A full-text transcription web app, written for the Tate Britain.
To install, clone the repo and run npm install
To develop locally, run npm start. Your browser will automatically be opened and directed to the browser-sync proxy address, and the server will serve files from the /build directory. Any changes in the /app directory will be automatically processed by Gulp, and the changes injected to any open browsers pointed at the proxy address.
NOTE: this app wasn't developed against staging, so while there is a staging app for testing auth, it probably won't work otherwise. To get it running locally as expected, run NODE_ENV=production npm start;
To upload new subjects and sets, use annotate-uploader, which will add all the required metadata needed to group by artist. You'll also need to add any new artists to app/modules/set-selector/artists.constants.js.
The project now uses Caesar to request aggregated lines via the GraphQL endpoint. To show an aggregated line, the minimum consensus score on a line must be > 2.5, or the number of views on the line must be > 4 . The minimum number of views is set as the consensus score can be artificially low if a user decides to type a bunch of words that aren't there/try to do the full page of text but only has dots round the first line.
Deploys happen automatically on the master and staging branches via Jenkins.
To deploy from your local machine, you'll need to set the BUCKET, PREFIX and BASE_URL environment variables before running the build and deploy scripts.
export BUCKET="zooniverse-static"
export PREFIX="anno.tate.org.uk/"
export BASE_URL="https://anno.tate.org.uk/"
npm run build && npm run deploy- AngularJS
- Gulp
- Browserify
- Stylus
- Nib
- Bootstrap
- Alpha version - as of 5c954ca
- Data aggregation algorithm by ggdhines