Option<>.collect() not to call PartialFunction on args where it is not defined

[sleepytomcat]

Fixes ##2579

Expected behavior: Option<>.collect() calls PartialFunction passed as argument only for the value on which this PartialFunction is defined.

Observed behavior: PartialFunction called on values for which it is not defined, resulting in undesired behavior (i.e. unchecked exception thrown).

Example code which fails with arithmetic exception (division by zero):

@Test void collect() {	PartialFunction<Integer, Integer> undefinedFor0 = new PartialFunction<>() {	private static final long serialVersionUID = 1L;	@Override public Integer apply(Integer t) {return 1/t;}	@Override public boolean isDefinedAt(Integer value) {return value != 0;}	};	Option<Integer> zero = Option.some(0);	assertTrue(zero.collect(undefinedFor0).isEmpty()); } 

Using io.vavr:vavr:0.10.2

I've checked the code and this is what I've found:

1. Optiona<>.collect() eventually calls PartialFunction.lift(), here

   vavr/src/main/java/io/vavr/control/Option.java

   Line 288 in 0b7a72c

   return flatMap(partialFunction.lift());

2. in turn, PartialFunction.lift() looks like this:

   default Function1<T, Option> lift() {
   return t -> Option.when(isDefinedAt(t), apply(t));
   }

— meaning apply(t) will be always called, no matter if isDefinedAt(t) returns true or false.

vavr/src/main/java/io/vavr/PartialFunction.java

Line 126 in 0b7a72c

return t -> Option.when(isDefinedAt(t), apply(t));

The proposed fix: changing the code to

default Function1<T, Option<R>> lift() { return t -> Option.when(isDefinedAt(t), () -> apply(t)); } 

[sleepytomcat]

I'm fairly new to open source collaboration and pull requests. @mincong-h , can you please guide what are my next steps to merge the changes to trunk?
Thank you!

[mincong-h]

@sleepytomcat , you need to wait until someone who has write permission to approve and merge it for you. My approval is useless because I don't have write permission thus cannot merge your PR. I just came across your PR by curiosity... Usually @danieldietrich will review.

[danieldietrich]

Hi @sleepytomcat, looks like I'm the sleepy tomcat 😅 Sorry for letting you wait a bit.
Your fix looks great, thanks for reporting the bug!