OceanBase Documentation Examples

This skill provides guidelines for creating SQL examples in OceanBase documentation.

Example structure

SQL statements

Always prefix with prompt:

• obclient> for default prompt
• obclient [SCHEMA]> when schema context is relevant

Include semicolons in executable statements.

Example:

obclient [KILL_USER]> SHOW PROCESSLIST;

Query results

Separate from SQL statements:

• Place SQL in one code block
• Place results in another code block
• Connect with descriptive text like “查询结果如下：” or “Query results:”

Example:

obclient [KILL_USER]> SHOW PROCESSLIST;

查询结果如下：

+------------+-----------+----------------------+-----------+---------+------+--------+------------------+
| ID         | USER      | HOST                 | DB        | COMMAND | TIME | STATE  | INFO             |
+------------+-----------+----------------------+-----------+---------+------+--------+------------------+
| 3221487726 | KILL_USER | 100.xx.xxx.xxx:34803 | KILL_USER | Query   |    0 | ACTIVE | SHOW PROCESSLIST |
+------------+-----------+----------------------+-----------+---------+------+--------+------------------+
1 row in set

Naming conventions

Use meaningful names

Avoid simple names:

• ❌ t1, t2, tg1, db1
• ❌ test_table, temp_db

Use business-meaningful names:

• ✅ Table groups: order_tg, product_tg, inventory_tg
• ✅ Tables: order_table, user_info, product_catalog
• ✅ Databases: sales_db, customer_db, warehouse_db

Why:

• Helps users understand real-world scenarios
• Makes examples more relatable
• Demonstrates practical applications

Example scenarios

E-commerce:

• Tables: orders, order_items, customers, products
• Table groups: order_tg, product_tg
• Databases: ecommerce_db

Financial:

• Tables: transactions, accounts, balances
• Table groups: transaction_tg, account_tg
• Databases: banking_db

Inventory:

• Tables: warehouses, inventory_items, stock_movements
• Table groups: warehouse_tg, inventory_tg
• Databases: logistics_db

What to include

Include:

• ✅ Query result tables (when helpful)
• ✅ Error messages (when demonstrating error handling)
• ✅ Descriptive output (when it adds value)

Exclude:

• ❌ “Query OK” messages
• ❌ “Query OK, 0 rows affected”
• ❌ “Query OK, 1 row affected”
• ❌ Generic success messages

Exception: Only include these if they provide meaningful information for understanding the example.

Example workflow

Step 1: Create context

Set up meaningful scenario:

obclient [SYS]> CREATE USER sales_user IDENTIFIED BY 'password123';
obclient [SYS]> GRANT CREATE SESSION TO sales_user;
obclient [SYS]> CREATE DATABASE sales_db;
obclient [SYS]> USE sales_db;

Step 2: Create objects

Use meaningful names:

obclient [SALES_DB]> CREATE TABLE order_table (
    order_id BIGINT PRIMARY KEY,
    customer_id BIGINT,
    order_date DATE,
    total_amount DECIMAL(10,2)
);

Step 3: Demonstrate feature

Show the SQL statement:

obclient [SALES_DB]> SELECT * FROM order_table WHERE order_date >= '2024-01-01';

Step 4: Show results

Separate code block with descriptive text:

查询结果如下：

+----------+-------------+------------+--------------+
| order_id | customer_id | order_date | total_amount |
+----------+-------------+------------+--------------+
|      101 |        1001 | 2024-01-15 |      1250.00 |
|      102 |        1002 | 2024-01-20 |       850.50 |
+----------+-------------+------------+--------------+
2 rows in set

Complex examples

Multi-step examples

For complex workflows, break into logical steps:

Step 1: Setup

obclient [SYS]> CREATE USER admin_user IDENTIFIED BY 'admin123';
obclient [SYS]> GRANT ALTER SYSTEM TO admin_user;

Step 2: Create table group

obclient [ADMIN_USER]> CREATE TABLEGROUP order_tg;

Step 3: Create table

obclient [ADMIN_USER]> CREATE TABLE order_table (
    order_id BIGINT PRIMARY KEY,
    customer_id BIGINT
) TABLEGROUP = order_tg;

Step 4: Verify

obclient [ADMIN_USER]> SHOW TABLEGROUPS;

查询结果如下：

+-----------+------------+
| TableName | TableGroup |
+-----------+------------+
| order_table | order_tg  |
+-----------+------------+
1 row in set

Error examples

When demonstrating error handling:

obclient [USER]> CREATE TABLE invalid_table (
    id INT PRIMARY KEY,
    name VARCHAR(10)
) PARTITION BY HASH(id) PARTITIONS 0;

错误信息如下：

ERROR 1235 (42000): Invalid partition count

Best practices

1. Start simple, add complexity gradually
2. Use consistent naming throughout example
3. Show realistic data in results
4. Include comments when helpful (but keep them concise)
5. Test examples to ensure they work
6. Follow test cases when they differ from parser definitions

Quality checklist

• SQL statements have obclient> prefix
• SQL and results in separate code blocks
• Meaningful, business-oriented names used
• No unnecessary “Query OK” messages
• Results formatted clearly
• Example demonstrates real-world usage
• All statements are executable and tested