Echo test mocks #708
Echo test mocks #708
Conversation
Codecov Report
@@ Coverage Diff @@
## dev #708 +/- ##
======================================
Coverage ? 96.63%
======================================
Files ? 33
Lines ? 1160
Branches ? 0
======================================
Hits ? 1121
Misses ? 39
Partials ? 0
Continue to review full report at Codecov.
|
test/echo.js
Outdated
| t.is(stdout, '555\n'); | ||
| t.end(); | ||
| }); | ||
| test.skip('simple test with silent(true)', t => { |
There was a problem hiding this comment.
Any reason why we're skipping this?
There was a problem hiding this comment.
I think because this is testing with config.silent = true;, but we already have set shell.config.silent = true above on line 7, so this test is redundant.
There was a problem hiding this comment.
In that case, why not delete it?
There was a problem hiding this comment.
Sure, I will delete this test.
test/utils/mocks.js
Outdated
| @@ -0,0 +1,73 @@ | |||
| /* eslint-disable prefer-rest-params, strict */ | |||
test/utils/mocks.js
Outdated
|
|
||
| function init() { | ||
| resetValues(); | ||
| console.log = consoleLog; |
There was a problem hiding this comment.
Do we need to mock console.log and console.error? What happens if we just mock process.stdout & .stderr objects? That might give us more reliable results overall.
There was a problem hiding this comment.
Right, that's probably a better thing to do in this scenario. I actually just copied this over from the shx mocks. I'll reduce this down to just the process.stdout and process.stderr.
There was a problem hiding this comment.
@freitagbr can you manually confirm that ava still outputs the result of the test (pass/fail)? I want to make sure the mocks don't interfere with test reporting.
test/utils/mocks.js
Outdated
| process.exit = oldProcessExit; | ||
| } | ||
| exports.restore = restore; | ||
| function Mocks() { |
There was a problem hiding this comment.
Any reason we're using mocks as a prototype? I like the idea in general, but we're mocking global values. Using prototypes gives the illusion that multiple mocks can be created, which wouldn't work very well (the second mock would be mocking the first one).
There was a problem hiding this comment.
This is a good point. I implemented it as a prototype because I though it was a good way to implement this. But, if you have several mocks going at once, the true process.stdout.write will get swapped around in them, and it may be hard to figure out where it went, unless you restore in the right order. This seems like it could be problematic, so I will revert this to the global version.
There was a problem hiding this comment.
Might be good to enforce that init() can only be called once though (since calling it twice will lose the real process.stdout). Maybe we can just implement this as 2 streams for process.stdout and process.stdin, then we don't have to reinvent the stream logic.
There was a problem hiding this comment.
I've re-implemented the mocks using globals, and calling init() or restore() any number of times will not lose the real process.stdout.write, nor the mock functions
|
Yes, the tests still output during these tests, even with the mocks. In fact, that spinny thing is still shown, so I don't think the mocks interfere with anything. |
|
Also, the reason I merged this into dev instead of master is because of the extra unit tests added in the newline echo branch. I figured it would just be easier to refactor all the tests at once, instead of fixing merge conflicts later on. |
* Add stdout/stderr test mocks * Mock stdout/stderr during echo tests * Fix lint issues * Use 'use strict' * Re-implement mocks as a prototype * Implement mocks as a single-instance * Remove redundant test * Create mocked stdout/stderr.write methods once
* Add stdout/stderr test mocks * Mock stdout/stderr during echo tests * Fix lint issues * Use 'use strict' * Re-implement mocks as a prototype * Implement mocks as a single-instance * Remove redundant test * Create mocked stdout/stderr.write methods once
Fixes #622
Adds mocks for
process.stdout.writeandprocess.stderr.writethat allows the written values to be extracted. This way, we can inspect the output ofechowithout actually outputting to stdout.