aggregate-class-calendar

Aggregate schedules from various classes (e.g. yoga studios) into one place, given search parameters.

Installing

1. Meteor -- https://www.meteor.com/install for info
2. NPM modules for project, from top dir $ meteor npm install
3. mongodb 3.4 -- https://docs.mongodb.com/manual/administration/install-on-linux/
4. google-chrome on Ubuntu $ wget https://dl.google.com/linux/direct/google-chrome-stable_current_amd64.deb $ sudo dpkg -i google-chrome-stable_current_amd64.deb
5. Other dependencies $ sudo apt install -y nginx supervisor

Chrome dependency

Run headless Chrome using: $ google-chrome --headless --disable-gpu --remote-debugging-port=9222 If there's an issue getting to the webpages, it can be debugged using an X session $ ssh -X ...

Deploying

0. Create the user to run the project $ adduser aggregate
1. Build the application, tar the deploy directory $ meteor build ../app --architecture os.linux.x86_64 $ tar -czvf ../app/aggregate-deploy.tar.gz deploy/
2. Copy the resulting tar files $ scp ../app/aggregate-class-calendar.tar.gz aggregate@:/home/aggregate $ scp ../app/aggregate-deploy.tar.gz aggregate@:/home/aggregate
3. Log into the machine
4. Unbundle the tars $ tar -xvzf aggregate-class-calendar.tar.gz $ tar -xvzf aggregate-deploy.tar.gz
5. Update or install dependencies $ cd bundle/programs/server && meteor npm install
6. Setup nginx $ cp deploy/nginx-prof.conf /etc/nginx/sites-available/aggregate-class-calendar.conf $ ln -s /etc/nginx/sites-available/aggregate-class-calendar.conf /etc/nginx/sites-enabled/aggregate-class-calendar.conf $ rm /etc/nginx/sites-enabled/default # if present $ service nginx reload
7. Setup database $ systemctl enable mongod $ service mongod start
8. Setup supervisor to run the app $ cp deploy/supervisor-proj.conf /etc/supervisor/conf.d/aggregate-class-calendar.conf $ supervisorctl update aggregate-class-calendar
9. If needed, restarting the whole project $ supervisorctl restart aggregate-class-calendar

Important Components

Upload courses from file

1. Run the subtree's toplevel.js to dump courses.json file
2. Test the file is OK $ python -m json.tool courses.json
3. Copy the file $ scp courses.json :/home/aggregate/bundle/programs/server/assets/app
4. Use the admin panel to upload from file

Cron jobs

Currently under construction, but cron job customization will be possible through the admin panel and percolate:synced-cron.

admin panel configuration, lib/adminconfig.js

To work with the admin dashboard, which uses yogiben:meteor-admin, an admin user must be created through console and added to the "admin" role. Currently, this only works with one user within 'adminEmails'. To add a user during initial setup, use the following command from meteor shell, through the shell-server package: $ var id = Accounts.createUser({email: "..", password: "..", profile: { name: ".." }}); $ Roles.addUsersToRoles(id, ['admin'], 'default-group');

logger setup, server/logging.js

This app uses winston through votercircle:winston and a papertrail config. See https://papertrailapp.com/systems/setup for setup instructions on a machine to send logs to their severs. See the logs at https://papertrailapp.com/dashboard The file logging.js adds the transport as necessary to Papertrail.

dependency on class-scraper

This repo depends directly on class-scraper through git subtree (repo found at https://github.com/joncinque/class-scraper), so to update this repo on any changes, run the command: $ git subtree pull --prefix .class-scraper https://github.com/joncinque/class-scraper.git master --squash Or, using aliases: $ git sbu https://github.com/joncinque/class-scraper.git .class-scraper In order to push back up, run the command: $ git subtree push --prefix .class-scraper class-scraper master Or, using aliases: $ git sbp .class-scraper class-scraper master Quick explanation: the prefix of ".class-scraper" is looking in that folder for the push information, "class-scraper" is the name of the remote we're pushing to, and "master" is the branch we're pushing from.