Commands and SQL statements executable in Firebird isql WIP

firebird.sql

SHOW TABLE EMPLOYEE; -- isql-specific command

SELECT EMP_NO, FIRST_NAME FROM EMPLOYEE;
SELECT EMP_NO, FULL_NAME FROM EMPLOYEE;

SELECT * FROM EMPLOYEE;

SELECT EMP_NO, LAST_NAME, SALARY, SALARY + 500 FROM EMPLOYEE;

SELECT DISTINCT DEPT_NO FROM EMPLOYEE;

SELECT LAST_NAME, DEPT_NO FROM EMPLOYEE WHERE DEPT_NO = 600;
SELECT EMP_NO, LAST_NAME, DEPT_NO FROM EMPLOYEE WHERE LAST_NAME = '';
SELECT LAST_NAME, DEPT_NO FROM EMPLOYEE
  WHERE LAST_NAME = 'O''Brien'; -- how to escape quotes

SELECT EMP_NO, LAST_NAME, HIRE_DATE FROM EMPLOYEE WHERE HIRE_DATE = '1990-05-01';
SELECT EMP_NO, LAST_NAME, HIRE_DATE FROM EMPLOYEE WHERE HIRE_DATE != '1990-05-01';
SELECT EMP_NO, LAST_NAME, HIRE_DATE FROM EMPLOYEE WHERE HIRE_DATE > '1990-05-01';

SELECT EMP_NO, LAST_NAME, DEPT_NO FROM EMPLOYEE WHERE DEPT_NO BETWEEN 600 AND 700;
SELECT EMP_NO, LAST_NAME, DEPT_NO FROM EMPLOYEE
  WHERE DEPT_NO IN (000, 100, 200, 300, 400, 500, 600);

SELECT EMP_NO, LAST_NAME FROM EMPLOYEE
  WHERE LAST_NAME LIKE 'B%';

SELECT EMP_NO, LAST_NAME, PHONE_EXT FROM EMPLOYEE
  WHERE PHONE_EXT IS NULL;

SELECT EMP_NO, LAST_NAME, SALARY, DEPT_NO FROM EMPLOYEE
  WHERE DEPT_NO = 671 AND SALARY > 50000;
SELECT EMP_NO, LAST_NAME, SALARY, DEPT_NO FROM EMPLOYEE
  WHERE DEPT_NO = 671 OR SALARY > 50000;

SELECT EMP_NO, LAST_NAME, DEPT_NO FROM EMPLOYEE
  ORDER BY DEPT_NO;
SELECT EMP_NO, LAST_NAME, DEPT_NO FROM EMPLOYEE
  ORDER BY DEPT_NO ASC, LAST_NAME DESC;
SELECT EMP_NO, LAST_NAME, PHONE_EXT FROM EMPLOYEE
  ORDER BY PHONE_EXT DESC; -- null の位置に注意
SELECT EMP_NO, LAST_NAME, PHONE_EXT FROM EMPLOYEE
  ORDER BY PHONE_EXT DESC NULLS FIRST;

SELECT FULL_NAME FROM EMPLOYEE WHERE LOWER(LAST_NAME) = 'green';
SELECT 'a' || 'b' FROM RDB$DATABASE; -- equivalent to Oracle's DUAL
SELECT SUBSTRING('Firebird' from 4 for 3) from RDB$DATABASE; -- 'ebi'
SELECT CHAR_LENGTH('Firebird') from RDB$DATABASE;
SELECT POSITION('A' in 'Firebird') from RDB$DATABASE; -- 0 if not found
SELECT LPAD(PHONE_EXT, 3, ' ') FROM EMPLOYEE;
SELECT TRIM(both 'O' from 'OEDO') FROM RDB$DATABASE;
SELECT TRIM(leading 'O' from 'OEDO') FROM RDB$DATABASE;
SELECT REPLACE('ORACLE', 'O', 'MI') FROM RDB$DATABASE;
SELECT REPLACE('ORACLE', 'O', '') FROM RDB$DATABASE;

SELECT LAST_NAME, ROUND(SALARY, -2) FROM EMPLOYEE;
SELECT LAST_NAME, TRUNC(SALARY) FROM EMPLOYEE; -- 0
SELECT MOD(50, 6) FROM RDB$DATABASE;

