wjf136 / final Goto Github PK

== README

My app is to show the details of MMA fights with the particular interest of developing heuristics on judges scoring.
CUrrently, it just maintains all the details required to develop such statistics. That would include a fight, the fighters
in that fight, the fight card that fight was on, as well as the judges who scored that fight and their particular scoring
of each round. Everything should seem failry straight forward, with some minor points.

1) I have thre levels of user
  1:  Somone viewing the site with no login
  2: A user who has logged in. This allows them to alter most data on the site. I have limited each user to a unique email
  3: An Admin account, as shown in my seeds file. I have in my seed file an admin account. The primary difference is that 
    an admin can delete a judge as well as view all user accounts.
The seed file contains two users. 
admin with a password of test
billy with a password of test

2) Fighters cannot be deleted through Web Application. I originally included the functionality, but it just didn't make sense.
If you delete a fighter, than you will have a hole for one of the fighters in the fight. If you had the wrong fighter in a 
fight, you can just edit the fight. So deleting a fighter really didn't make much sense and cause more chaos than required
because of it's dependencies. If a fighter needs to be deleted, it would be done by the administrator on the backend. Deleting
fighters doesn't seem like a very common task.

3) This is running off my home server I built from a raspberry pi running Raspbian (Version of Debian). Currently, it's using
the basic rails server. The rails server in combination with cpu on the raspbian has some lag. Additionally, I am using simple
Ip forwarding from my router. To configure the domain name to the external ip of my computer, I used a free and easy http://www.noip.com/
My next step is to deploy with Apache and Passenger, and then configure without no-ip. Then when I ran rails serve, I had to bind to
the local ip given by my router. If for whatever reason, you cannot hit the url, please email me at [email protected]. While I 
haven't had an issue yet, because I am running from my home, there is no redundancy if my internet goes down.

4) A scorecard is automatically created when a fight is created. A scorecard can only be deleted through deleting a fight. There
is a clear capability that will wipe out all the data on scorecard, but it will not delete it. This is for multiple reasons. One,
it makes sense. Even if you don't have the data for a scorecard, it still exists with every fight. Additionally, it allows much cleaner
method calling thoughout the views because it will provide a null for missing data in the scorecard, but not a null for the scorecard
itself. This dramatically reduces error checking throughout the views and provides much cleaner overall code.

Instructions:
Github Repo: https://github.com/wfortin136/final.git
URl: wjf13.ddns.net


Rubric: 
*****Domain Modeling:
- At least 5 database-backed models, which must consist of at least 4 domain models, plus one User model
Models:
- User

- Weightclass
- Fight
- Fighter
- Judge
- Fightcard
- Scorecard
- Judgescore

- A least 1 many-to-many model association
many-to-many
- Fightcard to Weightclass through fight
- To test
- :run rails console
- :Fightcard.first.weightclasses
- :wc = Weightclass.find_by(:name => "Bantomweight")
- :wc.fightcards

Proper association methods
- The models has many asscoiations. This was actually the most difficult part of my web app. Because there are so many
dependencies, many that are non standard such as requiring fighter_1_id and fighter_2_id from a scorecard, many of the
standard associations did not fit very nicely. I think the best example is getting the fights from each fighter. This is
shown in the Fighter model. In particular, I had to use sql querry string to match fighter_1 OR fighter_2 in the scorecard.
Then I needed a join from Fight to scorecard using the scorecard ids from the previous sql querry. This join leveraged a
previous association between the two 

Model validations as necessary
-Validated input fileds for primary pieces such as fights had to have 2 fighters. A user required a unique email address.
A fightcard and a fighter required a name. Basic things like that.

A proper database seeds file
The seeds file contains just enough fights, fight cards, fighters, and judges to test all the functionality. Then next step
would be finishing the web scraping the fully populating the database with automated seeds file.

********Controllers and Views
At least one complete, RESTful database-backed resource
- Fights
- Fightcards
- Judge
- Scorecard

Common site navigation throughout
- I tried to be consistent as possible. This is primarily done through the navbar that allows consistent navigation. But it
also shows through the embedded targets throughout the application.
Secure password-based signup/sign-in/sign out functionality
- This is shown through either the 2 users in the seeds file, or you can create a new standard user through the sign up functionality.

Feature Choice: 
- Deployed from my homeserver with the URL: wjf136.ddns.net

Best Practices:

-------------
To Do:
- Fix an IPv6 error I was getting from the router. Currently, I disabled IPv6 in order to properly deploy.
- Add input for user scoring of live events

- Finish writing ruby script that web scrapes Fight metric
1) Get UFC fight cards
2) Get fights from each card
3) Get fighters at each card
4) Get round by round stats from each fighter in each fight

- Web scrape http://mmadecisions.com/
1) Obtain judge round by round scores
2) Map scorecards to fights
---
Environment:

- Ruby version
1) ruby 2.2.2p95
2) rails 4.1.8

- System dependencies

* Configuration
1) bundle install

* Database creation
1) rake db:migrate

* Database initialization
1) rake db:seed

* Deployment instructions
1) rails server :b <internal ip-address>
2) or rails server (if running locally for testing)
I also had additional dependencies with my unique system. Specifically, I had to configure the ports at my router, I had to disable IPv6
on the specific ip address, and I required a couple additional gems in order to run rails on Debian. I needed execjs and nodejs, from 
what I gathered, for the javascript dependencies. Either way, these are only on the deployed version of my app. Within the repo, I have only
the gem files required to run on Mac Yosemite.