[ZEPPELIN-1874] fix: sparkR doens't render output properly

[1ambda]

What is this PR for?

Zeppelin sparkr only shows first element of all outputs due to invalid implementation

// https://github.com/apache/zeppelin/blob/master/spark/src/main/scala/org/apache/zeppelin/spark/ZeppelinRDisplay.scala#L71

private def textDisplay(body: Element): RDisplay = {
  // we are extracting only the first element. it shouldn't
  RDisplay(body.getElementsByTag("p").first().html(), TEXT, SUCCESS)
}

FYI

SparkR interpreter sends message which containing multiple whitespaces, newlines. But zeppelin frontend breaks the space characters. This should be handled by other issue (ZEPPELIN-1914). As you can see the screenshot below, the paragraph result sent from websocket keeps consequent whitespaces and newlines until rendering.

What type of PR is it?

[Bug Fix]

Todos

N/A

What is the Jira issue?

ZEPPELIN-1874

How should this be tested?

1. Build Zeppelin with Spark 2.0+, R, sparkR
2. Execute this paragraph

%spark.r
mtcarsDF <- createDataFrame(mtcars)
model <- glm(vs ~ mpg + disp + hp + wt , data = mtcarsDF, family = "binomial")
summary(model)

3. Check you can get all result.

Deviance Residuals:
(Note: These are approximate quantiles with relative error <= 0.01)
Min 1Q Median 3Q Max
-2.11025 -0.08567 -0.00069 0.13214 1.10483
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) -23.623 23.144 -1.0207 0.30739
mpg 0.78475 0.65026 1.2068 0.2275
disp -0.031549 0.027342 -1.1539 0.24854
hp -0.072188 0.045509 -1.5862 0.11269
wt 7.335 5.2336 1.4015 0.16105
(Dispersion parameter for binomial family taken to be 1)
Null deviance: 43.860 on 31 degrees of freedom
Residual deviance: 12.873 on 27 degrees of freedom
AIC: 22.87
Number of Fisher Scoring iterations: 9

Screenshots (if appropriate)

Buggy (doesn't show full result)

Questions:

• Does the licenses files need update? - NO
• Is there breaking changes for older versions? - NO
• Does this needs documentation? - NO

[Leemoonsoo]

Tested and LGTM

[1ambda]

Failed tests: 
  ZeppelinSparkClusterTest.sparkRTest:186 expected:<[[1] 3]> but was:<[
[1] 3



]>

The above test is failing. I will take a look into.

[felixcheung]

LGTM

[Leemoonsoo]

Merge to master if no more comments