React>> markov_draftjs>> 返回
项目作者： thibaudcolas

项目描述 ：
Draft.js sample content generated with Markov chains of Project Gutenberg books.
高级语言： JavaScript
项目地址: git://github.com/thibaudcolas/markov_draftjs.git
创建时间： 2017-10-24T13:11:41Z
项目社区：https://github.com/thibaudcolas/markov_draftjs
开源协议：MIT License
下载


markov_draftjs PyPI npm

Draft.js sample content generated with Markov chains of Project Gutenberg books.

This sample content is meant to be used while testing projects based on Draft.js, in particular Draftail and draftjs_exporter.

Why

Sample content can be useful to stress-test and benchmark tools built to handle Draft.js content. For the exporter, this is a great way to reliably assess its performance.

The content from this repository isn’t generated randomly – while the text and metadata values are fake, the content’s structure and the distribution of rich text formatting amongst the text is representative of that of 3 big CMS sites combined.

Here are rich text formats used in the content:

  • Blocks
    • unstyled
    • header-two
    • header-three
    • header-four
    • ordered-list-item, depth: 0 or 1
    • unordered-list-item, depth: 0 or 1
    • atomic
  • Inline styles
    • BOLD
    • ITALIC
  • Entities
    • LINK, MUTABLE with url (URL), linkType (page|external|email), optionally id (number)
    • DOCUMENT, MUTABLE with label (plain text), id (string containing a number)
    • IMAGE, IMMUTABLE with title (plain text), id (string containing a number), src (URL)
    • HORIZONTAL_RULE, IMMUTABLE without data

Using the sample content

In order to simplify using the samples across multiple projects, they are published as packages on npm and PyPI.

  1. # JavaScript projects.
  2. npm install markov_draftjs
  3. # Python projects.
  4. pip install markov_draftjs

Then, in JavaScript:

  1. const contentStates = require("markov_draftjs");

And in Python:

  1. from markov_draftjs import get_content_sample
  3. content_states = get_content_sample()

The sample content is also available from GitHub, eg. with RawGit (warning - big file): https://cdn.rawgit.com/thibaudcolas/markov_draftjs/44827d98/markov_draftjs/content.json.

Development

Requirements: virtualenv, pyenv, twine

  1. git clone git@github.com:thibaudcolas/markov_draftjs.git
  2. cd markov_draftjs/
  4. # Install the git hooks.
  5. ./.githooks/deploy
  7. # Install dependencies
  8. nvm install
  9. npm install
  11. # Unarchive sample text.
  12. cd corpora/
  13. tar -xzvf *.tar.gz
  14. cd ..
  16. # Install the Python environment.
  17. virtualenv .venv
  18. source ./.venv/bin/activate
  19. make init
  21. # Install required Python versions
  22. pyenv install --skip-existing 3.10.0
  23. # Make required Python versions available globally.
  24. pyenv global system 3.10.0
  26. # Generate new sample content.
  27. npm run start

Releases

  • Use irish-pub to confirm the content of the npm package.
  • Make a new branch for the release of the new version.
  • Update the CHANGELOG.
  • Update the version number in markov_draftjs/__init__.py, and package.json, following semver.
  • Make a PR and squash merge it.
  • Back on main with the PR merged, use make publish (confirm, and enter your password) and npm publish.
  • Finally, go to GitHub and create a release and a tag for the new version.
  • Done!

See also