Raw File
Tip revision: 57d704e8ca05108db3526b7181d70fa33add3519 authored by Cameron Yick on 13 December 2023, 03:03:53 UTC
fix(docs): Update typos in public copy (#10)
Tip revision: 57d704e
# No Grammar Supplementary Materials

This repo contains of the supplementary materials for the paper ["No Grammar To Rule Them All: A Survey of JSON-Style DSLs for Visualization"]( by Andrew McNutt.
Specifically it consists of

- a repository of example programs for each of the surveyed languages (code-examples/). (Nb: this contains a number of languages which were at some point included in our survey, such as tracery, which were eventually ruled to be excluded by our survey criterion.)
- an interactive website for exploring those samples and our associated analysis as well as several charts (src/)
- a notebook for constructing the "family tree" figure from the paper (analysis-nbs/analysis.ipynb)
- a list of relationships between languages in the survey (analysis-nbs/relationships.tsv)
- a table showing our analysis for each of the languages (public/lang-meta.tsv)

The interactive website is made available at

## Local instructions

While it is much easier to simply view these materials at the link provided above, if you wish to run them locally you can do so by running the following commands in the terminal

npx ts-node scripts/build-bundle.js
yarn dev

And then pointing a browser to http://localhost:5000/ (or whatever port is described as being opened in the CLI messages). This requires that node/yarn be already installed. See the [node website]( and [yarn website]( respectively for installation instructions.

If you wish to run the analysis notebook, run the following terminal commands

mkdir venv
python3 -m venv venv
source venv/bin/activate
pip3 install -r py-requires.txt

And then open a notebook as normal by running `jupyter notebook` in the terminal. Finally navigate to the analysis notebook (/notebooks/analysis-nbs/analysis.ipynb) through jupyter's browser. In development of that notebook the vscode version of notebooks was used, however it should be appropriately supported using any Jupyter notebook container. (The slightly unusual name for the requirements file is to prevent netlify from installing python requirements).

Finally, we have provided a script to reproduce one of the figures in the paper (namely Figure 3, which shows the analysis of each language across a variety of categories). This can be run via

sh scripts/

This script installs the dependencies, spins up a dev server, and then opens a puppeteer instance to save the image created on the web page. If you like, you can skip these steps and view this figure directly at  

back to top