Implement API v3

[jag3773]

Story

As a developer|admin I want an API endpoint that provides access to all the content in our Door43 ecosystem so that I can write awesome apps to make use of the content.

Notes

Current API v3 Spec

The latest version of these specifications are at:

• https://github.com/unfoldingWord-dev/door43.org/wiki/API-v3-Resource-Catalog-Endpoint
• http://resource-container.readthedocs.io/en/latest/

This should include documentation of the new setup, possibly in another readthedocs site that overviews our API setup.

Blocked Issues Until This is Implemented

The following is a non-comprehensive list of issues that can't be completed until we implement v3 of the API:

• Add USFM Downloads for Bible unfoldingWord.github.io-old#142
• OBS Streaming Video unfoldingWord.github.io-old#102
• OBS Audio Streaming unfoldingWord.github.io-old#101
• Start Including Chunk Zero Pages in tN unfoldingWord-dev/tools#166
• update indexing to include the chunks.json file unfoldingWord-dev/ts-requirements#75
• Final chunk of chunks.json has incorrect id unfoldingWord-dev/tools#150
• Need to make attribution for content in app unfoldingWord-dev/ts-requirements#110
• QA Routine for USFM unfoldingWord-dev/tools#157
• Fix Regex in tN Publishing unfoldingWord-dev/tools#156
• Alternate Versification unfoldingWord-dev/translationDatabaseWeb#448
• Import Hebrew for tS unfoldingWord-dev/Door43#381
• Import Greek Text for tS unfoldingWord-dev/Door43#380
• Update OBS JSON Format unfoldingWord-dev/translationDatabaseWeb#290
• Publishing Schema: Include URLs to Documents unfoldingWord-dev/translationDatabaseWeb#221
• Update date_modified on terms unfoldingWord-dev/Door43#302
• OBS ODT and DOCX unfoldingWord-dev/Door43#246
• Staging Area for API Content unfoldingWord-dev/translationDatabaseWeb#88
• Audio: Publishing to API unfoldingWord-dev/translationDatabaseWeb#79
• Bible Publishing unfoldingWord-dev/translationDatabaseWeb#69

In addition, this will significantly speed up the Update mechanism in tS since they will only have to make 1 HTTP request to get the entire catalog instead of hundreds.