MultiHTMLParse

This project was made to test multiple server-side HTML parsers, to compare their functionality to each other, and web browsers.

Running

docker-compose down && docker-compose build --parallel; docker-compose up;

Then visit http://127.0.0.1/

Input HTML and click parse. The result will replace __PARSER__ with whatever parser was used, and (if checked) render the result in an iframe.

For example, <svg onload=console.log('__PARSER__')>

Running CLI

docker-compose down && docker-compose build --parallel; docker-compose up;

echo "html=<h1>lol</h1>"|docker exec -i parse_dotnet "/usr/lib/cgi-bin/parsehtml"

Demo

http://htmlparse.hackaplaneten.se/

FAQ

• Why CGI? WTF?
  • I didn't feel like implementing web server/framework for every language.
• Why not newer docker-compose version?
  • Because then it didnt work with podman
• Can you add XYZ parser/language?
  • Maybe, but I wont be actively maintaining this project. Feel free to send a PR, or make a fork.
• Why GPL MIT license?
  • Even though no future customer deserves to pay for this crappy code, I don't want any contributor to withhold code for the sake of avoiding "locking it into" a GPL-like license. So MIT it is.

Currently supported parsers

• dotnet/AngleSharp
• dotnet/HtmlAgilityPack
• elixir/floki
• elixir/floki+fasthtml
• elixir/floki+html5ever
• golang/goquery
• golang/htmlquery
• java/jsoup
• node/node-html-parser
• node/cheeriojsdom
• node/parse5
• node/jsdom
• php/DOMDocument
• php/SimpleXML
• python/html.parser
• python/html5lib
• python/lxml
• ruby/nokogiri
• ruby/ox
• rust/scraper

TODO

• Use smaller base docker images
  • Use a builder to make rebuilds faster
• More parsers
  • Options for specific version of parser
• Make parsers return interesting standard data like tag count, first <a> etc
• Make frontend better at comparing results
  • Make an actual frontend