Closed edhager closed 7 years ago
This is fix 1 of 2 for dojo/cli-test-intern#17. The second fix is in dojo\cli-test-intern
.
Merging #104 into master will increase coverage by
<.01%
. The diff coverage is100%
.
@@ Coverage Diff @@
## master #104 +/- ##
=========================================
+ Coverage 95.59% 95.6% +<.01%
=========================================
Files 11 11
Lines 613 614 +1
Branches 127 127
=========================================
+ Hits 586 587 +1
Misses 14 14
Partials 13 13
Impacted Files | Coverage Δ | |
---|---|---|
src/main.ts | 95.6% <100%> (+0.04%) |
:arrow_up: |
Continue to review full report at Codecov.
Legend - Click here to learn more
Δ = absolute <relative> (impact)
,ø = not affected
,? = missing data
Powered by Codecov. Last update a5db155...9f58e33. Read the comment docs.
Seeing actually how this works, I realise we are now passing node
from the test command? Does this mean we are testing a bundle in one environment and then shipping it to another? Is this a good idea?
We have gone with not code splitting for the test bundle (which is used for running unit tests) which has been addressed https://github.com/dojo/cli-test-intern/pull/26
The builder is not setting Webpack's
target
option and the default is "web" which builds chunks for the browser environment. Using@dojo/core/load
causes Webpack to inject code into the chunks that assumes the globalwindow
exists. This fails when the chunks are loaded in a node.js environment.This PR adds a
target
option to the builder which maintains the "web" target by default.Fix dojo/cli-test-intern#17