Fix bug with observeAndSample function from Histogram

[cdepillabout]

The Histogram metric has a function observeAndSample, which lets you add a new observation, and sample it at the same time. Before this commit, this observeAndSample function would return the previous sample, not the sample you just added.

This commit changes the behavior of observeAndSample so that it returns the sample that was just added. This matches the behavior of addAndSample (from Counter), and modifyAndSample (from Gauge), which both return the value of the sample that was just added.

Here's an example program that demonstrates this problem:

{-# LANGUAGE OverloadedStrings #-}

module Main where

import Control.Monad.IO.Class (liftIO)
import System.Metrics.Prometheus.Concurrent.RegistryT
import System.Metrics.Prometheus.Metric.Counter (addAndSample)
import System.Metrics.Prometheus.Metric.Histogram (observeAndSample)
import System.Metrics.Prometheus.MetricId


main :: IO ()
main = runRegistryT $ do
    exampleCounter <- registerCounter "example_counter" mempty
    exampleHist <- registerHistogram "example_hist" mempty [10, 20 .. 100]

    liftIO $ do
      counterResult <- addAndSample 29 exampleCounter
      histResult <- observeAndSample 59 exampleHist
      print counterResult
      print histResult

Running this on master gives output like the following:

CounterSample {unCounterSample = 29}
HistogramSample {histBuckets = fromList [(10.0,0.0),(20.0,0.0),(30.0,0.0),(40.0,0.0),(50.0,0.0),(60.0,0.0),(70.0,0.0),(80.0,0.0),(90.0,0.0),(100.0,0.0),(Infinity,0.0)], histSum = 0.0, histCount = 0}

You can see that while the result of addAndSample for the Counter returns 29 (which was the value just added), the result of observeAndSample for the Histogram returns all zeros (which is the initial state of a histogram).

Running this same program in this PR results in:

CounterSample {unCounterSample = 29}
HistogramSample {histBuckets = fromList [(10.0,0.0),(20.0,0.0),(30.0,0.0),(40.0,0.0),(50.0,0.0),(60.0,1.0),(70.0,1.0),(80.0,1.0),(90.0,1.0),(100.0,1.0),(Infinity,1.0)], histSum = 59.0, histCount = 1}

Now you can see that the resulting Histogram sample correctly has a single observation.

[cdepillabout]

This PR also adds Show for all of our *Sample types.

I'm assuming that we want this for ease of debugging. The only problem I can see with this is if we ever want to add a sample type with a value that isn't showable, we might have to figure out what to do in that case. I imagine the likelihood of that is pretty low.

[clf-cklf]

Might we want Eq as well? Maybe even Ord. Even though we don't use those typeclasses they seem appropriate for what the data appears to represent.

[cdepillabout]

I don't really have a strong opinion on this one.

My personal opinion is to derive Show for everything, since it is helpful for debugging. But only derive other typeclasses as-needed, since it can add additional compile time: https://taylor.fausak.me/2017/08/09/deriving-type-classes-in-haskell-is-slow/

But if we have some sort of coding rule here for what to do, I'd be happy to follow it.