Difference between BREAK and REJECT in PARSE

Question

Both in Rebol and Red, there are two PARSE key words: break and reject which have similar behavior: break out of a match loop (such as any, some, while), the difference is that break always indicating success while reject indicate failure. I know what it means literally, but can't figure out a proper scenario where which of the two keywords should be used.

blk: [ 1 #[none] 2 #[none] #[none] 4 5 6 #[none] ]
count: 0
result: parse blk [
    any [ 
            remove none! insert 2 
            if ((count: count + 1) >= 2) break
            | 
            skip
    ]
]
probe blk          ;will get [1 0 2 0 none 4 5 6 none]
probe result       ;will get false
probe count        ;will get 2

In the code above, either break or reject will produce the same result.

So anyone can show me what is the difference between these two keywords?

Answer 1

The truth value produced by break depends on whether the minimum number of iterations in an iterative rule have been reached or not.

>> parse "aaabbb"  [ some [ "a" break ] to end]
== true

Here we have matched "a" at least once, and then broken out of the some rule and continued to the end. This is true.

>> parse "aaabbb"  [ some [ "a" reject ] to end]
== false

Here although we matched "a" once, reject causes the some rule to fail.

See Carl's blog for more details.