table_changes table-valued function

Applies to: check marked yes Databricks SQL check marked yes Databricks Runtime

Returns a log of changes to a Delta Lake table with Change Data Feed enabled.

To invoke this function you need to have at least one of the following:

  • SELECT privilege on the specified table
  • Be the owner of the table
  • Have administrative privileges

Syntax

table_changes ( table_str, start [, end ] )

Arguments

  • table_str: A STRING literal representing the optionally qualified name of the table.
  • start: A BIGINT or TIMESTAMP literal, representing the first version or timestamp of change to return.
  • end: An optional BIGINT or TIMESTAMP literal, representing the last version or timestamp of change to return. If not specified all changes from start up to the current change are returned.

Returns

A table including all columns of the table identified in table_str, plus the following columns:

  • _change_type STRING NOT NULL

    Specifies the change: delete, insert, update_preimage, or update_postimage

  • _commit_version BIGINT NOT NULL

    Specifies the commit version of the table associated with the change.

  • _commit_timestamp TIMESTAMP NOT NULL

    Specifies the commit timestamp associated with the change.

If table_str does not represent a qualified table name the name is qualified with the value of current_schema. If the table name contains spaces or dots use back quotes within the string to quote that part of the name.

Examples

-- Create a Delta table with Change Data Feed;
> CREATE TABLE myschema.t(c1 INT, c2 STRING) TBLPROPERTIES(delta.enableChangeDataFeed=true);

-- Modify the table
> INSERT INTO myschema.t VALUES (1, 'Hello'), (2, 'World');
> INSERT INTO myschema.t VALUES (3, '!');
> UPDATE myschema.t SET c2 = upper(c2) WHERE c1 < 3;
> DELETE FROM myschema.t WHERE c1 = 3;

-- Show the history of table change events
> DESCRIBE HISTORY myschema.t;
 version timestamp                    userId           userName      operation    operationParameters                                            ...
       4 2022-09-01T18:32:35.000+0000 6167625779053302 alf@melmak.et DELETE       {"predicate":"[\"(spark_catalog.myschema.t.c1 = 3)\"]"}
       3 2022-09-01T18:32:32.000+0000 6167625779053302 alf@melmak.et UPDATE       {"predicate":"(c1#3195878 < 3)"}
       2 2022-09-01T18:32:28.000+0000 6167625779053302 alf@melmak.et WRITE        {"mode":"Append","partitionBy":"[]"}
       1 2022-09-01T18:32:26.000+0000 6167625779053302 alf@melmak.et WRITE        {"mode":"Append","partitionBy":"[]"}
       0 2022-09-01T18:32:23.000+0000 6167625779053302 alf@melmak.et CREATE TABLE {"isManaged":"true","description":null,"partitionBy":"[]","properties":"{\"delta.enableChangeDataFeed\":\"true\"}"}

-- Show the change table feed using a the commit timestamp retrieved from the history.
> SELECT * FROM table_changes('`myschema`.`t`', 2);
 c1 c2     _change_type    _commit_version _commit_timestamp
  3 !      insert                        2 2022-09-01T18:32:28.000+0000
  2 WORLD  update_postimage              3 2022-09-01T18:32:32.000+0000
  2 World  update_preimage               3 2022-09-01T18:32:32.000+0000
  1 Hello  update_preimage               3 2022-09-01T18:32:32.000+0000
  1 HELLO  update_postimage              3 2022-09-01T18:32:32.000+0000
  3 !      delete                        4 2022-09-01T18:32:35.000+0000

-- Show the ame change table feed using a point in time.
> SELECT * FROM table_changes('`myschema`.`t`', '2022-09-01T18:32:27.000+0000') ORDER BY _commit_version;
 c1 c2     _change_type    _commit_version _commit_timestamp
  3 !      insert                        2 2022-09-01T18:32:28.000+0000
  2 WORLD  update_postimage              3 2022-09-01T18:32:32.000+0000
  2 World  update_preimage               3 2022-09-01T18:32:32.000+0000
  1 Hello  update_preimage               3 2022-09-01T18:32:32.000+0000
  1 HELLO  update_postimage              3 2022-09-01T18:32:32.000+0000
  3 !      delete                        4 2022-09-01T18:32:35.000+0000