search

concord-consortium / rigse

Concord Consortium Portal (somewhat like an LMS)
MIT License
9 stars 11 forks source link

readme

CC Rails Portal Activity Authoring, Deployment, and Reporting System

Code Climate

Setup

Prerequisites

Working git, ruby, and rubygems, wget

Core Extensions

Simple Getting Started

Using Docker

Install Docker and make sure that docker-compose is installed too (it should be part of the standard Docker installation).

git clone git@github.com:concord-consortium/rigse.git portal
cd portal
docker-compose up # this will take 15 minutes to download gems

Increase memory available to Docker to 4GiB-5GiB (OSX: Preferences... -> Advanced tab).

Now open your browser to http://0.0.0.0:3000. On OS X this might take more than 5 minutes to load the first page. Look in the terminal where you ran docker-compose up to monitor progress.

Visit the Docker docs for how to use your portal running in docker. This also includes: instructions on speeding things up on OS X, using a local dns+proxy system to avoid port conflicts, and setting up ssh for capistrano deploys.

Tests

After getting the server running it's good to confirm that all the tests pass before changing any code.

Prepare a database for use when running the spec tests:

RAILS_ENV=test rake db:create
rake db:test:prepare
rake db:feature_test:prepare
RAILS_ENV=cucumber rake app:setup:create_default_data

Start SOLR in test environment (it works with cucumber tests too):

RAILS_ENV=test rake sunspot:solr:run

Run the rspec unit tests:

rspec spec/

Prepare a database for use when running the cucumber tests:

RAILS_ENV=feature_test rake db:create
rake db:feature_test:prepare
RAILS_ENV=cucumber rake app:setup:create_default_data

Run the cucumber integration tests:

cucumber features/

All these tests should pass. If you add features make sure and add tests for these new features.

SSO Clients and LARA (authoring) integration

These instructions assume that you are setting up LARA and Portal using docker-compose files. It also assumes some prerequisites:

With the following settings you can:

Note: Student data and resource structure will be stored in the hosted Firestore database of report-service-dev. This is a used by the portal-report. Because of this, you must be online.

Starting from scratch on a Mac
  1. in the Portal: cp .env-osx-sample .env
  2. in the Portal .env set PORTAL_PROTOCOL=https
  3. start up the Portal: docker-compose up
  4. Add a Firebase App for report-service-dev with docker-compose exec app bundle exec app:setup:add_report_service_firebase_app, it will ask for the "private_key", paste in the private key from report-service-dev firebase app in learn.staging.concord.org
  5. in LARA: cp .env-osx-sample .env
  6. in LARA .env:
    1. set REPORT_SERVICE_TOKEN (see the comment in the .env-osx-sample file)
    2. set LARA_PROTOCOL=https
    3. set PORTAL_PROTOCOL=https
  7. start up LARA: docker-compose up
  8. Setup admin access to LARA using a portal SSO login:
    1. Go to https://app.lara.docker
    2. Click log in, and choose "localhost"
    3. This will take you to the portal (app.portal.docker)
    4. Log in with admin, password
    5. You are now logged in with admin@concord.org in LARA, however this user is not actually an admin in LARA
    6. Run the following command in terminal in the LARA folder: docker-compose exec app bundle exec rake lightweight:admin_last_user

Notes:

