Skip to content

Statistical Tests

adrianb82 edited this page Jul 20, 2016 · 1 revision

STATISTICAL TESTS

In order to test the API it is enough to simply run the same commands that are included in the API together with a set of files that are known to follow / not follow the specification.

To run the tests simply run the following bash scripts: test_server

Do not forget to change the permission of the script previously:

sudo chmod u+x pathToScript/test_server.sh

It contains the following tests:

POST validation

3 examples (valid_observation.json, valid_observation2.json, valid_observation3.json) are availalbe due to the fact that one can upload JSON files that contain observations with different fields. The webLyzard API defines a set of required and optional fields.

The first example (valid_observation.json) only contains required fields: id, uri, added_date (indexing date), date (observation/document date), indicator_id, indicator_name, value).

Some examples of optional fields are presented in the next example (valid_observation2.json): target_country, target_type, target_location, etc.

The last example is somewhat similar to the first example.

All these examples are expected to pass!

invalid POST

2 examples (invalid_observation.json, invalid_observation2.json) are available in order to highlight various types of errors.

In the first example a fictional field (field that does not exists in the required or optional sets of fields described in the webLyzard API) named testerror is defined which will make this test fail.

The second example is expected to fail, as the id of the observation is missing.

PUT test

The current example (update_observation.json) corresponds to an UPDATE statement.

The expected output is a value of 2000 instead of 1000 for the first observation (id=1).

GET test

This test simply returns the data for a single observation.

DELETE test

This test simply deletes the data for a single observation.

Clone this wiki locally