dianfu commented on a change in pull request #13273: URL: https://github.com/apache/flink/pull/13273#discussion_r479712335
########## File path: docs/dev/python/user-guide/table/10_minutes_to_table_api.md ########## @@ -0,0 +1,712 @@ +--- +title: "10 Minutes to Table API" +nav-parent_id: python_tableapi +nav-pos: 25 +--- +<!-- +Licensed to the Apache Software Foundation (ASF) under one +or more contributor license agreements. See the NOTICE file +distributed with this work for additional information +regarding copyright ownership. The ASF licenses this file +to you under the Apache License, Version 2.0 (the +"License"); you may not use this file except in compliance +with the License. You may obtain a copy of the License at + + http://www.apache.org/licenses/LICENSE-2.0 + +Unless required by applicable law or agreed to in writing, +software distributed under the License is distributed on an +"AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY +KIND, either express or implied. See the License for the +specific language governing permissions and limitations +under the License. +--> + +This document is a short introduction to PyFlink Table API, which is used to help novice users quickly understand the basic usage of PyFlink Table API. +For advanced usage, please refer to other documents in this User Guide. + +* This will be replaced by the TOC +{:toc} + +Common Structure of Python Table API Program +-------------------------------------------- + +All Table API and SQL programs for batch and streaming follow the same pattern. The following code example shows the common structure of Table API and SQL programs. + +{% highlight python %} + +from pyflink.table import EnvironmentSettings, StreamTableEnvironment + +# 1. create a TableEnvironment +table_env = StreamTableEnvironment.create(environment_settings=EnvironmentSettings.new_instance().in_streaming_mode().use_blink_planner().build()) + +# 2. create source Table +table_env.execute_sql(""" +CREATE TABLE datagen ( + id INT, + data STRING +) WITH ( + 'connector' = 'datagen', + 'fields.id.kind' = 'sequence', + 'fields.id.start' = '1', + 'fields.id.end' = '10' +) +""") + +# 3. create sink Table +table_env.execute_sql(""" +CREATE TABLE print ( + id INT, + data STRING +) WITH ( + 'connector' = 'print' +) +""") + +# 4. query from source table and caculate +# create a Table from a Table API query: +tapi_result = table_env.from_path("datagen").select("id + 1, data") +# or create a Table from a SQL query: +sql_result = table_env.sql_query("SELECT * FROM datagen").select("id + 1, data") + +# 5. emit query result to sink table +# emit a Table API result Table to a sink table: +tapi_result.execute_insert("print").get_job_client().get_job_execution_result().result() +sql_result.execute_insert("print").get_job_client().get_job_execution_result().result() +# or emit results via SQL query: +table_env.execute_sql("INSERT INTO print SELECT * FROM datagen").get_job_client().get_job_execution_result().result() + +{% endhighlight %} + +{% top %} + +Create a TableEnvironment +------------------------- + +The `TableEnvironment` is a central concept of the Table API and SQL integration. The following code example shows how to create a TableEnvironment: + +{% highlight python %} + +from pyflink.table import EnvironmentSettings, StreamTableEnvironment, BatchTableEnvironment + +# create a blink streaming TableEnvironment +table_env = StreamTableEnvironment.create(environment_settings=EnvironmentSettings.new_instance().in_streaming_mode().use_blink_planner().build()) + +# create a blink batch TableEnvironment +table_env = BatchTableEnvironment.create(environment_settings=EnvironmentSettings.new_instance().in_batch_mode().use_blink_planner().build()) + +# create a flink streaming TableEnvironment +table_env = StreamTableEnvironment.create(environment_settings=EnvironmentSettings.new_instance().in_streaming_mode().use_old_planner().build()) + +# create a flink batch TableEnvironment +table_env = BatchTableEnvironment.create(environment_settings=EnvironmentSettings.new_instance().in_batch_mode().use_old_planner().build()) + +{% endhighlight %} + +The `TableEnvironment` is responsible for: + +* Creating `Table`s +* Registering `Table`s to the catalog +* Executing SQL queries +* Registering user-defined (scalar, table, or aggregation) functions +* Offering further configuration options. +* Add Python dependencies to support running Python UDF on remote cluster +* Executing jobs. + +Currently there are 2 planners available: flink planner and blink planner. + +You should explicitly set which planner to use in the current program. +We recommend using the blink planner as much as possible. +The blink planner is more powerful