Updating an existing setup
  1. In the Portal, edit .env:
    1. Append docker/dev/docker-compose-lara-proxy.yml to the COMPOSE_FILE var.
    2. If your local portal domain is not app.portal.docker, then set PORTAL_HOST
    3. set PORTAL_PROTOCOL=https
    4. If your lara host name is not app.lara.docker, then set LARA_HOST
  2. Stop your portal services if they are running, and update them with docker-compose up
  3. In the Portal, as an administrator:
    1. Create or update an "Auth Client". Using the following settings: 
      Name: 'localhost'
App Id: 'localhost'
App Secret: 'unsecure local secret'
Client Type: confidential
Site Url: 'https://app.lara.docker'
Allowed Domains: (leave blank)
Allowed URL Redirects: 'https://app.lara.docker/users/auth/cc_portal_localhost/callback'
    2. Check the external report called DEFAULT_REPORT_SERVICE. It should have a URL of: https://portal-report.concord.org/branch/master/index.html?sourceKey=app.lara.docker.username Replace username with the username on your local system. If you don't know your username, run echo $USER. Note: this sourceKey param needs to match the value of LARA's REPORT_SERVICE_TOOL_ID environment variable. And by default this variable is configured to be app.lara.docker.${USER}.
    3. Add a Firebase App for report-service-dev, copy the values for the report-service-dev firebase app in learn.staging.concord.org
    4. Add a new "Auth Client" for the portal-report 
      Name: 'Portal Report SPA'
App Id: 'portal-report'
App Secret: (leave default value, this isn't used)
Client Type: public
Site Url: 'https://portal-report.concord.org'
Allowed Domains: 'portal-report.concord.org'
Allowed URL Redirects: 'https://portal-report.concord.org/branch/master/index.html'
  4. In LARA, edit .env:
    1. Append docker/dev/docker-compose-portal-proxy.yml to the COMPOSE_FILE var.
    2. Set PORTAL_HOST to app.portal.docker or whatever domain your local portal is
    3. Set PORTAL_PROTOCOL=https
    4. set LARA_PROTOCOL=https
    5. Set REPORT_SERVICE_TOKEN (see the comment in the .env file)
    6. set REPORT_SERVICE_URL (see the value in the .env-osx-sample file)
  5. Stop your Lara services if they are running, and update them with docker-compose up
  6. If you want admin access to Lara when signing in with a portal user, you will need to first login to LARA with this portal user. And then either:
    • use the rails console in LARA to set the is_admin flag of the newly created user.
    • use an existing admin in LARA to make the new user an admin.
  7. Setup Activity Player support
    1. Add a Tool to the portal with: 
      Name: ActivityPlayer
Source Type: ActivityPlayer
Tool ID: https://activity-player.concord.org
    2. Make an external report. In the Url field below replace the username in the sourceKey parameter with your local username: 
      Name: AP Report
Url: https://portal-report.concord.org/branch/master/index.html?sourceKey=app.lara.docker.username&answersSourceKey=activity-player.concord.org
Launch text: AP Report
Client: DEFAULT_REPORT_SERVICE_CLIENT
Report Type: offering
Allowed For Students: true
Default Report For Source Type:
Report available for individual students: true
Report available for individual activities: true
Use Query JWT: false
Move Students API URL:
Move Students API Token:
    3. Add this AP Report external report to each AP resource you publish the portal from LARA.
Troubleshooting

When you run the portal-report if you just see a spinner. Here are some steps to try:

  1. Verify the source is be added to firestore:
    1. Go to the firebase console and open the report-service-dev firestore.
    2. Look in the sources collection
    3. You should see a app.lara.docker.{$USER} collection (USER is your local username)
    4. If you don't see this collection then your LARA is not properly publishing the report structure to the report-service. In LARA check the values of REPORT_SERVICE_TOKEN and REPORT_SERVICE_URL (see above). A less common error would be a misconfigured LARA_HOST and TOOL_ID setup.
    5. After fixing these values update your lara app so it picks up the variables with docker-compose up. And make a change to your activity so it republishes the structure to Firestore. Check that the app.lara.docker.{$USER} collection is there now.
  2. Verify the resource structure added to firestore has the right URL:
    1. Go to the firebase console and open the report-service-dev firestore.
    2. Inside of the sources/app.lara.docker.{$USER} collection will be a resources collection. Inside of this will be a document for each activity or sequence that you've published from LARA.
    3. In these documents look at the url field. It needs to exactly matches what the portal-report is looking for. If it starts with http: instead of https: then you will see the spinner.
    4. To know what the portal-report is looking for in firestore: when you run the portal-report look at the network dev tools and look for the offering info request it makes to the portal. This request's url will look something like: https://app.rigse.docker/api/v1/offerings/[id]. Look at the response to this request and find the activity_url field.
    5. To fix the http: instead of https: problem make sure your LARA_PROTOCOL is set to https in your lara .env file. Then update your lara container with docker-compose up. And then make a change to the LARA activity or sequence to republish it. Verify the url field in firestore has been updated.

