codexplor

🧰🔧🔨 codexplor is a WIP 🧰🔧🔨

codexplor offers R functions dedicated to explore, analyze and monitor a programming project.

Given a programming project, codexplor compute standardized text mining metrics and dataviz’ about the project.

• Get rid of global complexity with networks of internal dependencies,
• Assess local complexity with
  • document-level (e.g., identify files with many functions defined in)
  • and function-level metrics (e.g., longest functions, those with numerous parameters or internal dependencies).

codexplor help me to figure out the big picture of a programming project faster, and to manage it more efficiently.

Installation

You can install the development version of codexplor with

devtools::install_github("clement-LVD/codexplor")

The default settings of codexplor are optimized for analyzing a project in language. Supported languages are : R, Python

---

Example

1. Turn a programming project into a corpus. Given folder(s) and/or github repo(s) and programming language(s), codexplor::construct_corpus will return a list of dataframes : the programming project is turned into a text-mining corpus.

library(codexplor)
# Construct a corpus with local foldercorpus<- construct_corpus(folders= getwd(), languages="R" )

This corpus of dataframes is a standardized way to analyze a programming project as a collection of documents and get insights on the functions and the files of the project, see the vignette of construct_corpus().

---

2. See a dataviz’ from a corpus.list. Given a corpus.list, look at the dataviz’ of an internal.dependencies network with codexplor::get_networkd3_from_igraph :

# Produce an interactive dataviz' with the network of internal.dependenciesdataviz<- get_networkd3_from_igraph(corpus$functions.network
, title_h1="codexplor. Graph of internal dependancies : functions network"
, subtitle_h2="Color and links = indegrees") # herafter an image (non-interactive) of the interactive dataviz ↓

These dataviz are useful for pinpointing where to start a polishing loop, identifying all the functions impacted by upcoming changes, […] or assessing the impact of a new dev loop on the project’s complexity.

Or look for a dataviz of the files with the following :

 get_networkd3_from_igraph(corpus$files.network)

See an example of a files network.

---

Vignettes

• See the vignette of construct_corpus().

• See the vignette of the citations.network of internal.dependencies dataframes.

• codexplor also offers helper functions, e.g., for create and filter a network with the igraph package, see the vignette of helper functions for igraph object and networkD3 dataviz

Features

Given programming project(s), codexplor::construct_corpus will compute several standardized metrics and answer a corpus.list of dataframes :

• The functionsdata.frame give insights about each functions of the project(s), e.g., number of parameters, number of internal dependencies, length of the code.

• The filesdata.frame give insights about each files, e.g., quantify number of functions within files and assess critical internal dependencies.

• The files.network and functions.network are networks of internal dependencies within the project(s).

Computed Methods	Level of insights
	Appreciate global complexity and reveal critical files, e.g., major internal dependancies, clusters of ‘difficult-to-read’ files
	Assess each file, e.g., length and files readability, number of functions within a file
	Assess each function, e.g., number of parameters or number of internal dependencies
	Assess each line, e.g., find the longest

Availables features :

Planned features :

Supported language(s) :

Other languages are planned.