Code Reviewer

STEPS:

1.Create a virtual environment $ py -m venv env and activate it by $ .\env\Scripts\activate
2.In the environment install the following libraries(flask, scikit-learn, lizard, numpy, pandas).
3.Change in app.py to the csv file in which output should be taken.
4.Run$ python app.py

---

Requirements for running the code- all can be installed using pip

1. flask
   
2. sklearn
   
3. lizard
   
4. numpy
   
5. pandas
   

---

Code is divided into 3 parts-

• app.py: Flask app
  
• code_review.py: Genrating Halstead Metrics and Cyclomatic Complexity
  
• predict_code.py: For giving output as good or bad using trained model
  

---

Explanation:

• app.py renders the HTML and CSS UI and handles the routing og templates.
  
• app.py calls main() function in code_review.py which returns the features of user code.
  
• code_review.py uses codeparams functions to calculate halstead metrics and Mccabe Cyclomatic complexity.
  
• After calculating code parameters app.py call arr_input() from predictcode.py.
  
• arr_input() predicts the output based on the code metrices.
  
• In arr_input() function saved trained model is stored in form .sav and we use it for prediction.