Virtual host settings (currently used for automation)

If you want to change the portal url from "app.portal.docker" to "learn.dev.docker", please follow the below steps:

  1. In the Portal, edit '.env' file and update PORTAL_HOST as learn.dev.docker
  2. In the Portal, edit '.env' file and update PORTAL_PROTOCOL as https for automation
  3. In the Portal, as an administrator, edit the Auth Client settings: 
    Site Url: 'https://learn.dev.docker'
Allowed URL Redirects: 'https://learn.dev.docker/users/auth/cc_portal_localhost/callback'

GitHub Codespaces

Github Codespaces is a cloud-based development environment. We are currently using it to do development work on LARA and Portal since it’s proven difficult to do local development on those codebases on M1 MacBooks.

Github’s documentation for Codespaces can be found at here.

You will need to set up separate codespaces for LARA and the Portal.

Use of Codespaces incurs an hourly cost. The amount is not a lot, but it should be kept in mind. Codespaces will shut themselves down automatically after a period of inactivity, but it would be best to manually shut them down when you’re done working in order to minimize cost.

You can use Codespaces in web browser or you can connect to selected machine from desktop Visual Studio Code if you install a Codespaces extension.

Basic setup

Once machine is up and running, most of the steps described for local development are still valid for GH Codespaces. The main difference is that you should copy .env-gh-codespaces-sample to .env (instead of .env-osx-sample), there's no need for Dinghy setup, and LARA and Portal hosts will be significantly different. However, everything you need to do in practice is described below.

  1. Run:

      cp .env-gh-codespaces-sample .env

  2. Open LARA GitHub Codespace, run echo ${CODESPACE_NAME} in terminal, and set LARA_CODESPACE_NAME variable in Portal's .env file.

  3. Run

      docker-compose up

  4. Once the app has started, open "Ports" tab in Visual Studio Code. Find a process that uses port 3000 and change its visibility to public (right click on "Private" -> Port Visibility -> Public). You should see an updated address in "Local Address" column. You can open this URL in the web browser and Portal should load. It seems it's necessary to do it each time you run docker-compose up.

  5. Open Portal, login as admin (password: password), and go to Admin tab.

    Go to Firebase Apps and create two new apps:

    • report-service-dev
    • token-service

    Client emails and private keys can be copied from learn.staging.concord.org.

Now, your Portal instance should work with LARA, Activity Player and basic reports.

Theme support & Rolling your own theme:

We are using the themes_for_rails gem Theme views go in app/themes/(name)/views/ Theme assets go in app/assets/theme/(name)/ Sample config files go in config/themes/(name)/settings.sample.yml

For now the best thing to do is to copy an existing theme. eg:

mkdir ./config/themes/<new_theme_name>

# configuration files:
cp ./config/themes/<old_theme_name>/settings.sample.yml
./config/themes/<new_theme_name>

# view files:
cp -r ./themes/<old_theme_name> ./themes/<new_theme_name>

# assets:
cp -r ./app/assets/themes/<old_theme_name>
./app/assets/themes/<new_theme_name>

# finally change the theme setting in your config/settings.yml
open config/settings.yml

NCES District and School Tables

When a rails-portal instance is created two tables containing data for schools and districts in the US are created from data supplied by the National Center for Education Statistics (NCES).

NCES maintains a database about US districts and schools called the Common Core of Data

The rake task: portal:setup:create_districts_and_schools_from_nces_data downloads 2006 NCES CCD data files from NCES website and imports data from these data files into the following models:

Only data from states and provinces identified in the config/settings.yml for the portal instance are imported.

The NCES district and school models are used to provide data from which districts and schools actively using the portal are be created.

The Portal::Nces06District includes about 50 different fields of data for each district.

The Portal::Nces06Schoolincludes about 500 different fields of data for each school.

PDF documentation for the NCES data schemas

Testing

Testing Frameworks

Rspec, Rspec-rails

Cucumber / Capybara

Feature specs that require javascript are run by Chrome via Selenium. By default Chrome will run in headless mode and there is nothing special you need to do inside of a Docker development environment.

