Skip to content

Latest commit

 

History

History
executable file
·
51 lines (46 loc) · 3.37 KB

README-MSM.md

File metadata and controls

executable file
·
51 lines (46 loc) · 3.37 KB
Raw

Alexa

Overview

The Alexa skill for reading the Samson House daily devotion is hosted on a Raspbery Pi server which is physically located at Mike Moore's house in Florida. (See documentation for Samson House Raspberry Pi for additional details.)

Workflow Concept

Step 1

Tom Mouka records the audio file for the daily devotion and saves it with a numeric sequence file name.

Step 2

Tom Mouka uploads the recorded audio file into the Samson House Google Drive account ([email protected] - see password spreadsheet for credentials) into the Alexa Files folder

Step 3

An automated script runs every morning at 00:05a ET which processes the next numerically sequenced audio file and moves it into a publication folder where Alexa can retrieve it via API.

Step 4

Alexa pulls the audio file via API on-demand as needed to play the skill

Requirements

Daily Automation for Samson House Alexa Skill Depends on Google Drive API for Python. (pip install --upgrade google-api-python-client google-auth-httplib2 google-auth-oauthlib) Requires a file on the root containing the path to the working directory. Requires a credentials.json file in the working folder.

Script Process

Downloads the days file from Google Drive (folder??) Uses FFMPEG to increase the volume by 10dB Renames it to change all spaces into underscores Posts it in the alexa directory to be hosted by a webserver Also creates the xml for Alexa to identify the stream. Uses FFMPEG again to convert the boosted file to ogg for the webpage hosted audio file Creates a symlink to the m4a version that is used by the webpage hosted audio files

Setup

  1. Run the script the first time with --noauth_local_webserver to authenticate with the appropriate Google Drive to convert the credentials.json to a token.json. The --noauth_local_webserver argument is not required if a valid token.json exists in the working directory.
  2. Finally place a text file containing the current 3 digit number of the Samson House .m4a file. The script will search for this file to determine what tomorrow's number is to download it from google drive.
  3. Create a crontab entry to call alexa.py at an appropirate time each day. The script uses the current date when building the xml file for the rss feed, so I recommend after midnight, but it should work either way.
  4. A second crontab entry should be created to run the verifyNext.py at an apporpriate time to send a reminder if the next days file isn't uploaded to Google Drive.

Troubleshooting & FAQ

  • If the Alexa script does not find the file needed for the day after tomorrow, then an Alert message is posted into the automation channel in the Samson House slack workspace and an email is sent to Rhys and Tom Mouka.
  • If the Alexa script does not find the file needed for tomorrow, then a Warning message is posted into the automation channel in the Samson House slack workspace and a notification e-mail is sent to Rhys and Tom Mouka
    • Contained in that warning e-mail is a link to the Google folder to upload the next audio file and a link to a button that will manually trigger the script
  • If the next audio file is uploaded after 00:05 ET on the day of publication, then Alex will continue to play the previous day's audio file until the new file is processed and moved into the publication folder. An on-demand link is available to trigger the Alexa processing script after the file is uploaded.