SELECT HIRE_DATE FROM EMPLOYEE;
SELECT CURRENT_DATE FROM RDB$DATABASE; -- equivalent to Oracle's SYSDATE
SELECT DATEDIFF(month, date '2009-08-15', date '2009-03-15') FROM RDB$DATABASE;
SELECT DATEADD(5 day to date '2009-08-15') FROM RDB$DATABASE;
SELECT LAST_DAY(of month from date '2009-08-15') FROM RDB$DATABASE;

SELECT CAST('today' as date) FROM RDB$DATABASE;
SELECT CAST(CURRENT_DATE as varchar(10)) FROM RDB$DATABASE;
SELECT CAST(CURRENT_TIMESTAMP as date) FROM RDB$DATABASE;

SELECT LAST_NAME, COALESCE(PHONE_EXT, 'N/A') FROM EMPLOYEE;

SELECT LAST_NAME, SALARY,
CASE
  WHEN (SALARY BETWEEN 0.0 AND 50000.0) THEN 'C'
  WHEN (SALARY BETWEEN 50000.0 AND 100000.0) THEN 'B'
  ELSE 'A'
END GRADE
FROM EMPLOYEE;

SELECT LAST_NAME, DECODE(JOB_GRADE,
  0, 'Novice',
  1, 'Aspirant',
  2, 'Battler',
  3, 'Fighter',
  4, 'Adept',
  5, 'Chevalier',
  6, 'Veteran',
  'Warrior')
FROM EMPLOYEE;

SELECT AVG(SALARY), SUM(SALARY), MIN(SALARY), MAX(SALARY), COUNT(SALARY)
  FROM EMPLOYEE;
SELECT MAX(FIRST_NAME), MIN(FIRST_NAME), COUNT(FIRST_NAME) FROM EMPLOYEE;

SELECT DEPT_NO, AVG(SALARY) FROM EMPLOYEE GROUP BY DEPT_NO;
SELECT JOB_CODE, AVG(SALARY) FROM EMPLOYEE GROUP BY JOB_CODE;

SELECT DEPT_NO, JOB_CODE, AVG(SALARY) FROM EMPLOYEE
  WHERE JOB_CODE = 'SRep'
  GROUP BY DEPT_NO, JOB_CODE
  HAVING AVG(SALARY) >= 50000;
SELECT DEPT_NO, AVG(SALARY) FROM EMPLOYEE
  GROUP BY DEPT_NO
  HAVING COUNT(1) > 2;
SELECT MIN(AVG(SALARY)) FROM EMPLOYEE
  GROUP BY DEPT_NO; -- ERROR -104

----------------------------------------------------------------------

SELECT EMP_NO, LAST_NAME, DEPARTMENT
  FROM EMPLOYEE NATURAL JOIN DEPARTMENT;
SELECT EMP_NO, FULL_NAME, JOB_TITLE
  FROM EMPLOYEE NATURAL JOIN JOB;

SELECT EMP_NO, LAST_NAME, DEPARTMENT
  FROM EMPLOYEE
  JOIN DEPARTMENT USING (DEPT_NO);
SELECT PO_NUMBER, SALES_REP, CUSTOMER, CURRENCY
  FROM SALES
  JOIN CUSTOMER USING(CUST_NO)
  JOIN COUNTRY USING(COUNTRY);
SELECT EMP_NO, LAST_NAME, DEPARTMENT
  FROM EMPLOYEE
  JOIN DEPARTMENT USING (DEPT_NO)
  WHERE DEPT_NO = 600;

SELECT PO_NUMBER, CUST_NO, CUSTOMER
  FROM SALES
  JOIN CUSTOMER USING (CUST_NO);

SELECT PO_NUMBER, SALES_REP, FULL_NAME
  FROM SALES
  JOIN EMPLOYEE ON SALES.SALES_REP = EMPLOYEE.EMP_NO;
SELECT PO_NUMBER, SALES_REP, CUSTOMER, CURRENCY
  FROM SALES
  JOIN CUSTOMER ON SALES.CUST_NO = CUSTOMER.CUST_NO
  JOIN COUNTRY ON CUSTOMER.COUNTRY = COUNTRY.COUNTRY;

SELECT d.DEPT_NO, d.DEPARTMENT, m.DEPT_NO, m.DEPARTMENT
  FROM DEPARTMENT d
  JOIN DEPARTMENT m ON d.HEAD_DEPT = m.DEPT_NO;

SELECT e.EMP_NO, e.LAST_NAME, e.SALARY, h.EMP_NO
  FROM EMPLOYEE e JOIN SALARY_HISTORY h
  ON e.SALARY BETWEEN h.OLD_SALARY AND h.NEW_SALARY; -- 意味はない

