fix: make a single webpack bundle #203
Conversation
Codecov Report
@@ Coverage Diff @@
## master #203 +/- ##
=========================================
+ Coverage 78.16% 78.47% +0.3%
=========================================
Files 6 6
Lines 142 144 +2
=========================================
+ Hits 111 113 +2
Misses 31 31
Continue to review full report at Codecov.
|
|
@vmx will .spec continue to work? |
|
@diasdavid Yes. This change should be backwards compatible. It will run all |
|
I've tested it locally with js-ipfs-api, js-ipfs and js-ipld-zcash. |
|
@vmx I have a suggestion - we can test this on CI's as well by using this branch as a git url in the package.json of a project. I think it would be a good way of making sure it works on CI without an actual release. js-ipfs-api should be a good candidate as it's the one that fails the most often. |
src/test/browser-config.js
Outdated
| // that's a single entry point, but it could also be a single test that | ||
| // is given as command line parameter | ||
| const preprocessors = getPatterns(ctx).reduce((acc, pattern) => { | ||
| acc[pattern] = ['webpack'] |
There was a problem hiding this comment.
missing 'sourcemap' preprocessor
Prior to this commit there was a webpack bundle created for every test file. That lead to very high memory usage when running the Browser tests. Now there's a single bundle.
vmx
force-pushed
the
karma-webpack-bundle-alt2
branch
from
9a10fb6
to
f06086c
Compare
|
Thanks @dignifiedquire, I (force) pushed a new version. @dryajov Good suggestion, I create PR on js_ipfs-api using this PR: ipfs-inactive/js-ipfs-http-client#705. |
|
As it got approved by @diasdavid and I fixed what @dignifiedquire suggested I dare to merge it. If it really turns out breaking things, we can always revert it. |
|
@vmx you have publish permissions. Go for it :) |
Prior to this commit there was a webpack bundle created for every test
file. That lead to very high memory usage when running the Browser tests.
Now there's a single bundle.