Skip to main content Link Menu Expand (external link) Document Search Copy Copied

CREATE TABLE statement

Create a FeatureBase table with the specified columns and data types.

BNF diagrams

expr

COLUMN_LIST

expr

Data types

expr

Table_options

expr

Syntax 

CREATE TABLE
  [IF NOT EXISTS]
  <table_name>
  (COLUMN_LIST)
  [WITH COMMENT 'comment'];

Arguments

Argument Description Required? Additional information
<table_name> String literal table identifier Yes Naming standards
IF NOT EXISTS Stop statement execution if a table of the same name already exists Optional  
COLUMN_LIST List of column names, data types and optional constraints with leading _id column that can be ID or STRING data type Yes * _id column
* Naming standards
* ID data type
* STRING data type
WITH COMMENT Optional string literal that describes the table No  

Additional information

Naming standards

FeatureBase identifiers (including object names such as databases, tables and columns) start with a lower-case alphabetic character and can include:

  • lower-case alphabetic characters
  • numbers 0-9

  • dash - and underscore _ characters.

  • table names can be up to 230 characters in length

  • column names can be up to 64 characters in length

_id column

The _id column in each table represents the primary key and supports two data types:

Data type Description Example Advantage Disadvantage
ID data type One or more contiguous blocks of positive integers 7,8,9,10,11,12,...20,21,22,23,... Faster queries Warning: performance and storage are adversely affected if values are non-contiguous
String data type String literal values can be supplied in any order if used as primary key “a23”, “s93kk”, “h82k”, “2023-02-10”,… Automatically mapped to blocks of contiguous integers Higher storage and performance costs

CREATE TABLE examples

Statement Data types Table options Relates to
CREATE TABLE cosvec-target * ID
* STRING
* VECTOR		   INSERT INTO cosvec-target
CREATE TABLE csv-target * INT
* STRING		   BULK INSERT csv-target
CREATE TABLE ndjson-target * BOOL
* ID
* STRING
* TIMESTAMP		 * IF NOT EXISTS
* WITH COMMENT		 BULK INSERT ndjson-target
CREATE TABLE orc-target * BOOL
* ID
* INT
* STRING		 * IF NOT EXISTS
* WITH COMMENT		 BULK INSERT orc-target
CREATE TABLE parquet-target * ID
* INT
* DECIMAL		   BULK INSERT parquet-target
CREATE TABLE tuple-set-target * IDSET
* IDSETQ
* INT
* STRING
* STRINGSET
* STRINGSETQ		   BULK INSERT tuple-set-target
CREATE TABLE tan-target * ID
* STRINGSET		   INSERT INTO tan-target

Further information