SELECT e.EMP_NO, e.LAST_NAME, d.DEPARTMENT
  FROM EMPLOYEE e RIGHT OUTER JOIN DEPARTMENT d
  USING (DEPT_NO); -- WHERE e.EMP_NO IS NULL;
SELECT e.EMP_NO, e.LAST_NAME, d.DEPARTMENT
  FROM DEPARTMENT d LEFT OUTER JOIN EMPLOYEE e
  USING (DEPT_NO); -- WHERE e.EMP_NO IS NULL;
SELECT e.LAST_NAME, d.DEPARTMENT
  FROM EMPLOYEE e CROSS JOIN DEPARTMENT d; -- 意味は当然ない

----------------------------------------------------------------------

SELECT LAST_NAME, SALARY FROM EMPLOYEE WHERE SALARY >=
  (SELECT SALARY FROM EMPLOYEE WHERE LAST_NAME = 'Green');

SELECT * FROM
  (SELECT AVG(SALARY) SALARY_MEAN, DEPT_NO
    FROM EMPLOYEE
    GROUP BY DEPT_NO);

SELECT LAST_NAME FROM EMPLOYEE
  WHERE SALARY > (SELECT AVG(SALARY) FROM EMPLOYEE); -- why only Japanese?

----------------------------------------------------------------------

SELECT LAST_NAME, JOB_CODE, JOB_COUNTRY FROM EMPLOYEE
  WHERE JOB_CODE IN
    (SELECT JOB_CODE FROM EMPLOYEE
      WHERE JOB_COUNTRY = 'USA');
SELECT LAST_NAME, SALARY, DEPT_NO FROM EMPLOYEE
  WHERE SALARY < ANY -- ALL も試そう
    (SELECT SALARY FROM EMPLOYEE WHERE DEPT_NO = 623);
    AND DEPT_NO != 623
  ORDER BY DEPT_NO;

SELECT EMP_NO, LAST_NAME FROM EMPLOYEE
  WHERE DEPT_NO IN (
    SELECT DEPT_NO FROM DEPARTMENT
      WHERE DEPARTMENT IN ('Customer Support', 'Customer Service'));

----------------------------------------------------------------------

SELECT DEPT_NO, MIN(SALARY) FROM EMPLOYEE
  GROUP BY DEPT_NO
  HAVING MIN(SALARY) > (
    SELECT MIN(SALARY) FROM EMPLOYEE WHERE DEPT_NO = 623);

----------------------------------------------------------------------

SELECT DEPT_NO, DEPARTMENT FROM DEPARTMENT
  WHERE DEPT_NO IN (
    SELECT HEAD_DEPT FROM DEPARTMENT);
SELECT DEPT_NO, DEPARTMENT FROM DEPARTMENT
  WHERE DEPT_NO NOT IN (
    SELECT HEAD_DEPT FROM DEPARTMENT
      WHERE HEAD_DEPT IS NOT NULL);

----------------------------------------------------------------------

SELECT JOB_CODE, JOB_GRADE, JOB_COUNTRY FROM EMPLOYEE
UNION -- ALL も試そう
SELECT JOB_CODE, JOB_GRADE, JOB_COUNTRY FROM JOB;

-- Use below instead of INTERSECT:
SELECT DISTINCT JOB_CODE, JOB_GRADE, JOB_COUNTRY
  FROM EMPLOYEE JOIN JOB
  USING (JOB_CODE, JOB_GRADE, JOB_COUNTRY);

-- Use below instead of MINUS or EXCEPT:
SELECT DISTINCT e.JOB_CODE, e.JOB_GRADE, e.JOB_COUNTRY
  FROM EMPLOYEE e
  LEFT OUTER JOIN JOB j
  USING (JOB_CODE, JOB_GRADE, JOB_COUNTRY)
  WHERE j.JOB_CODE IS NULL;

----------------------------------------------------------------------

INSERT INTO COUNTRY (COUNTRY, CURRENCY) VALUES ('Korea', 'Won');
INSERT INTO COUNTRY VALUES ('Korea', 'Won');

INSERT INTO SALARY_HISTORY
  (EMP_NO, CHANGE_DATE, UPDATER_ID, OLD_SALARY)
  VALUES
  (138, CURRENT_TIMESTAMP - 1, 'updater', 20.0);
INSERT INTO SALARY_HISTORY
  (EMP_NO, CHANGE_DATE, UPDATER_ID, OLD_SALARY)
  VALUES
  (138, '2023-11-25', LOWER(USER), 30.0);

