Allow anonymous_id in alias and group#188
Conversation
fdb7569 to
7cf11d0
Compare
|
|
131e281 to
3a1a89e
Compare
3a1a89e to
5f78817
Compare
This comment has been minimized.
This comment has been minimized.
Codecov Report
@@ Coverage Diff @@
## master #188 +/- ##
==========================================
+ Coverage 98.37% 99.04% +0.67%
==========================================
Files 10 11 +1
Lines 430 419 -11
==========================================
- Hits 423 415 -8
+ Misses 7 4 -3
Continue to review full report at Codecov.
|
5f78817 to
153bc46
Compare
Ah, okay. What is the intended usage here? Is this: the same as this? |
8294fa9 to
e2336c2
Compare
|
@f2prateek reminder, there's an open question here 🙂 |
|
@rohitpaulk ah missed this. correct, they should be considered similar. I would prefer if the library provided only one way for this though (rather than offering both)! What's more natural for Ruby developers to use? |
The natural way is to have different top-level keys for each parameter. i.e. rather than: I'm going to merge this PR as it retains backward compatibility. If we agree that |
This was a bug, introduced in #188
This was a bug, introduced in #188
Fixes #181.
This PR refactors the handling of common fields, extracts a
FieldParserthat adheres to the Segment spec.This will result in the following behaviour changes:
aliaswill now acceptanonymous_idgroupwill now acceptanonymous_idpagewill now validate thatnameis non-emptyAs I refactored this to adhere to the Segment spec, I found a couple of deviations from the spec:
optionsis accepted in every call, but not mentioned in the specscreentakes acategoryparameter, which is not mentioned in the specI've retained these for backwards compatibility.