Quick links#🠑
Some sources of inspiration:
Overview#🠑
The de-facto way of looking at code is as a dry, machine-like, monospaced, almost monotonous form of giving instructions to the computer. But I think code is just as creative, colorful, vivid, and textual as all of the other forms of information we consume. Our unconscious perception of a divide between computer programs and other forms of text prevents us from critiquing the actual language of code, down to its individual variables and functions.
This (small) project tries to push back on that by providing a method to make folders of code into books — or really, just printable PDFs. It features syntax highlighting with highlight.js, automagical formatting with prettier, and a table of contents with page numbers from my own hacky coding.
The actual PDF generation is done with html-pdf with some processing done in pdf-parse.
View an example PDF of what this tool can make here.
Usage (CLI)#🠑
Install with npm install --global codexer
.
If you’re new to Node, feel free to check out this guide.
codexer [options]
Options:
-V, --version output the version number
-t, --target <path> The path to a directory to be made into a PDF. The flag can be omitted.
-a, --author <name> will be filled in on the title page and every header
-o, --outPath <path> output location. Tilde notation currently not accepted :(
-d, --dry Add -d to enable a dry run that produces only a JSON
representing the order in which files will be assembled.
-j, --json <path> Pass in the location of a JSON file to specify your own order of files instead.
Most useful after trying out the -d option to see what a config should look like.
-dh --dryHTML Like the dry run option, but produces just HTML output. No page numbers though :(
-s --stylePath <path> Path to a HTML file with configurations that will be used to style the PDF.
-q --quietly Suppress all debugging messages
-w --width <length> Page width
-h --height <length> Page height
-e --exclude <patterns...> Specify regex patterns for files to exclude. Default excludes node_modules,
.git, yarn.lock and package-lock.json, and .env files.
Codexer also excludes any non-text encoded files; this cannot be altered.
--help display help for command
Usage (as a Node module)#🠑
You should probably stick to using it as a CLI tool or for
just messing around, as it’s not really ready for
production. But if you want, you can install with
npm install codexer
in your Node project. Then use it as
follows:
const codexer = require('codexer')
// or
import codexer from 'codexer'
codexer('.')
// Finished! PDF is located at /tmp/codexer/[your directory basename].pdf
codexer('.', {outPath: 'output.pdf'})
// Finished! PDF is located at [path to your directory].pdf
All of the CLI options are available for use in the npm version.
npx?#🠑
You can also call it as a single-use tool with npx if you
have Node installed by prefixing npx to the codexer
commands, e.g. by running npx codexer .
.
The problem is that phantomJS and pdf-parse, which this tool depends on, are incredibly large (the local node_modules/ folder on my machine comes out to 184Mb). Download, load, and install times are quite long because of this.
If you intend on using this tool more than once, I’d recommend just installing it globally.
If this doesn’t fit your use case#🠑
All in all, it probably doesn’t! Please, please feel free to download and alter it if you want an adjustment.
Personal notes#🠑
Honestly, this is just one of those cool-but-not-that-cool side projects I’ve feel like I’ve distracted myself with these days. But it is cool!
I’m looking for more tangible ways to hold and digest code, and more than anything else critically analyze code itself (e.g. rather than its periphery — the funders, the things it is used for, etc.). If you know of any links on this or have creative ideas, please send them my way!!