-- INSERT INTO SELECT のための一時テーブルを作成
CREATE TABLE ENGINEER_EMPLOYEE (
  EMP_NO SMALLINT NOT NULL,
  FIRST_NAME VARCHAR(15) NOT NULL,
  LAST_NAME VARCHAR(20) NOT NULL,
  DEPT_NO CHAR(3) NOT NULL,
  CONSTRAINT PK_ENGINEER PRIMARY KEY (EMP_NO)
);
-- DROP TABLE ENGINEER_EMPLOYEE;

INSERT INTO ENGINEER_EMPLOYEE
   SELECT EMP_NO, FIRST_NAME, LAST_NAME, DEPT_NO FROM EMPLOYEE
   WHERE JOB_CODE = 'Eng';

UPDATE SALARY_HISTORY
  SET UPDATER_ID = LOWER(UPDATER_ID)
  WHERE UPDATER_ID = USER;
UPDATE EMPLOYEE
  SET LAST_NAME = 'Scott';

UPDATE EMPLOYEE SET
  DEPT_NO = (
    SELECT DEPT_NO FROM EMPLOYEE WHERE EMP_NO = 138),
  SALARY = (
    SELECT MAX(SALARY) FROM EMPLOYEE WHERE JOB_CODE = 'Eng'); -- ある理由で失敗するが後学のために実行するといい

DELETE FROM SALARY_HISTORY;
DELETE FROM EMPLOYEE
  WHERE DEPT_NO = (
    SELECT DEPT_NO FROM EMPLOYEE WHERE EMP_NO = 110)
  AND SALARY = (SELECT MAX(SALARY) FROM EMPLOYEE WHERE JOB_CODE = 'Eng'); -- これも

----------------------------------------------------------------------
-- COMMIT/ROLLBACK

INSERT INTO COUNTRY VALUES ('Korea', 'Won');
INSERT INTO COUNTRY VALUES ('Macau', 'Pataca');
COMMIT; -- or ROLLBACK;

-- SAVEPOINT then ROLLBACK TO SAVEPOINT

INSERT INTO COUNTRY VALUES ('Korea', 'Won');
INSERT INTO COUNTRY VALUES ('Macau', 'Pataca');
SAVEPOINT A;
INSERT INTO COUNTRY VALUES('Israel', 'Shekel');
INSERT INTO COUNTRY VALUES('Nigeria', 'Naira');
ROLLBACK TO A;

----------------------------------------------------------------------
-- Lock

-- 試すためにはユーザーアカウントを追加する必要がある。
-- SYSDBA で localhost:employee に接続して次を実行：
-- CREATE USER username PASSWORD '<password>';

-- User A:
SELECT EMP_NO, LAST_NAME FROM EMPLOYEE FOR UPDATE;
UPDATE EMPLOYEE SET LAST_NAME = 'XXXX' WHERE EMP_NO = 138;

-- User B:
UPDATE EMPLOYEE SET LAST_NAME = 'YYYY' WHERE EMP_NO = 138;
-- Waiting...

-- User A:
COMMIT; -- or ROLLBACK

-- User B:
-- Execution...

----------------------------------------------------------------------
-- TRUNCATE TABLE: Firebird にはこの構文がない。

-- CREATE TABLE PEOPLE (
--   id int not null,
--   nickname varchar(12) not null,
--   country char(4),
--   constraint pk_people primary key (id),
--   constraint uk_nickname unique (nickname) using index ix_nick
-- );

----------------------------------------------------------------------
-- TABLE
-- 表コマンド実行には権限が必要だ。SYSDBA でログインしたセッションにて
-- GRANT CREATE TABLE TO USER <username>; しておくこと。

CREATE TABLE EMP(
  EMP_NO SMALLINT,
  LAST_NAME VARCHAR(20),
  EMAIL VARCHAR(30),
  DEPT_NO CHAR(3)
);
-- show table emp;

--CREATE TABLE COPY_EMP(EMP_NO, LAST_NAME, EMAIL, DEPT_NO)
--  AS SELECT EMP_NO, LAST_NAME, NULL, DEPT_NO FROM EMPLOYEE;
-- SELECT * FROM COPY_EMP;

CREATE TABLE EMP_DEF(
  EMP_NO SMALLINT,
  LAST_NAME VARCHAR(20),
  HIRE_DATE TIMESTAMP DEFAULT 'NOW',
  DEPT_NO CHAR(3)
);

