features_and_other_notes.txt

testClass.py:	
-------------
Driver file for getFeatures.py which extracts the features from the given delimiter separated file containing all the input scripts/


train.py:
------------
To train the machine learning classifier using data from data/train2.py and to write out the trained object to disk


classifier_test.py:
-------------
Takes a test .csv file as input and gets the classifier's prediction for the data in that file.


getFeatures.py:
-------------
Constructs Javascript AST for the input script file. Extracts features from the AST and returns to thecalling driver file in the form of a dictionary
Features extracted are:

1. No of eval() functions 
2. No of setInterval() functions
3. No of setTimeout() functions
4. No of Javascript string link() functions
5. No of Javascript string search() functions
6. No of exec() functions
7. No of escape() functions
8. No of unescape() functions
9. Ratio of no of Javascript keywords to no of normal words (from predefined list)
10. Average entropy of all words
11. Average entropy of the script
12. No of long strings in the program (with length >40)
13. Maximum entropy of the program
14. Average string length in the program
15. Maximum string length in the program
16. No of functions or variables with long names(with length>=40)
17. No of direct string assignments
18. No of string-modiyfing functions
19. No of event attachment functions
20. No of DOM modifying functions in the script
21. No of suspicious strings (lookup from predefined list)
22. Whitespace ratio (No of whitespace characters to actualy number of characters in the script)
23. No of strings containing only hexadecimal characters
24. Max non printable characters in strings	
25. Average line length in the script
26. No of times 'iframe' is a string in the script
27. No of tags with malicious names in the script (from predefined list)
28. Length of the script