diff options
Diffstat (limited to 'docs/1_Write_the_tests.md')
| -rw-r--r-- | docs/1_Write_the_tests.md | 246 |
1 files changed, 246 insertions, 0 deletions
diff --git a/docs/1_Write_the_tests.md b/docs/1_Write_the_tests.md new file mode 100644 index 0000000..500af54 --- /dev/null +++ b/docs/1_Write_the_tests.md @@ -0,0 +1,246 @@ +# Write the tests + +## Create the test tree + +At the root of your project, create a test dedicated directory that will hold +all your tests materials. A classic test tree looks like the following: + +```tree + test + ├── CMakeLists.txt + ├── etc + │ ├── CMakeLists.txt + │ └── aft-middlename.json + ├── fixtures + │ ├── CMakeLists.txt + │ ├── helper.sh + │ ├── data + │ └── plugin.lua + └── tests + ├── CMakeLists.txt + ├── test01.lua + ├── test02.lua + └── test03.lua + ... +``` + +Here is a description of each subdirectory purpose: + +- etc: holds the test binding configuration in a JSON file. +- fixtures: contains all external needed files to run your tests. This is mainly + used to inject data or a plugin with the mock-up APIs code in a LUA or C plugin +- tests: Contains only the tests written in LUA for your binding. + +## Create your configuration file + +The configuration file describes your test api and how it launches the tests. A +test binding doesn't provide verbs and this configuration will describe the api +verb(s), mocked-up apis. Here are the `key` descriptions for this file: + +### `metadata` section + +- `uid`: A simple label mainly useful at debugging purpose +- `version` (optional): the test's version +- `info` (optional): self-explanatory +- `api`: the name that your test binding will take. Used to be the test api's + name prefixed with `aft` (ie: `aft-<tested-api-name>`) +- `require`: the tested api's name. This key ensure that the tested api is + correctly launched and initialized so the test binding could test it. + +### `testVerb` section + +- `uid`: the verb name +- `info` (optional): self-explanatory +- `action`: special string indicating which function to trigger when the verb is + called. It will always be the same about the test binding. +- `args` section: + - `trace`: the name of the tested api that you are going to test. This is + needed to let the test binding tracing the tested api and retrieve through + the binder monitoring feature `calls` and `events`. + - `files`: A string or an array of strings of the LUA files with your tests. + Only provide the file name without the path. + +### `mapis`, stand for Mock-up api, section + +- `uid`: the mocked up api's name +- `info` (optional): self explanatory +- `libs`: LUA script or C plugin file name + +#### `verbs` section + +Describe the implemented mocked up API verbs. Each verb is mapped to a function +name that will be executed at the verb call. + +- `uid`: verb's name +- `info` (optional): self explanatory +- `action`: an URI string that point to a C plugin or LUA script's function that + will be executed at the verb call. The format of the action URI is: + `<lua|plugin|api>`://`<C plugin's name|api's name|lua script name>`#`<function|verb's name>` + +#### `events` section + +This section allows you to trigger a function when a described event is received +. It could be for any event which you need to apply modifications on it. You DO +NOT have to enumerate each possible events that the mocked up api would +generate, you could avoid this section if you do not want to execute a function +at events reception. + +- `uid`: event's name (`<api>/<event-name>`) +- `info` (optional): self explanatory +- `action`: an URI string that point to a C plugin or LUA script's function that + will be executed at the event reception. The format of the action URI is: + `<lua|plugin|api>`://`<C plugin's name|api's name|lua script name>`#`<function|verb's name>` + the action `lua://AFT#_evt_catcher_` is the default `afb-test` events listener. + +### Configuration examples + +Here is a simple example: + +```json +{ + "id": "http://iot.bzh/download/public/schema/json/ctl-schema.json#", + "$schema": "http://iot.bzh/download/public/schema/json/ctl-schema.json#", + "metadata": { + "uid": "Hello_Test", + "version": "1.0", + "api": "aft-example", + "info": "Binding made to test other bindings", + "require": [ + "hello" + ] + }, + "testVerb": { + "uid": "testing-hello", + "info": "Launch the tests against hello api", + "action": "lua://AFT#_launch_test", + "args": { + "trace": "hello", + "files": ["aftTest.lua","helloworld.lua"] + } + } +} +``` + +and another example that mock-up