DBMS

A simple relational DB built during the Bradfield Sabbatical program.

Objective

To learn how a relational DBMS system works from first principles

Overview

• /data directory will contain the csv and heap files.
• heap_file.c, given a schema and a CSV file, it outputs their corresponding heapfile
• executor.c contains all implemented nodes:
  • scan: it scans tuples from heap files one at a time.
  • select: given a predicate functions, it tests a tuple against it if the test returns true.
  • project: it filters out the fields not present in the result schema from a tuple and yields a new tuple that conforms to the result schema
  • average
• serializer.c, used to serialize char, int into binary and vice versa.

Run

You will first need to download the movie lens data into the /data directory. You'll only need the movies.csv and ratings.csv.

Once that's done you need to run the following command to convert the csv files to a heap file that the DB can use.

  gcc -o heap_file serializer.c heap_file.c -I.
  ./heap_file

After that's done you can run the query executor

 gcc -o executor serializer.c executor.c heap_file.c -I.
 ./executor