-
-
Notifications
You must be signed in to change notification settings - Fork 1.2k
Testing
When you do integration testing within your own apps you'll eventually need to close your node-postgres clients otherwise the test process will never exit. There are various strategies I've employed throughout the node-postgres test code to accomplish this. While they all work they are not all very pretty due to having to bootstrap the testing & disconnecting of node-postgres while testing it. (chicken v. egg) So here's what I recommend.
If you're using the pg object to manage a pool of clients you can call pg.end()
whenever you're finished testing. This is currently a brute-force way to shut down your clients and may potentially kill any queries which have not finished running during your tests (note: I plan on looking at better ways to terminate the client pool in the future). Internally it loops through all the clients and basically forces their sockets to disconnect. Your mileage may vary with this approach.
Another option is to set the client pool size to 1. This ensures the pg object will always return the same client object whenever pg.connect()
is called. This way you can do the following in your 'cleanup' or 'teardown' area:
tearDown: function() {
var connectionString = /*whatever connection you use for testing*/
pg.connect(connectionString, function(err, client) { client.end() });
}
If you're manually managing your own Client instances without using pg then you are free to end them whenever you want. Generally I do the following:
var client = new Client(/*connectionInfo*/);
client.query('bla bla');
client.on('drain', client.end.bind(client)); //auto disconnect client after last query ends
Note: as mentioned other places...do not create a new client instance for each http request you receive. You will exhaust available connections to your PostgreSQL server and you will be sorry.
In case you want to contribute....
Unit tests do not depend on having access to a running PostgreSQL server. They run in memory by stubbing out the layer below the object under test. If the Client object is being tests, the lower level Connection object is stubbed out and controlled within the test. This allows to simulate specifically timed error events and more complex interactions between the Client and the Connection such as when executing a Prepared Statement. All unit test files exist under test/unit/
and by convention end with -tests.js
. They can be executed either by make test
, make test-unit
, or npm test pg
The integration tests operate on an actual PostgreSQL server instance. You can supply credentials via a quick and dirty command line option parser which is executed before any test is run. Also, the integration tests require populating your database with some test data. The script is located at node-postgres/script/create-test-tables.js
. All integration tests files exist under test/integration/
and by convention end with -tests.js
. Further instructions below...
You can run any test file directly by doing the node test/unit/connection/inbound-parser-tests.js
or something of the like.
However, you can specify command line arguments after the file and they will be picked up and used in the tests. None of the arguments are used in unit tests since they have no external dependencies, so you're safe to just blast away with the command like above, but if you'd like to execute an integration test, you outta specifiy your database, user to use for testing, and optionally a password.
To do so you would do something like so:
node test/integration/client/simple-query-tests.js pg://user:password@host:port/databaseName
If you'd like to execute all the unit or integration tests at one time, you can do so with the "Makefile"
the make file is used as a shortcut to running all the tests. Since many of the integration tests slam the database with load/connection-pool based testing the tests are executed sequentially file-by-file. This ensures each file finishes before the next begins and reduces test complexity and raises test isolation. The command line parameters passed to individual test files can be passed to the Makefile which will in turn pass them along to each test file.
make test-unit
or optionally, since test-unit
is the default
note this means integration tests under test/integration
do not run with from default makefile task
make test
make test-integration connectionString=pg://user:password@host:port/databaseName
make test-all connectionString=pg://user:password@host:port/databaseName
In short, I tried to make executing the tests as easy as possible. Hopefully this will encourage you to fork, hack, and do whatever you please as you've got a nice, big safety net under you.
In order for the integration tests to not take ages to run, I've pulled out the script used to generate test data. This way you can generate a "test" database once and don't have to up/down the tables every time an integration test runs. To run the generation script, execute the script with the same command ine arguments passed to any other test script.
node script/create-test-tables.js pg://user:password@host:port/databaseName