in functionality and performance, and the flink planner is reserved for compatibility. + +{% top %} + +Create Tables +------------- + +`Table` is the core component of the Table API. A `Table` represents a intermediate result set during a Table API Job. + +A `Table` is always bound to a specific `TableEnvironment`. It is not possible to combine tables of different TableEnvironments in same query, e.g., to join or union them. + +### Create From A List Object + +You can create a Table from a list object: + +{% highlight python %} + +# create a blink batch TableEnvironment +from pyflink.table import EnvironmentSettings, BatchTableEnvironment +table_env = BatchTableEnvironment.create(environment_settings=EnvironmentSettings.new_instance().in_batch_mode().use_blink_planner().build()) + +table = table_env.from_elements([(1, 'Hi'), (2, 'Hello')]) +table.to_pandas() + +{% endhighlight %} + +The result is: + +{% highlight text %} + _1 _2 +0 1 Hi +1 2 Hello +{% endhighlight %} + +You can also create the Table with column names: + +{% highlight python %} + +table = table_env.from_elements([(1, 'Hi'), (2, 'Hello')], ['id', 'data']) +table.to_pandas() + +{% endhighlight %} + +The result is: + +{% highlight text %} + id data +0 1 Hi +1 2 Hello +{% endhighlight %} + +By default the table schema is extracted from the data automatically. + +If the table schema is not as your wish, you can specify it manually: + +{% highlight python %} + +table_without_schema = table_env.from_elements([(1, 'Hi'), (2, 'Hello')], ['id', 'data']) +# by default the type of the "id" column is 64 bit int +default_type = table_without_schema.to_pandas()["id"].dtype +print('By default the type of the "id" column is %s.' % default_type) + +from pyflink.table import DataTypes +table = table_env.from_elements([(1, 'Hi'), (2, 'Hello')], + DataTypes.ROW([DataTypes.FIELD("id", DataTypes.TINYINT()), + DataTypes.FIELD("data", DataTypes.STRING())])) +# now the type of the "id" column is 8 bit int +type = table.to_pandas()["id"].dtype +print('Now the type of the "id" column is %s.' % type) + +{% endhighlight %} + +The result is: + +{% highlight text %} +By default the type of the "id" column is int64. +Now the type of the "id" column is int8. +{% endhighlight %} + +### Create From Connectors + +You can create a Table from connector DDL: + +{% highlight python %} + +table_env.execute_sql(""" + CREATE TABLE random_source ( + id BIGINT, + data TINYINT) + WITH ( + 'connector' = 'datagen', + 'fields.id.kind'='sequence', + 'fields.id.start'='1', + 'fields.id.end'='3', + 'fields.data.kind'='sequence', + 'fields.data.start'='4', + 'fields.data.end'='6' + ) +""") +table = table_env.from_path("random_source") +table.to_pandas() + +{% endhighlight %} + +The result is: + +{% highlight text %} + id data +0 2 5 +1 1 4 +2 3 6 +{% endhighlight %} + +### Create From Catalog + +A TableEnvironment maintains a map of catalogs of tables which are created with an identifier. + +The tables in catalog may either be temporary, and tied to the lifecycle of a single Flink session, or permanent, and visible across multiple Flink sessions and clusters. + +The tables and views created via SQL DDL, e.g. "create table ..." and "create view ..." is also stored in catalog. + +You can also access the tables in catalog via SQL directly. + +If you want to use the catalog tables in Table API, you can use the "from_path" method to create the Table API objects from catalog: + +{% highlight python %} + +# prepare the catalog +# register Table API tables to catalog +table = table_env.from_elements([(1, 'Hi'), (2, 'Hello')], ['id', 'data']) +table_env.create_temporary_view('source_table', table) + +# create Table API table from catalog +new_table = table_env.from_path('source_table') +new_table.to_pandas() + +{% endhighlight %} + +The result is: + +{% highlight text %} + id data +0 1 Hi +1 2 Hello +{% endhighlight %} + +{% top %} + +Write Queries +------------- + +### Write Table API Queries + +The `Table` object offers many methods to apply relational operations. +These methods return a new `Table` object, which represents the result of applying the relational operation on the input `Table`. +i.e. you can make a method chaining when using Table API. +Some relational operations are composed of multiple method calls such as table.groupBy(...).select(), where groupBy(...) specifies a grouping of table, and select(...) the projection on the grouping of table. + +The [Table API]({{ site.baseurl }}/dev/table/tableApi.html) document describes all Table API operations that are supported on streaming and batch tables. Review comment: It seems that this is not updated in the PR. ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org