However, if you would like to run Chrome in non-headless mode on your host machine, this is possible by making the following changes:

Factory Bot

factory_bot allows you to quickly define prototypes for each of your models and ask for instances with properties that are important to the test at hand.

Running the rspec tests

Running all the rspec tests:

bundle exec rake spec

Running a single file:

bundle exec rake spec SPEC=spec/routing/dataservice/bundle_contents_routing_spec.rb

Running a single directory:

bundle exec rake spec SPEC=spec/routing/dataservice

Running all the controller tests:

bundle exec rake spec SPEC=spec/controllers

Running the feature tests with cucumber

Running all the feature tests: bundle exec rake cucumber

Running all the feature tests using the ci_reporter gem that's used on the hudson CI system: bundle exec rake hudson:cucumber

Running a single feature:

bundle exec cucumber features/student_can_not_see_deactivated_offerings.feature

Running a single feature in non-headless mode:

HEADLESS=false bundle exec cucumber features/student_can_not_see_deactivated_offerings.feature

Using binding.pry with Cucumber tests

Problem:

Integration tests are difficult to debug without accessing the content in the browser and inspecting the relevant elements. Using debugging tools in the command line or trying to view the problem from a screenshot is not helpful when the problem might be a hidden link or different element type, for example.

Solution:

Using pry in non-headless mode in Chrome opens a new Chrome window showing you the state of the page where pry has paused the test. You can inspect elements in the page at that point in time to more easily identify the problem.

How to use:

Follow the instructions above to set up and start chromedriver.

For a particular cucumber test where JavaScript is enabled, find the step you want to test:

And I follow "Admin"

Find the corresponding step definition and insert binding.pry:

When /^(?:|I )follow "([^"]*)"$/ do |link|
  binding.pry
  first(:link, link).click
end

Make sure chromedriver is running and run the test with HEADLESS=false prepended to the path

$ HEADLESS=false bundle exec cucumber features/admin_accesses_special_pages.feature

When pry is hit, a new Chrome window will pop up where you can inspect element and use the pry in the command line as usual.

note: Please see documentation regarding running chromedriver on your host machine above ☝️.

Understanding the Codebase

Some video walk-throughs

The Page Elements Model Part I:

screencast: The Page Elements Model PartI

Install github version of railroad with aasm patches from ddolar's repo

Generate a graph of the projects models using railroad: railroad -o models.dot -M

Open that file with omnigraffle, or traslate to some other image format using the dot tool.

Page Elements Model Part II:

screencast: Page Elements Model PartII

Using mysql query browser to view schema: Mysql gui-tools

Use the generator to generate page elements eg:

./script/generate element xhtml content:text

PageElement View partials:

screencast: PageElement View partials

Shows the relationship between:

HAML, Compass and SASS:

screencast: HAML, Compass and SASS

Brief introduction to the technologies generally, and how we use them specifically

Javascript use in Portal

screencast: Javascript use in Portal

Javascript librararies we are using, and what things we have written by ourselves; Stuff we did:

Capistano Tasks:

Updating a staging server.

Once a development branch has been deployed to a development server, tested and found reliable enough to deploy to staging here's how to do it.

Our convention is to create dev, staging, and production branches in the git repository following that use the same name.

For example the xproject family have the following capistrano stages and branches in teh git repository:

In the code below I will assume that we are using the xproject series of stages and branches.

If you don't already have a local branch of staging

git branch  --track xproject_staging origin/xproject_staging

Switch to the staging branch and merge from xproject_dev

git co xproject_staging
git merge xproject_dev

Push your copy of the staging branch to the gihub repository:

git push origin xproject_staging

Dump the production database to this file db/production_data.sql on the production server, download it to the local folder db/production_data.sql, the cleans up the production db/ folder.

cap xproject_production db:fetch_remote_db

Push the production database from the local db/production_data.sql to the staging server, then import the data into the database on staging, then cleanup.

cap xproject_staging db:push_remote_db

Run any migrations on the staging server:

cap xproject_staging deploy:migrate

There may be rake tasks that need to be run to update or fix data in the database.

