be treated as sequences""",
+                        null,
+                        false),
+                booleanProperty(
+                        SchemaInfo.DICTIONARY_DETECTION_ENABLED,
+                        """
+                        If true, when creating a table using existing data, columns with a low number of distinct values
+                        will have the allowed_values column property populated with random values""",
+                        null,
                         false));
     }
 
@@ -143,6 +158,20 @@ public List<PropertyMetadata<?>> getTableProperties()
                         "Default limit of rows returned from this table if not specified in the query",
                         null,
                         defaultLimit -> checkProperty(1 <= defaultLimit, INVALID_TABLE_PROPERTY, "default_limit value must be equal or greater than 1"),
+                        false),
+                booleanProperty(
+                        TableInfo.SEQUENCE_DETECTION_ENABLED,
+                        """
+                        If true, when creating a table using existing data, columns with the number of distinct values close to
+                        the number of rows will be treated as sequences""",
+                        null,
+                        false),
+                booleanProperty(
+                        TableInfo.DICTIONARY_DETECTION_ENABLED,
+                        """
+                        If true, when creating a table using existing data, columns with a low number of distinct values
+                        will have the allowed_values column property populated with random values""",
+                        null,
                         false));
     }