Truncate long source context lines #329
Conversation
test/raven.utils.js
Outdated
| frame.context_line.trim().should.startWith('{snip}'); | ||
| frame.context_line.trim().should.endWith('{snip}'); | ||
| frame.post_context.should.be.an.instanceOf(Array); | ||
| frame.post_context[0].trim().should.endWith('{snip}'); |
There was a problem hiding this comment.
I feel it's not sufficient to just test for presence of {snip} – can we improve this test to also assert the content of the snipped code is as expected (e.g. what appears between {snip})?
LewisJEllis
force-pushed
the
truncate-long-stuff
branch
from
971089d to
24f6ec5
Compare
|
@benvinegar updated the test. I just check the length and the presence of the |
|
It's an improvement.
Yes. |
|
Has this been released yet? Just signed up for sentry (looks awesome) and this was unfortunately the first thing I hit. We deploy an isomorphic react app that leverages webpack asset url hashing so we webpack the server code as well. |
|
Not yet. We'll publish something soon. |
|
@benvinegar any progress / timeline on this? |
|
@joevbruno – I wanted to do this earlier in the week, but now that it's Friday I'd rather wait until the weekend is over first. There's a bunch of stuff on master that could potentially be problematic (it's well tested, but you never know), and even this truncation change could cause some issues to group differently. We want to be on-hand just in case. |
|
@benvinegar Understood. Hopefully it will happen sometime this week |
|
this has been released @joevbruno let me know if you have any problems with it! |
This should address most of the cases causing #153.
/cc @mattrobenolt @MaxBittker