Skip to content

Latest commit

 

History

History

manager

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
scripts
scripts
 
 
src
src
 
 
static
static
 
 
README.md
README.md
 
 
package.json
package.json
 
 
yarn.lock
yarn.lock
 
 

README.md

bmsurge-rebooted/manager

Environment variables configuration

# .env
MONGO_URL=mongodb+srv://<user>:<password>@<host>/<db>
WORKER_URL=https://<domain>.run.app
MP3_URL_PATTERN=https://<domain>/<path>/%s.mp3
SONG_UPDATER_URL=https://<user>:<password>@us-central1-be-music-surge-frontend.cloudfunctions.net

Install dependencies

yarn

Preparing a BMS event for rendering

Create a ZIP file for each song in the event. The ZIP file should meet the following requirements:

  1. Filenames must be encoded in UTF-8.

  2. Files must reside at the root of the ZIP file. They may not be inside a directory.

Upload all the ZIP files to a web server where it can be downloaded by the worker.

Importing URLs

Create a file containing list URLs to the ZIP files to render, name it as <eventId>.urls.json. For example, g2r2018.urls.json may look like this:

[
  "https://<url>/%21%21%21%21%21%21%21%21KANPAI%21%21%21%21%21%21OGG%E3%81%A0%E3%81%9E.zip",
  "https://<url>/%21littlegirl_complex.zip",
  "https://<url>/%28U.F.M.%29.zip",
  "https://<url>/03_twinkle_memories_001.zip",
  "https://<url>/5-Acht-Omegatische_Liebe.zip",
  "https://<url>/5192296858534827628530496329220096.zip",
  "https://<url>/6-LU-zemerim.zip",
  "https://<url>/7ost%20chapt3r%20ogg.zip",
  "https://<url>/8-lu_remixed_by_kju8-My_CAKES_AND_ALE.zip",
  "..."
]

Then, import the URLs into the database using:

node src/index.js import <path/to/event>.urls.json

The above command will do a dry run. To confirm the changes, add -f. This will import the songs into the MongoDB database.

Render the BMS songs to MP3

node src/index.js work

This will read from the database to find all songs that need to be rendered, then it will invoke the worker service that has been deployed to Google Cloud Run. Thanks to Google Cloud Run, we can render hundreds of songs simultaneously. From my test 1,000 songs can be rendered in less than 30 minutes!

Update the songlist on downstream services

node src/index.js songlist -o private/_songlist.json --update --index=YYYY-MM-DDT00:00:00Z

This will read from the database to find all songs that have been rendered successfully, and generate a songlist.

  • -o saves the generated songlist to a JSON file.
  • --update uploads the songlist to the dj system, which allows the song to be streamed on the station.
  • --index uploads the songlist to the song request system’s search index. Due to Algolia’s limits, we can’t upload the full songlist; we must upload only the changes to stay within the free operations limit. Send in a timestamp to specify the minimum update time to index.

Web-based dashboard (unused)

env PORT=8082 node src/index.js server

This will start a web server that displays the dashboard.