There are several different ways to count the number of LinesOfCode in StructuredQueryLanguage.
Here is the SQL example from the ManyToManyChallenge. [DeleteMe: Some spaces have been deleted for consistency; the number of lines has not been changed.]
CREATE TABLE Users
( user_id INTEGER NOT NULL PRIMARY KEY,
username VARCHAR NOT NULL,
CONSTRAINT UNIQUE (username));
CREATE TABLE Groups
( group_id INTEGER NOT NULL PRIMARY KEY,
group_name VARCHAR,
CONSTRAINT UNIQUE (group_name));
CREATE TABLE r_user_groups (
user_id INTEGER NOT NULL,
group_id INTEGER NOT NULL,
CONSTRAINT PRIMARY KEY (user_id, group_id),
CONSTRAINT FOREIGN KEY (user_id) REFERENCES users ON DELETE CASCADE ON UPDATE CASCADE,
CONSTRAINT FOREIGN KEY (group_id) REFERENCES groups ON DELETE CASCADE ON UPDATE CASCADE
)
CREATE INDEX r_user_groups_idx1 ON r_user_groups (group_id);
If a complete SQL query is considered to be a line of code, this example has 4 lines of code.
If the code is formatted as shown above, and each non-blank line as formatted is considered to be a line of code, this example has 16 lines of code.
If each constraint or action is formatted on its own line, the code looks like this:
CREATE TABLE Users
( user_id INTEGER
NOT NULL
PRIMARY KEY,
username VARCHAR
NOT NULL,
CONSTRAINT UNIQUE (username));
CREATE TABLE Groups
( group_id INTEGER
NOT NULL
PRIMARY KEY,
group_name VARCHAR,
CONSTRAINT UNIQUE (group_name));
CREATE TABLE r_user_groups (
user_id INTEGER
NOT NULL,
group_id INTEGER
NOT NULL,
CONSTRAINT PRIMARY KEY (user_id, group_id),
CONSTRAINT FOREIGN KEY (user_id)
REFERENCES users
ON DELETE CASCADE
ON UPDATE CASCADE,
CONSTRAINT FOREIGN KEY (group_id)
REFERENCES groups
ON DELETE CASCADE
ON UPDATE CASCADE
)
CREATE INDEX r_user_groups_idx1
ON r_user_groups (group_id);
If a complete SQL query is considered to be a line of code, this example still has only 4 lines of code.
If the code is formatted as shown above, and each non-blank line as formatted is considered to be a line of code, this example has 30 lines of code.
I prefer to have much of the schema setups in meta-data (See DataDictionary) and use a GUI tool and/or table browser rather than code for such. In fact, I think that is how it is usually stored. Code is just one representation. See CodeAvoidance.
See also: ManyToManyChallenge, SemiColon, LinesOfCode