HyukjinKwon commented on a change in pull request #26429: [SPARK-29777][SparkR]
SparkR::cleanClosure aggressively removes a function required by user function
URL: https://github.com/apache/spark/pull/26429#discussion_r346158731
##########
File path: R/pkg/R/utils.R
##########
@@ -546,8 +546,11 @@ processClosure <- function(node, oldEnv, defVars,
checkedFuncs, newEnv) {
ifelse(identical(func, obj), TRUE, FALSE)
})
if (sum(found) > 0) {
- # If function has been examined, ignore.
- break
+ # If function has been examined
+ if (identical(parent.env(environment(funcList[found][[1]])),
func.env)) {
Review comment:
Sorry I wasn't clear. It looks actually able to break a valid case.
```r
df <- createDataFrame(list(list(n=1)))
ff <- function(n) { ifelse(n == 1, 1, n * ff(n-1)) }
collect(dapply(df, function(x) {
data.frame(a=ff(x$n))
}, structType("a double")))
```
Before:
```
a
1 1
```
After:
```
Error: node stack overflow
```
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
