A simple CLI app that retrieves book information details from Google Books API. The app is able to filter results by:
- author name
- book title
- free form text that can match any book detail
(any combination of these factors is allowed)
A session with the app can look like this:
$ ./books -a Tolkien
1) Title: OS FILHOS DE HURIN
Date:
Authors: J. R. R. TOLKIEN, ALAN LEE, CHRISTOPHER TOLKIEN, LUZIA APARECIDA DOS SANTOS, RONALD EDUARD KYRMSE
2) Title: O Silmarillion
Date: 2015-08-11
Authors: J.R.R. Tolkien
3) Title: O Hobbit
Date:
Authors: J.R.R. Tolkien
4) Title: Senhor Dos Aneis, O - as Duas Torres - Vol 2
Date: 2000
Authors: John Ronald Reuel Tolkien
5) Title: THE FELLOWSHIP OF THE RING
Date:
Authors: J.R.R. TOLKIEN
...
To build the app, be sure to have rebar 2 laying around (I have it on my home /bin folder which is on my path).
NOTE: Yes, I know rebar 2 is (sort of) deprecated and people should be using rebar 3 but since I'll be maintaining a few rebar 2 apps, I used this opportunity to learn a bit about it.
After making sure you have rebar laying around on your path, you need to install all dependencies:
$ rebar get-deps
==> books (get-deps)
Pulling jsone from {git,"[email protected]:sile/jsone.git"}
Cloning into 'jsone'...
Pulling dactyl from {git,"[email protected]:basho/dactyl.git"}
Cloning into 'dactyl'...
Pulling hackney from {git,"[email protected]:benoitc/hackney.git"}
Cloning into 'hackney'...
...
This will download all dependencies and make them available to compile. Then all that is left is for you to compile and build the app:
$ rebar compile escriptize
==> jsone (compile)
Compiled src/jsone.erl
Compiled src/jsone_decode.erl
Compiled src/jsone_encode.erl
==> dactyl (compile)
Compiled src/dactyl.erl
...
==> books (escriptize)
This will compile all modules and dependencies and will also generate an binary escript that basically is the runnable app.
All you need to do now is to use the app!
$ ./books
Usage: "./books" [options] [search term]
Options:
-a Author
-t Title
-h shows this help screen
Search Term is any string you want to search for
Examples:
$ "./books" -a Tolkien # search for books written by Tolkien
$ "./books" -t Hobbit # search for books entitled Hobbit
$ "./books" Hitchhiker Guide # search for books that include the words 'Hitchhiker' or 'Guide'
$ "./books" -a 'Douglas Adams' Hitchhiker # search for books written by Douglas Adams and that include the word 'Hitchhiker'
For the test environment, there is a rebar.config.script file that will inject a few config details.
This works by setting the ENV variable RUNENV=test.
All tasks should be executed through make (and using the provided Makefile) as it accounts for the RUNENV detail.
The very first thing to do before running the tests is to install test dependencies. To do so, run:
$ make test-get-deps
This will install all regular dependencies and also the test ones.
To run the EUnit tests simply run:
$ make eunit
this is just a shortcut for RUNENV=test rebar eunit.
There are 2 ways to run the tests: With code coverage report or without it. Code coverage is not the best metric ever but can be helpfull when you are refactoring tests or covering non-tested libs/apps with tests.
So, to run tests with coverage reports simply:
$ make ct-cover
to run without coverage report (which is also way faster):
$ make ct
You can (and should) also run all tests at once. To do so, run:
$ make test-all
This will run both ct tests and eunit tests.
This is my very first CLI app in Erlang, I am in no manner an Erlang expert or something like that. This is meant to be a learning exercice and a reference (either good or bad) for future work.