rxw1/slices

A browser for sliced Haskell fragments

A browser for sliced Haskell fragments

Showcase project and playground utilizing Redux, Redux DevTools, React, React Router, RethinkDB, ElasticSearch, universal/server side rendering using Koa and Material Design Lite. It's written in ES7, compiled by Babel and the front end is served via WebPack. Originally based on khtdr's redux-react-koa-isomorphic-counter-example and the official Redux examples.

Questions, feedback and forks are welcome. I'm happy about any issues opened.

There are still some bugs to fix and some more tech I'd like to use or at least explore, like ImmutableJS, Rx.js, better ElasticSearch indexing, RethinkDB Changefeeds, WebSockets, authentication, caching, better and more complex routing and more. And the output files are still pretty big: 5.6 MB for common.js (third party library stuff) and 1.5 MB for app.js. Anyways, the code is quite clean and up to date. That's a good thing.

There's a slight chance that it's currently running here, so you can play around a little bit.

Requirements

• RethinkDB 2.1.2
• ElasticSearch 1.7.1

Setup

Make sure RethinkDB and ElasticSearch are correctly set up and running.

npm install
npm run dev
curl localhost:3000/setup/db
curl localhost:3000/setup/es

Insert slices into the database

You can add slices to the database by sending a POST request to api/slices, e.g.

curl -XPOST localhost:3000/api/slices -H "Content-Type: application/json" -d "$(cat ./slice)"`.

or

find slices -type f | while read -r slice; do
  xpost api/slices -H "Content-Type: application/json" -d "$(cat $slice)"
done

Copy some stylesheets around (sorry! will fix that)

mkdir public && cp -r node_modules/highlight.js/styles public

If things went well, the client should be reachable at localhost:3000/slices. You can query the API on e.g. localhost:3000/api/slices. See src/server/routes/slices.js for some of the available routes.

METHOD	ENDPOINT	DESCRIPTION
GET	api/slices/sample	get random slices, defaults to 3
GET	api/slices/sample/42
GET	api/slices/1719479834788995000	get a slice and all its dependencies
GET	api/slices/1719479834788995000/refs	get references only
GET	api/slices/search/hexquad	query elasticsearch to return slices with matching fragments
POST	api/slices/1719479834788995000/upvote
POST	api/slices/1719479834788995000/downvote
POST	api/slices/1719479834788995000/like
GET	api/slices/liked	get all liked slices
GET	api/slices/withReferences	get all slices w/ references
GET	api/slices/withoutReferences	get all slices w/o references
GET	api/slices/withInstances	get all liked w/ instances

Zsh love

export PORT=3000
pick () { jq ".[].$1" }
xget () { curl -s -XGET localhost:$PORT/$1 }

Query Elasticsearch and get sliceIDs only

xget api/slices/search/int\?size=50 | jq '.hits.hits[]._source.sliceID'

Pretty reference counts

(repeat 42 (
  a=$(xget api/slices/$(xget api/slices/sample/1 | pick sliceID)/refs | pick sliceID | wc -l)
  echo -n "$a "
  repeat $a printf .
) && echo) | lolcat -F 0.5

Count duplicate slices on each fetch

repeat 42 xget api/slices/sample/1 | pick "uses[].reference.otherSlice" |\
  grep -v null | uniq -d | wc -l

Insert a directory of slices into the database

slices=($(ls -1 slices))
for i in slices/*; do
  xpost api/slices -H "Content-Type: application/json" -d "$(cat $i)"
  echo
done