Importing Data from Parquet

-- Autodetect based on file suffix
SELECT * FROM 'test.parquet';

-- Use parquet_view function
SELECT * FROM parquet_view('test.parquet');

 a  | b  
----+----
  1 |  1
  2 |  2
  3 |  3
 .. | ..
(100 rows)

CREATE TABLE test AS (SELECT * FROM 'test.parquet');

CREATE TABLE test AS (SELECT * FROM parquet_view('test'));

CREATE TABLE test (a integer, b integer);

-- Use COPY in SQL (relative path to server)
COPY test from 'test.parquet' (format parquet);

-- Or \copy in psql (relative path to client)
\COPY test from 'test.parquet' (format parquet);

-- Create a table with the columns you need
CREATE TABLE onecol (c integer);

-- Select the columns you need for import
INSERT INTO onecol (SELECT a FROM 'test.parquet')

SELECT * FROM parquet_schema('test.parquet');

  file_name   |  name  | cedar_type |  type   | type_length | repetition_type | num_children | converted_type | scale | precision | field_id | logical_type 
--------------+--------+------------+---------+-------------+-----------------+--------------+----------------+-------+-----------+----------+--------------
 test.parquet | schema | boolean    | BOOLEAN |             | REQUIRED        |            2 |                |       |           |          | 
 test.parquet | a      | bigint     | INT64   |             | OPTIONAL        |            0 |                |       |           |          | 
 test.parquet | b      | bigint     | INT64   |             | OPTIONAL        |            0 |                |       |           |          | 
(3 rows)

SELECT * FROM parquet_file('test.parquet');

  file_name   | file_size | file_meta_size |            created_by            | num_rows | num_row_groups | num_columns | format_version 
--------------+-----------+----------------+----------------------------------+----------+----------------+-------------+----------------
 test.parquet | 2029      | 554            | parquet-cpp-arrow version 21.0.0 | 100      | 1              | 3           |              2
(1 row)

SELECT * FROM parquet_rowgroups('test.parquet');

  file_name   | id | num_rows | num_cols | total_bytes | compressed_bytes 
--------------+----+----------+----------+-------------+------------------
 test.parquet |  0 |      100 |        2 |        1463 |             1463

SELECT * FROM parquet_colchunks('test.parquet');

  file_name   | rg_id | col_id | type  | file_path | schema_path |         encodings          |    codec     | num_values | total_bytes | compressed_bytes 
--------------+-------+--------+-------+-----------+-------------+----------------------------+--------------+------------+-------------+------------------
 test.parquet |     0 |      0 | INT64 |           | {a}         | {PLAIN,RLE,RLE_DICTIONARY} | UNCOMPRESSED |        100 |         771 |              771
 test.parquet |     0 |      1 | INT64 |           | {b}         | {PLAIN,RLE,RLE_DICTIONARY} | UNCOMPRESSED |        100 |         692 |              692