Slate transcript editor

Work in progress

Building on the success and lessons learned from @bbc/react-transcript-editor. Mostly to be used in the context of autoEdit 3(digital paper edit), and other proejcts.

Criterias/Principles

Easy to reason around
Can handle transcript and media over 1 hour without loss in performance
Only essential features for correction of timed text
adapters to and from other STT services, are external, except for dpe (digital paper edit, adapter).
leverages existing libraries, such as bootstrap, and react-bootstrap, to focus on the diffuclt problems, and not wasting time re-inventing the wheel or fiddling around with css.

See project board for more details of ongoing work.

See draftJs vs slateJs in doc/notes for some considerations that inspired this version.

Storybook

Setup

git clone git@github.com:pietrop/slate-transcript-editor.git

cd slate-transcript-editor

npm install

Usage

Usage - dev

npm run storybook

npm start

Visit http://localhost:6006/

Usage - prod

slate-transcript-editor

npm install slate-transcript-editor

import  SlateTranscriptEditor  from 'slate-transcript-editor';
// you need to import bootstrap separatly
import 'bootstrap-css-only';

<SlateTranscriptEditor
  mediaUrl={DEMO_MEDIA_URL_KATE}
  transcriptData={DEMO_TRANSCRIPT_KATE}
  handleSaveEditor=// optional - function to handle when user clicks save btn in the UI
  />

or with more options, see table below

See storybook *.stories.js in src/components/ for more examples

Attributes	Description	required	type
transcriptData	Transcript json	yes	Json
mediaUrl	string url to media file - audio or video	yes	String
`handleSaveEditor`	function to handle when user clicks save btn in the UI	no	Function
`handleAutoSaveChanges`	returns content of transcription after there's a change, auto save has condierable performance lag on longer files, sudgest to not use for files over 45 min/1hour.	no	Function
`autoSaveContentType`	specify the file format for data returned by `handleAutoSaveChanges` and `handleSaveEditor`,falls default to `digitalpaperedit`, and runs alignement before export. Other option is `slate`,without alignement.	no	String
`isEditable`	set to true if you want to be able to edit the text	no	Boolean
`showTimecodes`	set to true if you want to show timecodes in the transcript at paragraph level	no	Boolean
`showSpeakers`	set to true if you want to show speaker labels in the transcript at paragraph level	no	Boolean
`title`	defaults to empty String, also used in file names for exported files.	no	String
`showTitle`	Whether to display the provided title	no	String
`mediaType`	can be `audio` or `video`, if not provided, it defaults to video, but the component also uses the url file type to determine and adjust the player (`.wav`, `.mp3`,`.m4a`,`.flac`,`.aiff` are recognised as audio )	no	String

see storybook for example code

System Architecture

Uses slate as editor, see their docs.
Uses align-diarized-text for restoring timecodes. This lib combined stt-align-node with alignment-from-stt to restore timecodes and preserve speaker labels.
It align-diarized-text when export of formats that require timecodes, eg dpe json, or docx and txt with timecodes. Also for the 'realignement'/sync UI btn.
If you export or save as slate json, at the moment it doesn't run alignement. The function to perform the alignement is also exported by the module, so that you can performe this computational intensive alignement elsewhere if needed, eg server side.

CSS

The project uses bootstrap, and react-bootstrap. And you'll need to include your own stylesheet in your React app.

npm install bootstrap-css-only

eg bootstrap-css-only is convinient because it doesn't ship with JQuery, that is not a dependency of react-bootstrap

and then import in your app

import 'bootstrap-css-only';

Alternativly this gives you the extra flexibility to write your own overriding the boostrap classes (see bootstrap and react-bootstrap on themeing) for more info.

Documentation

There's a docs folder in this repository.

docs/notes contains dev draft notes on various aspects of the project. This would generally be converted either into ADRs or guides when ready.
docs/guides contains walk through / how to.
docs/adr contains Architecture Decision Record.

The docs folder syncs with gitbook to make the documentation more pleasent to browse at autoedit.gitbook.io/slate-transcript-editor-docs/ - Work in progress

Development env

npm 6.13.6
node 12
storybook

If you have nvm you can run nvm use to change to the node version for this repo.

Linting

This repo uses prettier for linting. If you are using visual code you can add the Prettier - Code formatter extension, and configure visual code to do things like format on save.

You can also run the linting via npm scripts

npm run lint

and there's also a pre-commit hook that runs it too.

Build

build module

Following storybook Distribute UI across an organization guide.

build storybook

npm run build-storybook

Tests

TBC

Deployment

Deployment module

To publish module to npm

npm run publish:public

and for a test run use

npm run publish:dry:run

Deployment storybook

To publish storybook to github pages

npm run deploy:ghpages