[GH-788] Made the check to determine whether the current web-worker is owned by PapaParse more reliable#937
Conversation
|
Hi @flakey-bit, Thanks for working on this. I see there is an issue on your patch that match our test suite broken. |
ab87cf1 to
f2fe81f
Compare
NB: I found there was an issue with my implementation so I've had another go at fixing it 🤞 - I've force-pushed to my branch, which should have updated the PR. See latest build queued here. |
|
I've added a test case to cover this scenario. Please take a look |
|
Hi @flakey-bit, Test failed on node12 which is unsuported by now. |
57dbf5c to
76ac689
Compare
|
@pokoli rebase is done |
|
@flakey-bit Tests are still broken, could you please have a look? You can run them locally with: |
|
I'll take a look tomorrow. Sorry for taking up your time!
…On Fri, 10 Jun 2022, 22:39 Sergi Almacellas Abellana, < ***@***.***> wrote:
@flakey-bit <https://github.com/flakey-bit> Tests are still broken, could
you please have a look?
You can run them locally with: npm test
—
Reply to this email directly, view it on GitHub
<#937 (comment)>, or
unsubscribe
<https://github.com/notifications/unsubscribe-auth/ABBVTAKGN4AJKUM6CQW4MILVOMLMNANCNFSM5YE26S3A>
.
You are receiving this because you were mentioned.Message ID:
***@***.***>
|
…able library still targets ECMAScript 5
…ker is owned by PapaParse more reliable - this fixes mholt#788 Previously, the assumption was if we're in a web-worker AND the worker was loaded from a blob, the worker *must* be ours.
…d (this fails prior to f2fe81f) Added inside a new describe block for browser-specific tests (looks for existence of `window`). This test is browser-specific due to the way the script is being loaded into a web-worker using importScripts()
76ac689 to
085778e
Compare
|
@pokoli have force-pushed again.
Note I've included an additional to commit to allow the tests to use latest ECMAScript syntax (does not affect the targeting of the main library). |
|
@Silic0nS0ldier can you also take a look at this? The approach will likely have implications for the move to ES6 modules (#875) In particular, see the test case that this PR adds. |
|
@flakey-bit why is required to add the latest ECMAScript syntax? Does the test work without that? |
|
@pokoli the test is currently written using Could rewrite the test not to use such features, but would probably make it harder to understand. |
|
@flakey-bit I've lost quite a bit of context since opening #875, but I don't see any reason for this to cause problems for ES6. Not sure how relevant it will be here, but I did some work to solve the same problem with NodeJS worker threads some time ago. https://github.com/developit/web-worker/blob/9ee077b6c5df0acaf5d83d89e2087bd3a3b82b97/src/node.js#L138-L142 |
|
@pokoli any chance of getting a new release out to NPM? |
|
@flakey-bit I will prefer to wait a little if there is something else to be included on the release. |
|
@pokoli it would be awesome to get a release out if it's not too much trouble I know it's a bit selfish of me, but part of the reason I picked up #788 was to allow us to integrate PapaParse into our application without resorting to patch-package 🙂 |
This fixes #788
Previously, the assumption was if we're in a web-worker AND the worker was loaded from a blob, the worker must be ours.