H2 database column name “GROUP” is a reserved word

Question

How do I create a table in H2 with a column named GROUP? I saw an example that used something like [*] a while ago, but I can\'t seem to find it.

Answer 1

You have to surround the reserved word column name in quotes, like so

"GROUP"

Source (direct link): h2database.com

Keywords / Reserved Words

There is a list of keywords that can't be used as identifiers (table names, column names and so on), unless they are quoted (surrounded with double quotes). The list is currently:

CROSS, CURRENT_DATE, CURRENT_TIME, CURRENT_TIMESTAMP, DISTINCT, EXCEPT, EXISTS, FALSE, FOR, FROM, FULL, GROUP, HAVING, INNER, INTERSECT, IS, JOIN, LIKE, LIMIT, MINUS, NATURAL, NOT, NULL, ON, ORDER, PRIMARY, ROWNUM, SELECT, SYSDATE, SYSTIME, SYSTIMESTAMP, TODAY, TRUE, UNION, UNIQUE, WHERE

Certain words of this list are keywords because they are functions that can be used without '()' for compatibility, for example CURRENT_TIMESTAMP.

Answer 2

I've been having this problem with SQL generated by JPA... Turned out I was using a variable name called limit.

Caused by: org.h2.jdbc.JdbcSQLSyntaxErrorException: Syntax error in SQL statement "CREATE TABLE EXPENSE_LIMIT (ID BIGINT NOT NULL, LIMIT[*] DECIMAL(19,2), ACCOUNT_ID BIGINT, EXPENSE_CATEGORY_ID BIGINT, PERIOD_ID BIGINT, PRIMARY KEY (ID)) "; expected "identifier"; SQL statement:

Where my model class had a field called limit.

The fix is to specify column name as

@Column(name = "`limit`")

Answer 3

Trailing Underscore

Add a trailing underscore: GROUP_

The SQL spec explicitly promises^† that no keyword will ever have a trailing underscore. So you are guaranteed that any naming you create with a leading or trailing underscore will never collide with a keyword or reserved word.

I name all my columns, constraints, etc. in the database with a trailing underscore. Seems a bit weird at first, but you get used to seeing it. Turns out to have a nice side-effect: In all the programming as well as notes and emails, when I see the trailing underscore I know the context is the database as opposed to a programming variable or a business term.

Another benefit is peace-of-mind. Such a relief to eliminate an entire class of possible bugs and weird problems due to keyword collision. If you are thinking, "No big deal - what's a few SQL keywords to memorize and avoid", think again. There are a zillion keywords and reserved words, a zillion being over a thousand.

---

The answer by Shiva is correct as well: Adding quotes around the name, "GROUP", does solve the problem. The downside is that remembering to add those quotes will be tiresome and troublesome.

---

Further tip: For maximum compatibility across various SQL databases, do your naming in all lowercase. The SQL spec says that all names should be stored in uppercase while tolerating lowercase. But unfortunately some (most?) databases fail to follow the spec in that regard. After hours of study of various databases, I concluded that all-lowercase gives you maximum portability.

So I actually suggest you name your column: group_

Multiple word names look like this: given_name_ and date_of_first_contact_

---

^† I cannot quote the SQL spec because it is copyright protected, unfortunately. In the SQL:2011 spec, read section 5.4 Names and identifiers under the heading Syntax Rules item 3, NOTE 111. In SQL-92 see section 5.2, item 11. Just searching for the word underscore will work.