Skip to main content

Developer Guide

Follow the instructions in the README of the RookDB/RookDB to run the Storage Engine. After the system starts, the interface will display a list of available operations to choose from.

The Storage Engine currently provides the following options:

Show Databases​

Displays all databases available in the catalog with metadata (owner, creation timestamp).

Create Database​

Creates a new database with metadata and updates the catalog.

Steps:

  1. Enter a database name when prompted
  2. Enter the owner name (or use the default)
  3. The database is registered in pg_database and the data directory is created

Example:

Database name: users
Owner: admin

Select Database​

Sets the active database for performing operations.

Steps:

  1. Enter a database name from the displayed list

Show Tables​

Displays all tables in the selected database with statistics (row count, page count, creation timestamp).

Create Table​

Creates a new table with a schema and optional constraints.

Steps:

  1. Enter table name
  2. Enter columns using format:
column_name:data_type[:constraint]
  1. Press Enter on an empty line to finish

Supported Types:

  • INT (alias: INTEGER, INT32)
  • BIGINT (alias: INT64)
  • FLOAT (alias: REAL, FLOAT32)
  • DOUBLE (alias: FLOAT64)
  • BOOL (alias: BOOLEAN)
  • TEXT (alias: STRING)
  • VARCHAR(n)
  • DATE
  • TIMESTAMP
  • BYTES (alias: BYTEA, BLOB)

Supported Inline Constraints:

  • PRIMARY KEY — unique, not-null constraint with a backing B-Tree index
  • NOT NULL — disallows NULL values

Example:

id:INT:PRIMARY KEY
name:TEXT:NOT NULL
email:VARCHAR(100)

After defining columns, you can also define table-level constraints such as composite primary keys and foreign keys.

Load CSV​

Loads CSV data into an existing table. Constraint validation is performed during loading — rows that violate PRIMARY KEY, UNIQUE, FOREIGN KEY, or NOT NULL constraints are rejected.

Steps:

  1. Enter table name
  2. Enter CSV file path

Example:

examples/example.csv

Select Tuples​

=============================
Choose an option:
...
8. Select tuples
...
=============================
Enter your choice: 7
Enter table name: employees
Enter SQL (single SELECT with WHERE): SELECT * WHERE id > 4

=== Tuples in 'company.employees' ===
Total pages: 3

id (INT) | name (VARCHAR(10))
Tuple 1: id=5 name='xJIsV'
Tuple 2: id=6 name='KzCIUYD'
Tuple 3: id=9 name='Alice'

=== End of tuples ===

Output Format​

=== Tuples in '<db>.<table>' ===
Total pages: <num>

<col1_name> (<col1_type>) | <col2_name> (<col2_type>) | ...
Tuple 1: col1=<val> col2=<val> ...
Tuple 2: col1=<val> col2=<val> ...
...

=== End of tuples ===
  • NULL values are printed as colname=NULL.
  • Decode errors are printed as <decode-error: ...>.
  • Tuple numbering is 1-based and continuous across the filtered results.

Show Table Statistics​

Displays storage statistics like total number of pages.

Exit​

Exit from RookDB.