isaacgr / jaysonic / 190

95%

push

travis-ci-com

Improvement/#65/improve tests (#85)

* Correctly use before/after hooks within describe blocks
* Open and close server connections before each test block to remove need for unique ports
* Remove unecessary batchResponse listener tests
* Remove v1 server from client tests (since client version determines server response if server is v2)
* Resolve http status code with response/error.
* Add method to remove connected clients when server is closed.
* Overwrite _removeClients for ws server.
    - uses .close() instead of .destroy()
* Add 'version' parameter to _raiseError to format error response correctly.
* Assume version 2.0 response if the version passed to formatError isnt explicitly 1.
* open and close server/client connections after each test block
* use the same port and methods for all servers
* use imported servers and clients from test-client and test-server files
* had to comment out #54 request timeout test for http due to
socket hang up error
* Add test for get.message()
* Return 1.0 error response if 2.0 message request sent to 1.0 server.
* Add test for listening multiple times to ws server
* Add tests to client for connecting multiple times
* Add test for reconnect and client closing connection
* Add reconnect test for base client protocol
* Rename node ws client test
* Remove unecessary try/catch for ws server listen
* Add test for https server creation
* Add tests for verifyData in base client class
* Remove extra delimiter from server notifications
* Add chai spies to dev deps
* Add test for serverDisconnect method for client
* Add tests for ws-client subscriptions
* Test _getAllFuncs()
* Test rejecting .close() if error
* Comment out http request timeout test since it doesnt work for some reason
* Dont JSON format errors from http client notifications
* Add test for gotError in base server protocol.
* Add delimiter before calling writeToClient in gotError.
* Add tests for base client protocol error h... (continued)

229 of 250 branches covered (91.6%)

Branch coverage included in aggregate %.

680 of 707 relevant lines covered (96.18%)

139.7 hits per line