Motivation

The purpose of the lib to provide an easier way for understanding dependency graph in our apps. Usually, we want to see next dependencies:

1. Why that specific module or file was included in the bundle.
2. Which dependencies in the bundle are included due to the specific module.
3. Which modules use that specific module.

Usage

Tool uses stat.json file, provided by webpack, to build dependency graph. So to get a visual representation of dependency graph:

1. Run webpack build with saving stats file on disk. ie, webpack --json > stats.json.
2. Go to online version.
3. Upload generated stats.json file.

Next version will have the option to be used as webpack plugin as well.

Usage cases

Let's overview all 3 main cases:

Module Overview

In that mode, we can see all third party modules that are used in our code base and connection between them. In this way we can spot hidden dependencies or find duplicated modules.

Example:

Here we can see that stacktrace.js is used only in one place in our code and included some heavy third party modules. So we can conclude that it's a good candidate for async loading

Module usage

If we want to see all modules that require specific module, we can select appropriate module in select element that places at top of that page

Example:

We want to check how much it costs to switch to preact library. Selecting react-dom we can see next picture

Well, seems that we need to look more carefully at those libraries that use React-DOM API under the hood and check how they are compatible with preact

File usage

Another case is when we need to understand the whole picture why the file was included in the bundle. The specific files are listed in the table on the left side of the page

To see the whole inclusion tree (up to webpack entry point) you should select specific file with a click on its name. Than tree will be rendered:

1. inclusion tree

2. reference table with listing all places where that specific file was required and listing all modules that are required by our specific module

reference table

Example

We want to split our domain logic by pages, before we need to understand on which pages those files are utilized. Having selected desired module in the table at the right, we are able to see the whole inclusion tree.

To reduce visual noise use filtering by name(input element at top left corner of the page)

well, now we figured out that desired domain logic file is utilized only on one page and we can start thinking about code-splitting and inclusion that domain logic file into appropriate chunk

Additional notes

The lib is still under development and was created due to personal needs. If you have any ideas what should be improved or fixed don't hesitate to create an issue.