Logs parser

Tests execution status:

Task:

Logs parser (collect MOST PAGE VIEWS and MOST UNIQ PAGE VIEWS metrics)

Additional work:

Added Invalid Logs section Added Uniq visitors addresses section

Run:

Execute:

bundler
chmod +x ./parser
bundle exec ./parser ./webserver.log

Note: provided sample file doesn't contain any valid IP address, so to provide expected output, IP addres validation schould be disabled (see this file)

bundle exec rspec

Test Coverage - 100%

Script output

======MOST PAGE VIEWS======
0) /about/2 -- 89 visits
1) /contact -- 89 visits
2) /index -- 82 visits
3) /about -- 81 visits
4) /help_page/1 -- 79 visits
5) /home -- 78 visits
===============================
======MOST UNIQ PAGE VIEWS======
0) /index -- 23 visits
1) /home -- 23 visits
2) /contact -- 23 visits
3) /help_page/1 -- 23 visits
4) /about/2 -- 22 visits
5) /about -- 21 visits
===============================
======UNIQ VISITORS ADDRS======
0) /about/2 -- ["444.701.448.104", "184.123.665.067", "382.335.626.855", "543.910.244.929", "555.576.836.194", "802.683.925.780", "200.017.277.774", "126.318.035.038", "451.106.204.921", "235.313.352.950", "836.973.694.403", "217.511.476.080", "316.433.849.805", "061.945.150.735", "715.156.286.412", "646.865.545.408", "016.464.657.359", "897.280.786.156", "682.704.613.213", "722.247.931.582", "158.577.775.616", "336.284.013.698"] visits
(Actuall output in this section is shorten...)
===============================
======INVALID LOGS======
0) {:addr=>nil, :path=>"/help_page/1", :errors=>["Line format isn't valid!"]}
1) {:addr=>nil, :path=>"836.973.694.403", :errors=>["Line format isn't valid!"]}
(Broke some examples in log file for demonstration purpose)
===============================

Approach description:

Dependency schema:

.
├── adapters
│   └── file_adapter.rb
├── entries
│   ├── line.rb
│   └── statistic.rb
├── parser.rb
└── printers
    ├── artifacts_printer.rb
    ├── base_printer.rb
    └── visits_printer.rb

• Adapters - the goal of adapter is to provide data sources abstraction (so in future we will not be restricted to only File source, and we also will have possibility to scale horizontally with different adapters (e.g. DB, HTTP, etc))
• Entries - entries encapsulates behavior related to data source (validation, rules of calculation and exposing data in unified format to outer world)
• Printers - their job are actually print, format and processing related logic (sorting, ...)

All dependency injections are placed in the constructors to have possibility to test modules independently with mock's.

Possible Improvements (ideas)

• Move validations in separate namespace (for now validations are implemented in business logic)
• Implement less coupled Entries (now Statistic is highly depends on Line)