What is the difference between regexp.Compile and regexp.CompilePOSIX?

Question

Can anyone provide some examples to explain the differences between regexp.Compile and regexp.CompilePOSIX ? I read the documentation. But I can't get an intuitive understanding.

Answer 1

Perl- and POSIX-compatible regular expressions are similar in large parts, but differ in some key aspects, like for example submatching . This is mentioned here :

POSIX defines that to resolve submatches, first chose the match that starts leftmost in the string. (This is traditional Perl behavior but here things diverge.) Among the submatches starting at the leftmost position in the string, choose the longest one overall.

Say you have a regular expression (foo|foobar) . When matching this expression against a string that would match several of the subexpressions (for example, foobarbaz would match both subpatterns, foo and foobar ), a Perl-compatible regex would return the first match ( foo ), while a POSIX-compatible regex would return the longest match ( foobar ).

Some example code ( playground ):

package main

import "fmt"
import "regexp"

func main() {
    pattern := "(foo|foobar)"
    str := []byte("foobarbaz")

    rPCRE, _ := regexp.Compile(pattern)
    rPOSIX, _ := regexp.CompilePOSIX(pattern)

    matchesPCRE := rPCRE.Find(str)
    fmt.Println(string(matchesPCRE))
    // prints "foo"

    matchesPOSIX := rPOSIX.Find(str)
    fmt.Println(string(matchesPOSIX))
    // prints "foobar"
}