/ 
Spek-0.0.10
River stage zero No dependents
/ Spek

NAME

Spek

SYNOPSIS

Test centric web framework

INSTALL

cpanm Spek

USAGE

declare http endpoints

To declare http endpoint you should write swat tests:

GET /users

To get users list ...

$ mkdir users
$ touch users/get.txt

POST /user

To create a new user ...

$ mkdir -p user
$ touch user/post.txt

GET /user

To get user info ...

$ mkdir -p user/id
$ touch user/id/get.txt

Follow swat for full explanation of swat test harness.

run spek application

$ spek

... you should see:

HTTP::Server::PSGI: Accepting connections at http://0:5000/

run swat tests

# in parallel console

$ echo 127.0.0.1:5000 > host

$ swat

... you should get an errors like <HTTP/1.0 404 Not Found>
... as endpoints are not *defined* yet

127.0.0.1 - - [16/Mar/2016:09:00:03 +0300] "GET /users HTTP/1.1" 404 20 "-" "curl/7.47.0-DEV"
127.0.0.1 - - [16/Mar/2016:09:00:03 +0300] "GET /user/foo HTTP/1.1" 404 20 "-" "curl/7.47.0-DEV"
127.0.0.1 - - [16/Mar/2016:09:00:03 +0300] "POST /user HTTP/1.1" 404 20 "-" "curl/7.47.0-DEV"

implement endpoints

initialize application

app.pm :

our $USERS = {
    'alex' => {
        'age' => 39,
        'email' => 'melezhik@gmail.com'
    },
    'bot' => {
        'age' => 1000,
        'email' => 'iamarobot@...'
    },

};

populate controllers

It should be Kelp routes:

GET users/

users/get.pm :

my ( $self ) = @_;
my $list;
for my $id ( sort keys %$USERS ) {
    $list.="$id: $USERS->{$id}->{email}\n";
}

$list;

GET user/id

user/id/get.pm :

my ($self, $id ) = @_;
"id: $id email: ".($USERS->{$id}->{email});

POST user

user/post.pm

my ( $self ) = @_;
$USERS->{$self->param('name')} = {
    email => $self->param('email'),
    age => $self->param('age'),
};

"user ".($self->param('name'))." created OK"

refine your tests

Tests should be swat modules:

$ echo swat_module=1 > users/swat.ini
$ echo swat_module=1 > user/id/swat.ini
$ echo swat_module=1 > user/swat.ini

Some tests should pass http request parameters:

POST /user

$ echo 'curl_params="-d name=foo -d age=30 -d email='foo@bar'"' >> user/swat.ini

Define expected response:

POST /user :

$ echo 'user foo created OK' > user/post.txt

GET /users :

$ cat users/get.txt

alex: melezhik@gmail.com
bot: iamarobot@...
foo: foo@bar

GET /user/id

$ cat user/id/get.txt

id: foo email: foo@bar

And finally create a meta story:

$ mkdir crud
$ cat crud/meta.txt

application should be able
to perform CRUD operations

$ cat crud/hook.pm

run_swat_module( POST => '/user' );
run_swat_module( GET => '/user/id', { id => 'foo' }  );
run_swat_module( GET => '/users'  );

Rebuild spek app

In console running spek app:

$ <CTRL> + <C>
$ spek

... should see:


reiniting spek app ...
populate app.pm ...
populate post /home/vagrant/my/spek-example-app/user ...
populate get /home/vagrant/my/spek-example-app/user/id ...
populate get /home/vagrant/my/spek-example-app/users ...
HTTP::Server::PSGI: Accepting connections at http://0:5000/

Run tests

Now you have a specifi(K)ation simply just running:

$ swat

... should see:

/home/vagrant/.swat/.cache/5738/prove/crud/META/request.t ..
# @META
#        application should be able
#        to perform CRUD operations
#
ok 1 - 200 / 1 of 2 curl -X POST -k --connect-timeout 20 -m 20 -d name=foo -d age=30 -d email='foo@bar' -L -D - '127.0.0.1:5000/user'
ok 2 - output match 'user foo created OK'
ok 3 - 200 / 1 of 2 curl -X GET -k --connect-timeout 20 -m 20 -L -D - '127.0.0.1:5000/user/foo'
ok 4 - output match 'id: foo email: foo@bar'
ok 5 - 200 / 1 of 2 curl -X GET -k --connect-timeout 20 -m 20 -L -D - '127.0.0.1:5000/users'
ok 6 - output match 'alex: melezhik@gmail.com'
ok 7 - output match 'bot: iamarobot@...'
ok 8 - output match 'foo: foo@bar'
1..8
ok
All tests successful.
Files=1, Tests=8,  0 wallclock secs ( 0.02 usr  0.00 sys +  0.06 cusr  0.00 csys =  0.08 CPU)
Result: PASS

Application introspection

routes

To inspect all available routes:

$ spek --routes

meta stories

To inspect meta stories:

$ spek --meta

Read swat documentation to know more about swat meta stories.

Advanced usage

handling json responses

Modify Kelp controller

GET /user/id

$ cat user/id/get.pm

my ($self, $id ) = @_;
{
    id => $id,
    email => $USERS->{$id}->{email} ,
    age => $USERS->{$id}->{age}
}

Add json handler to swat test:

$ cat user/id/get.handler

my $headers   = shift;
my $body      = shift;
use JSON;
$hash = decode_json($body);
return "$headers\n".("id: $hash->{id}\nemail: $hash->{email}\nage: $hash->{age}");

Regenerate spek application

$ <CTRL> + <C> # to stop running application

$ spek

... should see:

reiniting spek app ...
populate app.pm ...
populate post /home/vagrant/my/spek-example-app/user ...
populate delete /home/vagrant/my/spek-example-app/user/id ...
populate get /home/vagrant/my/spek-example-app/user/id ...
populate get /home/vagrant/my/spek-example-app/users ...
generate hook for /user/id ...
inject response handler for get ...

Run swat tests

$ swat

... should see:

ok 7 - 200 / 1 of 1 curl -X GET -k --connect-timeout 20 -m 20 -L -D - '127.0.0.1:5000/user/foo'
ok 8 - output match 'Content-Type: application/json'
ok 9 - output match 'id: foo'
ok 10 - output match 'email: foo@bar'

Read more swat docs on how to process http responses.

To add handlers for other http methods simply create a proper handler file:

DELETE /user/id

$ nano user/id/delete.handler # and so on ...

Author

Alexey Melezhik

Download example spek application

You can upload source code here - https://github.com/melezhik/spek-example-app

See also