Use elasticsearch + PHP bindings ...

• Migrate content from WP: http://blog.openshakespeare.org/ to openliterature.net
• hmmm: some more recent blog posts not copied across - need to complete …
• Set up word of the day
  • Make it look nice in the sidebar ...
• Migrate across plays (see url structure)
  • We just take the main version for the present (forget all the alternate versions?)
• Sort out front page
  • Main authors with a nice image (shakespeare, milton)
• Get images in assets.okfn.org (and use them)
• Change theme
• Boot assets.openliterature.net - #10
• Move across pdfs to there #11
• Google analytics
• Redirects
• Shut down old system

Play migration

• Shakespeare
• Milton
• Philosophers (??)

New url structure

• URL structure: /{author}/{work}/
  • e.g. /shakespeare/hamlet/
  • alternative editions /{author}/{work}_version/ e.g. /shakespeare/hamlet_folio/

The media uploader explicitly blocks the loading of JSON files via the uploader. We need to find a better way of storing these to load back via the API.

Is it still working? https://github.com/okfn/annotator-wordpress

Vision for OpenLiterature

• A deployment of Textus designed for hosting lots of public domain texts
• Find, view, explore, annnotate public domain texts

Vision for Textus

See this set of slides

• Componentized system for working with texts (especially more static texts)
• Specifications
• Implementation of a
  • Viewer (JS + HTML)
  • Annotation backend
  • Flat-file or more elaborate storage for texts and markup
  • Integration with Wordpress - one-click install of the above

Can we integrate with this #4 (full text search)

API spec from Jan 2014 Sprint

API [Annotation]

cf. http://okfnlabs.org/textus/doc/api.html

Base location:

/api/textus/1

The API returns JSON

GET  /annotation/text/{text-id}/{start}/{end}   # very specific search
GET  /annotation/?text=text-id&start=&end=      # search
POST /annotation/          # create
POST /annotation/{id}    # update

Other items (Not sure we need yet):

GET ol.net/api/textus/1/annotation/{id} # get annotation
GET ol.net/api/textus/1/text/{id}/ # get text

Data Structure

Database table in e.g. wordpress named:

textus_annotation

{
"start" : 300,
"end" : 320,
"type" : "textus:comment",
"userid" : [wordpress-id]
"private": false
"date" : "2010-10-28T12:34Z",
"payload" : {
    "lang" : "en"
    "text" : "Those twenty characters really blow me away, man..."
}

"start" : 0,
"end" : 3095,
"type" : "textus:source",
"userid" : [wordpress-id]
"date" : "2010-10-28T12:34Z",
"payload" : {
    "type" : "textus:scanned-image",
    "url" : "http://my.transcription.site/texts/mytext/pages/1.html",
    "data-url" : "http://my.transcription.site/texts/mytext/pages/1/image.png" }

Questions:

• Replace user userid pointing to wordpress id?
• Privacy and wordpress integration
  • private: false [default] / true
• Only owner (and site administrators) can view
• permissions - For later - but more general (so i can share with X people, with a group ...)

POST /annotation/ # create

POST an annotation object but no userid (wordpress will add that) and type will be auto added as textus:comment

{
"start" : 300,
"end" : 320,
"date" : "2010-10-28T12:34Z",
"payload" : {
    "lang" : "en"
    "text" : "Those twenty characters really blow me away, man..."
}

Will write to DB table

POST /annotation/{id} # update

Check that current user id = id of existing annotation
And check that incoming userid = existing

GET /annotation/text/{text-id}/{start}/{end} # very specific search

• Return all public annotations in that range
• Return all private annotaitons in that range where userid of annotation = current user id (or user is adminstrator)

Return format looks like:

{
status: HTML style codes
error: 'string'
total: 
results: [
"start" : 300,
"end" : 320,
"type" : "textus:comment",
"userid" : [wordpress-id]
"private": false
"date" : "2010-10-28T12:34Z",
"payload" : {
    "lang" : "en"
    "text" : "Those twenty characters really blow me away, man..."
} & so on
,
{annotation}
]

    }

Here's the repo for it: https://github.com/okfn/textus-wordpress

URL: assets.openliterature.net ...

Step-by-Step Plan

• Formatter to convert plain text or other formats to Textus format - https://github.com/okfn/textus-formatter
• Viewer - javascript viewer to display a textus text - https://github.com/okfn/textus-viewer
• Wordpress plugin to take a shortcode for a url and embed the textus viewer etc - https://github.com/okfn/textus-wordpress
• Annotation support (to be planned!)

Anticipated Basic Workflow

• Prep a text (by hand if you want)
  • https://github.com/okfn/textus-formatter/issues esp rufuspollock-okfn/textus-formatter#1
  • rufuspollock-okfn/textus-viewer#1
• Get that loading in textus-viewer
• Create textus-formatter and extract into it code for generating Textus formatted texts
• New theme for this to embed the viewer in wordpress

Tech Overview

Basic idea: "Textus" system - that's Wordpress with Textus plugins. For overview and intro see:

http://okfnlabs.org/blog/2013/03/27/next-steps-for-textus.html

Details

This google drawing shows the technical architecture

This drawing shows the overview of how the web app could work

Texts are processed into 2 components

• raw text
• syntax formatting (paragraphs, strong, em etc etc)
• NO annotation atm

Wordpress

On wordpress page: /shakespeare/hamlet/

In the page we have shortcode:

[textus:url]

url = url or raw text
url.typography.json = typography formatting

Replace [textus:url] is replaced by

<div id=”raw”>
raw text
</div>
<script src=”textus-viewer.js”></script>
<script type=”text/javascript”>
var textusTypography = typography;
// now boot textus viewer
viewer = new Viewer(rawText, typography);
</script>
</pre>

Appendix

• (now inlined above) New architecture - Need new component architecture ...