Query Optimization by Analyzing and Pre-caching Frequent Queries in Compile-Time

** Course project for Advanced Compilers (EECS583) **

The src/ folder contains the codes for out library (under src/queryManager/) and benchmarks (under src/benchmark/). All parts are written in java. We have created scripts for making easy to use with the tool:

• To simply compile and run a application use ./javac.sh appName <parameters to run>. It will use the installed javac command in your system.
• To optimize your queries with our tool use our "SQL Pre-Fetcher Compiler" command: ./sqlpfc.sh appName <parameters to run>. This script will profile and analyze the queries and it will inject our sql pre-fetcher into your code. The output would be the class files of the .class files with injected pre-fetcher. You can optionally give a last parameter run to the command, and it will run you app after optimization.
• To run the experiments you can simply use the run_github_expr.sh and run_yelp_expr.sh script. They will take care of everything and will save the results in the expr-result folder. Before being able to run the experiments you'll need to set up the datasets. The instructions to do so are available in README under the folder of each application in src/benchmark/.