These should have corresponding capistrano tasks.

Test the staging server: http://xproject.staging.concord.org/

If the authors confirm that there are no blockers then let people know when the update will take place and perform these tasks on the production server.

other Rake tasks:

haml

We use haml for some templates, see: http://haml.hamptoncatlin.com/

To install this plugin we followed this procedure:

  1. gem install --no-ri haml
  2. haml --rails path/to/rigse_app

Rendering

haml

We use haml for some templates, see: http://haml.hamptoncatlin.com/

to install this plugin we followed this procedure:

  1. gem install --no-ri haml
  2. haml --rails path/to/rigse_app

Authentication, Sessions, and Cookies

User authentication with Devise

Devise is already setup. The routes are setup, along with the mailers and observers. Forgotten password comes setup, so you don't have to mess around setting it up with every project.

Devise uses the pepper parameter within settings.yml to encrypt user passwords. A default pepper is provided in settings.samles.yml You need to change this when deploying to a public server.

Devise is also setup to use user activation. Users which require activation are sent emails automatically.

Single Sign On

The app supports user registration and authentication via third party single sign using OAuth. Using this feature requires setting up OAuth credentials with the third party.

The app currently supports OAuth registration and authentication using Google and Schoology.

To set up single sign on with Google, follow the steps below.

1) Create a new Google app in console.developers.google.com.

2) Create a set of OAuth credentials for the app. For the "Authorized JS origin" value use the valid public domain that resolves to your copy of the app on the web. For the "Authorized redirect URI" value use https://[your domain]/users/auth/google/callback.

3) Take the Client ID and Client Secret values created in step two and add them as values for GOOGLE_CLIENT_KEY (Client ID) and GOOGLE_CLIENT_SECRET (Client Secret) in your app's .env file.

4) Restart the app.

Testing Single Sign On with Google and a Local Portal

Create a set of OAuth credentials for the app following the steps above, but for the "Authorized JS origin" value use a valid public top level domain that resolves to 127.0.0.1. Google won't accept 127.0.0.1 or a domain like app.portaldocker.local. An easy option is to use https://lvh.me which resolves to 127.0.0.1 without requiring any special configuration of your computer.

For the "Authorized redirect URI" value use https://[your domain]:[your port number]/users/auth/google/callback. If, for example, you use lvh.me and Docker is serving your portal over port 32789, the value would be https://lvh.me:32789/users/auth/google/callback.

Uses the Database for Sessions

Will Paginate

We use will_paginate in pretty much every project we use.

Exception Notifier

You don't want your applications to crash and burn so Exception Notifier is already installed to let you know when everything goes to shit.

config/initializers/exception_notifier.rb does the setup. Currently it reads "admin_email" from config/settings.yml and use it as the destination address. The setup can be modified to include multiple email addresses. See the homepage readme of exception notifier.

Bug

It seems rails 2.3.3 and 2.3.4 fails to deliver emails when someone passes multiple destination addresses as an array, which exception notifier does. config/initializers/fix_mailer_on_rails_2.3.4.rb fixes the problem.

The code is borrowed from Dmitry Polushkin

Databases

On OS X the mysql2 gem usually can't find the mysql client library that it needs to run. The command below fixes that. It assumes your mysql is installed in the default basedir of /usr/local/mysql/lib. And it assumes you are using bundler.

install_name_tool -change libmysqlclient.16.dylib /usr/local/mysql/lib/libmysqlclient.16.dylib \
  `bundle show mysql2`/lib/mysql2/mysql2.bundle

For newer versions of rvm and mysql2, you will see an error like this

dlopen(/Users/scytacki/.rvm/gems/ruby-1.9.3-p545/extensions/x86_64-darwin-13/\
  1.9.1/mysql2-0.3.15/mysql2/mysql2.bundle, 9): Library not loaded: libmysqlclient.18.dylib
  Referenced from: /Users/scytacki/.rvm/gems/ruby-1.9.3-p545/extensions/x86_64-darwin-13/\
    1.9.1/mysql2-0.3.15/mysql2/mysql2.bundle
  Reason: image not found - /Users/scytacki/.rvm/gems/ruby-1.9.3-p545/extensions/x86_64-darwin-13/\
    1.9.1/mysql2-0.3.15/mysql2/mysql2.bundle