-- 不明
-- ALTER TABLE DEPARTMENT READ ONLY;

DROP TABLE EMP;
DROP TABLE COPY_EMP;
DROP TABLE EMP_DEF;

----------------------------------------------------------------------
-- NOT NULL, UNIQUE, PK, FK, CHECK

CREATE TABLE EMP(
  EMP_NO SMALLINT NOT NULL,
  LAST_NAME VARCHAR(20),
  EMAIL VARCHAR(30),
  DEPT_NO CHAR(3)
);
DROP TABLE EMP;

CREATE TABLE DEPT(
  DEPT_NO CHAR(3) UNIQUE,
  DEPARTMENT VARCHAR(25)
);
DROP TABLE DEPT;

CREATE TABLE DEPT(
  DEPT_NO CHAR(3) PRIMARY KEY,
  DEPARTMENT VARCHAR(25)
);
DROP TABLE DEPT;

-- 制約を一列に二つ付ける場合
CREATE TABLE DEPT(
  DEPT_NO CHAR(3) NOT NULL,
  DEPARTMENT VARCHAR(25) UNIQUE,
  PRIMARY KEY (DEPT_NO)
);
DROP TABLE DEPT;

CREATE TABLE DEPT(
  DEPT_NO CHAR(3),
  DEPARTMENT VARCHAR(25),
  MNGR_NO SMALLINT REFERENCES EMPLOYEE(EMP_NO)
);
DROP TABLE DEPT;

CREATE TABLE DEPT(
  DEPT_NO CHAR(3) CHECK (
    DEPT_NO = '000' OR (DEPT_NO > '0' AND DEPT_NO <= '999') OR DEPT_NO IS NULL),
  DEPARTMENT VARCHAR(25)
);
DROP TABLE DEPT;

----------------------------------------------------------------------
-- View の作成にも GRANT が必要であることに注意。

CREATE VIEW EMP_V AS
  SELECT EMP_NO, LAST_NAME, HIRE_DATE, DEPT_NO FROM EMPLOYEE
    WHERE JOB_CODE = 'Eng';
-- show table EMP_V;

CREATE OR ALTER VIEW EMP_V (EID, LAST_NAME, HIRE_DATE, DID) AS
  SELECT EMP_NO, LAST_NAME, HIRE_DATE, DEPT_NO FROM EMPLOYEE
    WHERE JOB_CODE = 'Eng';
-- show table EMP_V;

CREATE VIEW DEPT_V (NAME, MAX_SALARY) AS
  SELECT d.DEPARTMENT, MAX(e.SALARY)
    FROM EMPLOYEE e JOIN DEPARTMENT d
    USING (DEPT_NO)
    GROUP BY d.DEPARTMENT;
-- SELECT * FROM DEPT_V;

CREATE OR ALTER VIEW COUNTRY_V AS
  SELECT * FROM COUNTRY
  WHERE CURRENCY NOT LIKE 'E%'
  WITH CHECK OPTION; -- CONSTRAINT は指定不能
UPDATE COUNTRY_V SET CURRENCY = 'En' WHERE CURRENCY = 'Yen'; -- expectedly fail

-- Oracle の WITH READ ONLY はない。

DROP VIEW EMP_V;
DROP VIEW DEPT_V;
DROP VIEW COUNTRY_V;

----------------------------------------------------------------------
-- SEQUENCE の作成（どころか閲覧）にも GRANT を要する。

CREATE SEQUENCE EMP_SEQ
  START WITH 0 INCREMENT BY 1;  -- EXT VALUE FOR seq_name expression.
-- show sequence EMP_SEQ;

SELECT NEXT VALUE FOR EMP_SEQ FROM RDB$DATABASE;

ALTER SEQUENCE EMP_SEQ
  RESTART WITH 100;
SELECT NEXT VALUE FOR EMP_SEQ FROM RDB$DATABASE;

DROP SEQUENCE EMP_SEQ;

----------------------------------------------------------------------
-- INDEX 作成

-- CREATE TABLE EMP(
--   EMP_NO SMALLINT,
--   LAST_NAME VARCHAR(20),
--   EMAIL VARCHAR(30),
--   DEPT_NO CHAR(3)
-- );
CREATE INDEX ix_emp_no ON EMP (EMP_NO);
-- show index ix_emp_no;

DROP INDEX ix_emp_no;

----------------------------------------------------------------------
-- SYNONYM はおそらくない

参考：『Oracle Master Bronze 11g SQL 基礎 I 必修教本』