SANER 2024: Compare performance between in terms of exection speed between procedural programming and imperative programming

This research project is a comparative study between for loops and list comprehension in terms of performance.

The python files

rules.py

This file contains the database of rules used to convert for into list comprehension and list comprehension into for loops. The files contains the following functions:

• for_rewrite_rules: represent the rules for converting for loops into their equivalent list comprehension.
• list_comp_rewrite_rules: represent the rules for converting list comprehensions into their equivalent for loops.

model.py

The model.py file contains the differents models used to create the symbol table and retrace the value of a variable. It has:

• Class Symbol: it makes that for every occurence of any string we get each the same symbol.
• Class ScopeType: Enum used to represent different type of scope that we can have across the program.
• Class VariableType: Enum used to represent different type of variable that we can have across the program.
• Class Scope: a scope is represented by: - a set of instructions (symbol, line, value) - a set of variable declared in the scope - its type - list of inner scope - link to the parent scope - iter variable if it is an iterable scope (i.e for-loop, while-loop)

symbolTable.py

This file contains:

• The class SymbolTable: The symbol table contains for each symbol the set of all scope where each symbol will be potentially used.

source.py

This file contains the code used to retrace the origin of a variable i.e In a symbol table when you want to retrace the value that contains a symbol. This file has the:

• Class Source that given a specific symbol at a specific line retrace its value if possible.

unit.py

This file contains:

• Class Unit: This class allow the insertion of new code snippet inside an AST tree
• The function extend_ast_tree: This function extend an AST tree by adding a new attribute "parent" to each node

testCases.py

This file contains all the test cases used to verify the good behavior of the codes written. It has:

• ForTestCases
• SymTableTestCases
• ListCompTestCases
• SourceTestCases

process.py

This file contains different function to help extract useful informations on AST tree node. For now the code can extract informattion about ast.For, ast.If and ast.Expr nodes.

performance.py

Testing the timing performance between the for loop transformation and list comp transformation

helper.py

Just a file to help avoid duplication. The file contains function to interact with files, folder and csv.

stats.py

The most important file of the project. It takes all the python files of a given project and produce a json containing the statistics of the project about the number of for transformed.

Using the code

The use this project you must clone this project on your local computer and in your own file you will do the following:

in the source folder, open the terminal and run the following command:  python3 stats.py --p path_to_your_project