So then to fix a command like this is needed:

install_name_tool -change libmysqlclient.18.dylib /usr/local/mysql/lib/libmysqlclient.18.dylib \
  /Users/scytacki/.rvm/gems/ruby-1.9.3-p545/extensions/x86_64-darwin-13/1.9.1/mysql2-0.3.15/mysql2/mysql2.bundle

CSS

Rails 3 Asset Pipeline

The portal uses the"Rails 3 Asset Pipeline":http://guides.rubyonrails.org/asset_pipeline.html stylesheets, images, javascript should all be placed in the app/assets/ directory structure. theme specific assets should be placed in app/assets/theme/themename/.

You should read the Rails guide for more information about the asset Pipeline.

themes_for_rails has been configured to play nice with the asset pipeline. See the initializer: config/initializers/themes_for_rails.rb

cap deploy should trigger the rake task assets:precompile to run.

You can also run locally by hand: bundle exec rake assets:precompile

The assets will be compiled to public/assets which should be ignored by .gitignore

When running in development mode you do not need to pre-compile your assets.

Solr & Sunspot

Sunspot is being used to provide search capabilities.

You will need to create solr cores if you want to update materials or publish. At the least you will need a 'development' solr core. Here is a basic set of directions:

  1. Make sure there are no solr processes running with ps auxxwww | grep solr
  2. scp -r deploy@learn.staging.concord.org:/web/portal/shared/solr-template solr
  3. cp -r ./solr/production ./solr/development
  4. vim ./solr/development/core.properties (change name from production to development)
  5. bundle exec rake sunspot:solr:start
  6. bundle exec rake sunspot:solr:reindex (edited)

You could also create a test core by repeating steps 3 & 4.

In development mode you will need to create an index and start sunspot:

bundle exec rake sunspot:solr:start
bundle exec rake sunspot:reindex

You can then visit the web interface to the solar server by visiting localhost:8982/solr/admin/. Though I haven't found any good reason to do so.

Rspec testing with sunspot disabled & enabled:

For rspec tests see the helper methods defined in spec/support/solr_spec_helper.rb

For cucumber tests, you can use "Given The materials have been indexed" to update solr indexes after fixture data has been loaded.

https://github.com/sunspot/sunspot/wiki/RSpec-and-Sunspot

Solr delpoyment and index-updating

If you make changes to how Solr does its indexing, you will have to run a cap task to tell it to reindex:

In theory a simple bundle exec cap <host> solr:reindex should work, but to be sure use: bundle exec cap <host> solr:hard_reindex to restart and reindex.

Application Settings & Settings YAML

There is a settings.yml file that contains site-wide stuff. The site name, url and admin email are all used in the Devise mailers, so you don't need to worry about editing them.

Database YAML

Enabling features via environment variables

Certain features of the portal are controlled via environment variables.

The PORTAL_FEATURES environment variable can take a string of the form "feature1 feature2" to include the following features:

If CORS is enable, by default it will allow any request from '*.concord.org', to any route, but can be controlled by two additional environment variables:

Technical debt.

Here is a brief list of things which need to be looked into:

Misc

Archiving a portal

A set of rake tasks is available under the archive_portal namespace that enable the portal data to be archived. These tasks are:

The rake tasks use a config file at /config/archive_portal.yml to specify the S3 bucket parameters to use when extracting images and to use when generating the url to those images in the reports. A /config/archive_portal.sample.yml file exists to be copied and updated with real values.

These tasks will take a long time. Easiest way to run them is to ssh to running server and run them in the background using nohup, e.g.:

nohup bundle exec rake archive_portal:extract_and_upload_images &

You can close your ssh session and the task will be still running. Logs will be saved in nohup.out.

New Admin interfaces

As of 2020-06-12 we are in the process of moving some administrative functions to a new technology stack. See the New Admin interface documentation

License

CC Rails Portal is released under the MIT License.