Feature Extractor

V 1.2.4
This app extracts features from source code files.
This project is part of the Guardjs organization.

This application has a command line interface thanks to "yargs".
Simply extacts features from an string and save them as a "csv" file.
It is for getting these specific features from js files, but thanks to beautiful architecture of this app and it's designer (me), it has the capability to easily reduce or increase a feature model to the feature set.

Usage

• install requisites by npm i
• use $ node app --help command to get helped with this app
• app works with serve command like $ node app serve >>resultfile

Commands

1. $ node app feature >>result/1.csv to show just a the feature set
2. $ node app serve >>result/1.csv it should work totally

alternatives for the 2th command

• $ node app serve --directory="$workspace\samples" >>result/1.csv
• $ node app serve -d="$workspace\samples" >>result/1.csv

Feature set

Attention: the order of features are by the order of files of feature-model that the app uses and the order of features here isn't a default order! THERE'S NOT ANY ORDER

no.	feature	description
1	eval	The number of eval
2	setTimeout	The number of setTimeout
3	iframe	The number of strings containing “iframe”
4	unescape	The number of unescape
5	escape	The number of escape
6	classid	The number of classid
7	parseInt	The number of parseInt
8	fromCharCode	The number of fromCharCode
9	ActiveXObject	The number of ActiveXObject
10	strDirect	The number of directly string assignments
11	concat	The number of concat
12	indexOf	The number of indexOf
13	substring	The number of substring
14	replace	The number of replace
15	document.addEventListener	The number of document.addEventListener
16	attachEvent	The number of attachEvent
17	createElement	The number of createElement
18	getElementById	The number of getElementById
19	document.write	The number of document.write
20	words	The number of words in JavaScript
21	keywords	The number of JavaScript keywords
22	chars	The number of characters in JavaScript
23	keyOnWords	The ratio between keywords and words
24	wordperline	The entropy of the script as a whole
25	longest	The length of the longest JavaScript word
26	longer200	The number of long strings(> 200) characters
27	shortstring	The length of the shortest JavaScript word
28	wordperlongline	The entropy of the longest JavaScript word
29	space	The number of blank spaces in the JavaScript
30	avg	Average length of words in the JavaScript
31	hexvalue	The number of hex values in the javascript
32	utfvalue	The number of utf values in the JavaScript
33	vertexes	The number of vertexes
34	Largest	label The length of longest label of a vertex
35	edges	The number of edge
36	Largest	degree The most degree of a vertex
37	E/V	Ratio number of edges on vertexes
38	Loop	The number of loops in graph
39	longestLoop	The length of longest loop
40	condition	The number of condition branch
41	exEdge	The number of edges of exceptions
42	exEdge/edges	Ratio number of exception edges on all edges
43	strVertex	The number of vertexes with string assignment

Anatomy of project

app folder contains feature-models. middleware folder is for some initializing and some functionality provided in the way to do the main job. app.js file is the main process of the algorithem! I tried to make it simple and readable :)).

.
+-- - app  
|   +-- - syntax  
    |   +-- index.js # this will handle all suntax features  
    |   +-- + rgxPatternCounter # the models inside this folder are simple regex functions   
    |   +-- + common  
|   +-- - graph  
    |   +-- + common  
|   +-- index.js  
+-- - interface # commands going here   
|   +-- serve.js  
|   +-- list.js  
|   +-- count.js  
|   +-- index.js  
+-- app.js

farther question

Don't hesitate to contact me about this project
me on github
email me
me on telegram (fast way)