Add timeout and unit to TimeoutException message

[artem-zinnatullin]

This is a small enhancement to help with crash/log debugging.

Right now often times you get a stacktrace that points only to RxJava:

java.util.concurrent.TimeoutException
        at
io.reactivex.internal.operators.flowable.FlowableTimeoutTimed$TimeoutSubscriber.onTimeout(FlowableTimeoutTimed.java:137)
        at
io.reactivex.internal.operators.flowable.FlowableTimeoutTimed$TimeoutTask.run(FlowableTimeoutTimed.java:169)
        at
io.reactivex.internal.schedulers.ScheduledRunnable.run(ScheduledRunnable.java:66)
        at
io.reactivex.internal.schedulers.ScheduledRunnable.call(ScheduledRunnable.java:57)
        at java.util.concurrent.FutureTask.run(FutureTask.java:266)
        at
java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$201(ScheduledThreadPoolExecutor.java:180)
        at
java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:293)
        at
java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
        at
java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
        at java.lang.Thread.run(Thread.java:748) 

Adding timeout and unit values can help find related user code faster.

I'm not advocating for particular message, it can be something even shorter, like $timeout $unit.

It can also be extracted in a method if you see value in that.

[akarnokd]

I'm for better error messages but I'm not convinced a timeout value is generally unique enough in a project that will allow you to locate the flow. Especially when the timeout value comes from configuration and thus it is not prevalent in the source code.

An alternative would be an overload on the operator with a String argument that would be passed onto the exception constructor, tagging the particular source location.

[codecov]

Codecov Report

Merging #6234 into 2.x will decrease coverage by 0.04%.
The diff coverage is 100%.

@@             Coverage Diff              @@
##                2.x    #6234      +/-   ##
============================================
- Coverage     98.26%   98.21%   -0.05%     
+ Complexity     6201     6200       -1     
============================================
  Files           667      667              
  Lines         44883    44887       +4     
  Branches       6216     6216              
============================================
- Hits          44103    44088      -15     
- Misses          243      261      +18     
- Partials        537      538       +1

Impacted Files	Coverage Δ	Complexity Δ
...ernal/operators/flowable/FlowableTimeoutTimed.java	99.18% <100%> (+0.81%)	3 <0> (ø)	⬇️
...rnal/operators/completable/CompletableTimeout.java	100% <100%> (ø)	2 <0> (ø)	⬇️
...va/io/reactivex/internal/util/ExceptionHelper.java	100% <100%> (ø)	18 <1> (+1)	⬆️
...activex/internal/subscribers/FutureSubscriber.java	100% <100%> (ø)	32 <0> (ø)	⬇️
...o/reactivex/internal/observers/FutureObserver.java	87.69% <100%> (ø)	28 <0> (ø)	⬇️
...ivex/internal/observers/BlockingMultiObserver.java	100% <100%> (ø)	24 <0> (ø)	⬇️
...l/operators/observable/ObservableTimeoutTimed.java	100% <100%> (ø)	3 <0> (ø)	⬇️
...tivex/internal/operators/single/SingleTimeout.java	100% <100%> (+1.72%)	2 <1> (ø)	⬇️
...tivex/internal/observers/FutureSingleObserver.java	94.33% <100%> (-3.78%)	24 <0> (-1)
...io/reactivex/subscribers/SerializedSubscriber.java	94.31% <0%> (-4.55%)	25% <0%> (-1%)
... and 31 more

---

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 1ea1e2a...827d1fd. Read the comment docs.

[artem-zinnatullin]

I like your suggestion, I still however think that adding timeout and value is a good option for now and for the case when user didn't specify a message.

This will be a "free" improvement for existing users who will just receive it with an update of the library and 0 changes in their code.

Especially when the timeout value comes from configuration and thus it is not prevalent in the source code.

Often times these configurations are traceable, thus knowing the values can be helpful enough.

But still, optional message sounds good to me, I can work on it once we decide on the destiny of this PR

[akarnokd]

Okay, let's have a default message for now, but more like along these lines:

The source did not signal an event for $timeout ${unit.toLowerCase()}(s) and has been terminated.

[artem-zinnatullin]

sgtm, implementing

[artem-zinnatullin]

Here and below I'm not sure what's the purpose of mocking, TestSubscriber/TestObserver are capable of checking what's needed in these test cases

[akarnokd]

Probably some v1 hybrid remnant.

[artem-zinnatullin]

I've updated the message and added tests, turns out in some cases we didn't check for TimeoutException, now these cases are covered

@vanniktech please take a look again

[vanniktech]

Nice even getting rid of some mockito nonsense.