Scv-bilara released

Overview

The scv-bilara v0.2.2 Javascript library is now available for installation via npm install scv-bilara. Anagarika Sabbamitta and I have tested the API to meet all use cases needed by Voice and she is currently using the scv-bilara search script to assist with her own translation of Bhante Sujato’s EBTs into German for SuttaCentral. This library requires and has been tested with NodeJS v10.17.0. It is licensed as MIT open-source without restrictions.

The primary purpose of this library is to provide scriptable programmatic access to all segmented EBTs in bilara-data format for Linux platforms. Although translation import is supported, the primary use cases are for search and retrieval of segments and/or documents from any bilara-data repository. The scv-bilara library is tri-lingual, which may be of benefit to non-English translators working with Bhante Sujato’s own English translations as the source. PM us if interested.

Now that scv-bilara is stable, Voice will migrate its codebase to use scv-bilara for all segmented documents. Voice will still use the REST SuttaCentral API for legacy documents, but scv-bilara has more features and better performance than the existing Voice codebase for managing segmented documents.

We will be releasing maintenance fixes for scv-bilara over the coming year, but this will primarily be for tuning the English and Pali classification AI in the library. The Javascript API itself is stable and will be released as v1.0.0 early next year.

Motivation

The Voice codebase has grown too large to be maintained as a single package. We’re splitting up stuff into independent libraries that can be used without Voice. We’ll also be looking at packaging the Voice TTS engine, which will become an independent npm module suitable for arbitrary TTS applications.

10 Likes

scv-bilara v0.2.7 has been published.

Anagarika Sabbamitta and I have realized that it might be of some use to open an editor to look at the files found in the most recent search. This would be of particular use for global search and replace. The new script is ./scripts/bilara-edit.sh

For example, the following opens up vi to edit the seven files matching ‘root of suffering’:

./scripts/search.js root of suffering
./scripts/bilara-edit.sh vi

For vi and vim, the script will automatically set the search pattern. And if you make changes and save your file, it will change the actual bilara-data file. :warning:

We have also tested this with subl, which opens up Sublime Text.

NOTE: any given sutta id will correspond to two or three files (e.g., pli, en, de). The file chosen to edit is for the search language.

Bhante @Sujato, this new script may be of use for translators wishing to adjust their vocabulary and apply changes to past translations. Since standard editors are invoked, the translator has full control over changes made and can apply changes globally or locally as desired.

6 Likes

It’s just so incredible! I hardly do a little sigh over something that might be a little difficult with my work, and it doesn’t take long before @karl_lew comes up with another ingenious script that makes it easier!

Much heartfelt thanks for this! :pray: :heart:

4 Likes

Anagarika Sabbamitta, thank you very much for all your patience and hard work making scv-bilara possible. I really look forward to hearing DN33 and many other suttas auf Deutsch as well as other languages. If you or any other translator has Bilara tasks that may be automatable via scv-bilara, please RSVP and we will all work on it together. :pray:

4 Likes

@blake @HongDa @Aminah please have a look at this, it is a great step forward!

It may be that the work done here can help with building a search interface for Bilara itself, or anyway, that’s my hope.

4 Likes

scv-bilara v1.0.4 has been published to npm.

Anagarika Sabbamitta and I have put together scripts that translators can use to update sc-voice/bilara-data with their own translations using the Linux editor of their choice. These changes will not be in SuttaCentral bilara-data. They will be in sc-voice/bilara-data. Anagarika Sabbamitta has completed AN1 and AN2 in sc-voice/bilara-data and she is using the scripts for her translation.

The scripts basically allow one to create a Github branch for translating a new sutta and publish that branch to master when ready. For those interested, we recommend SSH agents to ease the burden of entering usernames and passwords for Git usage.

With Bhante Sujato’s permission and blessings, these translations can be pulled into SuttaCentral bilara-data for display on the main SuttaCentral web site. Until then, the translations on sc-voice/bilara-data will only be available on Voice v1.8, which will be released this year with samples of Anagarika’s Deutsche translations made from Bhante Sujato’s own English source translations.

3 Likes

scv-bilara v1.0.16 has been published to npm

This release supports retrieval of individual segments in English:

…or Deutsch:

Search also has a new setting for Markdown output that will be useful for quoting EBTs in forums such as Discuss & Discover. For example:

./scripts/search nandi dukkhassa -outMarkdown

The output is formatted with a Markdown hyperlink for every quoted segment. Just copy the desired lines into your post and you will get something like this:

mn1: Because he has understood that relishing is the root of suffering,

Bilingual Markdown quotes are also supported:

./scripts/search mulam dukkhassa -outMarkdown2

sn42.11: Chando hi mūlaṃ dukkhassā’ti.
sn42.11: Denn Sehnen ist die Wurzel des Leidens.‘

2 Likes

Hey, you are building ever cooler things!!! :open_mouth:

2 Likes