A simple relational database based on Stanford CS346 RedBase, implemented in elegant modern C++14.

• Indexing. We implement B+Tree index to accelerate single-table queries as well as table joins.
• Data Persistence. All changes to data are persistent on disk if the shell exits normally.
• SQL support. We support a subset of SQL, including basic DDL (create table, drop table, create index, drop index) and DML (insert, delete, update, select) statements.

Prepare a Linux / macOS machine, and clone this project to your local environment.

git clone https://github.com/li-plus/redbase-cpp && cd redbase-cpp

Install dependencies of this project. We use CMake as the C++ build system. Flex & Bison are used to generate the SQL parser. libreadline helps us to build a user-friendly shell.

sudo apt install cmake flex bison libreadline-dev

Then build the project and make optional unittest.

mkdir -p build/ && cd build
cmake ..
make -j
make test   # optional

Now start the RedBase shell and have fun!

./bin/redbase mydb

This command creates a database named mydb for the first time, since it does not exist yet. On the next time, it will directly load the existing database mydb. To drop the database mydb, simply remove this folder by rm -r mydb.

Below is a quick demo of the supported main features.

create table student (id int, name char(32), major char(32));
create index student (id);
create table grade (course char(32), student_id int, score float);
create index grade (student_id);

show tables;
desc student;

insert into student values (1, 'Tom', 'Computer Science');
insert into student values (2, 'Jerry', 'Computer Science');
insert into student values (3, 'Jack', 'Electrical Engineering');

select * from student;
update student set major = 'Electrical Engineering' where id = 2;
select * from student;
delete from student where name = 'Jack';
select * from student;

insert into grade values ('Data Structure', 1, 90.0);
insert into grade values ('Data Structure', 2, 95.0);
insert into grade values ('Calculus', 2, 82.0);
insert into grade values ('Calculus', 1, 88.5);

select * from student, grade;
select id, name, major, course, score from student, grade where student.id = grade.student_id;

drop index student (id);
desc student;

drop table student;
drop table grade;
show tables;

exit;

• Paged File Module (PF) enables higher-level modules to perform efficient file I/O in terms of pages.
• Record Management Module (RM) manages the storage of unordered records.
• Indexing Module (IX) manages persistent indexes over unordered data records stored in record files.
• System Management Module (SM) handles the data definition language (DDL), including create table, drop table, create index, drop index statements.
• Query Language Module (QL) handles the data manipulation language (DML), including insert, delete, update, select statements.
• SQL Parser translates a raw SQL statement into an Abstract Syntax Tree (AST), which is further interpreted and executed by the controller.