the `low-can` api: + +```json +{ + "id": "http://iot.bzh/download/public/schema/json/ctl-schema.json#", + "$schema": "http://iot.bzh/download/public/schema/json/ctl-schema.json#", + "metadata": { + "uid": "Other_Tests", + "version": "1.0", + "api": "aft-example", + "info": "Binding made to test other bindings", + "require": [ + "tested-api" + ] + }, + "testVerb": { + "uid": "Complete", + "info": "Launch all the tests", + "action": "lua://AFT#_launch_test", + "args": { + "trace": "low-can", + "files": [ "aftTest.lua", "mapis-tests.lua" ] + } + }, + "mapis": [{ + "uid": "low-can", + "info": "Faked low-can API", + "libs": "mapi_low-can.lua", + "verbs": [ + { + "uid": "subscribe", + "info": "Subscribe to CAN signals events", + "action": "lua://low-can#_subscribe" + }, + {...}, + { + "uid": "write", + "info": "Write a CAN messages to the CAN bus.", + "action": "lua://low-can#_write" + } + ], + "events": [{ + "uid": "low-can/diagnostic_messages", + "action": "lua://AFT#_evt_catcher_" + },{ + "uid": "low-can/messages_engine_speed", + "action": "lua://AFT#_evt_catcher_" + },{ + "uid": "low-can/messages_vehicle_speed", + "action": "lua://AFT#_evt_catcher_" + }] + }] +} +``` + +## The LUA test files + +The test framework uses the LUA language to describe the tests. + +You have two different things to have in mind when you write your tests using +this framework: *test* and *assertions* functions. + +- *Assertions* functions are meant to test an atomic operation result. + (ie: `1+1 = 2` is an assertion) +- *Test* functions represent a test (Unbelievable), they represent a set of one + or several *assertions* which are all needed to succeed to valid the test. + +The framework comes with several *test* and *assertion* functions to simply be +able to test verb calls and events receiving. Use the *test* ones and if you +need more use the ones that call a callback. If the test is more complex or +more comprehensive then *describe* your test function using *assert* functions. +See the example below. + +See the test framework functions [References](Reference/Introduction.md) for a +comprehensive list of *tests* and *assertions* functions available. + +### Tests example + +```lua + function _callback(responseJ) + _AFT.assertStrContains(responseJ.response, "Some String") + end + + function _callbackError(responseJ) + _AFT.assertStrContains(responseJ.request.info, "Ping Binder Daemon fails") + end + + function _callbackEvent(eventName, eventData) + _AFT.assertEquals(eventData, {data = { key = 'weird others data', another_key = 123.456 }}) + end + + _AFT.addEventToMonitor("hello/anEvent") + _AFT.addEventToMonitor("hello/anotherEvent", _callbackEvent) + + _AFT.testVerbStatusSuccess('testPingSuccess','hello', 'ping', {}) + _AFT.testVerbResponseEquals('testPingSuccess','hello', 'ping', {}, "Some String") + _AFT.testVerbResponseEquals('testPingSuccess','hello', 'ping', {}, "Unexpected String") + _AFT.testVerbCb('testPingSuccess','hello', 'ping', {}, _callback) + _AFT.testVerbStatusError('testPingError', 'hello', 'pingfail', {}) + _AFT.testVerbResponseEqualsError('testPingError', 'hello', 'pingfail', {}, "Ping Binder Daemon fails") + _AFT.testVerbResponseEqualsError('testPingError', 'hello', 'pingfail', {}, "Ping Binder Daemon succeed") + _AFT.testVerbCbError('testPingError', 'hello', 'pingfail', {}, _callbackError) + + _AFT.testVerbStatusSuccess('testEventAdd', 'hello', 'eventadd', {tag = 'event', name = 'anEvent'}) + _AFT.testVerbStatusSuccess('testEventSub', 'hello', 'eventsub', {tag = 'event'}) + _AFT.testVerbStatusSuccess('testEventPush', 'hello', 'eventpush', {tag = 'event', data = { key = 'some data', another_key = 123}}) + + _AFT.testVerbStatusSuccess('testEventAdd', 'hello', 'eventadd', {tag = 'evt', name = 'anotherEvent'}) + _AFT.testVerbStatusSuccess('testEventSub', 'hello', 'eventsub', {tag = 'evt'}) + _AFT.testVerbStatusSuccess('testEventPush', 'hello', 'eventpush', {tag = 'evt', data = { key = 'weird others data', another_key = 123.456}}) + + _AFT.testVerbStatusSuccess('testGenerateWarning', 'hello', 'verbose', {level = 4, message = 'My Warning message!'}) + + _AFT.testEvtGrpReceived("TestEventGroupReceived",{"hello/anEvent","hello/anotherEvent"},300000) + _AFT.testEvtGrpNotReceived("TestEventGroupNotReceived",{"hello/anEvent","hello/anotherEvent"},300000) + + _AFT.testEvtReceived("testEvent", "hello/anEvent",300000) + _AFT.testEvtReceived("testEventCb", "hello/anotherEvent",300000) + + _AFT.describe("myTestLabel", function() + _AFT.assertEquals(false, false) + end) +``` |