Strip block comments
You are encouraged to solve this task according to the task description, using any language you may know.
A block comment begins with a begining delimiter and ends with a ending delimiter, including the delimiters.
Task : Strips block comments from text (of codes). Your demos should at least handle simple, non-nested and multiline block comment delimiters.
Sample text for striping.
/** * Some comments * longer comments here that we can parse. * * Rahoo */ function subroutine() { a = /* inline comment */ b + c ; } /*/ <-- triky comments */ /** * Another comment. */ function something() { }
D
<lang d>import std.stdio ; import std.regexp, std.algorithm ;
string[] sepComment(string s, string cpat[] ...) {
assert(cpat.length == 2, "sepComment : 2 pattern arguments for comment begin & end") ; string[] res = new string[](2) ; int p = 0, q = 0 /* cursors */, ic = 0 ; // inside comment? int[] plen = new int[](2) ; // this's for handling /*/
bool advCursor() { auto m = std.regexp.search(s[p..$], cpat[ic]) ; if(m is null) return false ; plen[ic] = max(0, plen[ic], m[0].length) ; q = p + m.pre.length ; // got comment head if(ic) { q += m[0].length ; } // or comment tail if(std.regexp.find(m[0], "\n|\r") != -1) // special adjust for \n\r q-- ; return true ; }
while(true) { if(!advCursor()) break ; res[ic] ~= s[p..q] ; // save slice of result if( ic && (q - p < plen[0] + plen[1])) { // this handle /*/ pattern p = q ; if(!advCursor()) break ; res[ic] ~= s[p..q] ; // save result again } p = q ; // advance cursor ic = 1 - ic ; // toggle search type } if(ic) throw new Exception("Mismatched Comment") ; res[ic] ~= s[p..$] ; // save rest(non-comment) return res ;
}
void main() {
string s = ` /** * Some comments * longer comments here that we can parse. * * Rahoo */ function subroutine() { a = /* inline comment */ b + c ; } /*/ <-- triky comments */
/** * Another comment. */ function something() { }` ;
writefln("==original:\n%s", s) ; auto t = sepComment(s, `/\*`, `\*/`) ; writefln("==comment stripped:\n%s\n==stripped comment:\n%s", t[0], t[1]) ;
s = "apples, pears # and bananas
apples, pears ; and bananas " ; // test for line comment
writefln("==original:\n%s", s) ; t = sepComment(s, `#|;`, `[\n\r]|$`) ; writefln("==comment stripped:\n%s\n==stripped comment:\n%s", t[0], t[1]) ;
}</lang> part of output:
==comment stripped: function subroutine() { a = b + c ; } function something() { } ==stripped comment: /** * Some comments * longer comments here that we can parse. * * Rahoo *//* inline comment *//*/ <-- triky comments *//** * Another comment. */
Perl 6
<lang perl6>sample().split(/ '/*' .+? '*/' /).print;
sub sample { ' /**
* Some comments * longer comments here that we can parse. * * Rahoo */ function subroutine() { a = /* inline comment */ b + c ; } /*/ <-- triky comments */
/** * Another comment. */ function something() { }
'}</lang>
Output:
function subroutine() { a = b + c ; } function something() { }
PicoLisp
<lang PicoLisp>(in "sample.txt"
(while (echo "/*") (out "/dev/null" (echo "*/")) ) )</lang>
Output:
function subroutine() { a = b + c ; } function something() { }
Tcl
<lang tcl>proc stripBlockComment {string {open "/*"} {close "*/"}} {
# Convert the delimiters to REs by backslashing all non-alnum characters set openAsRE [regsub -all {\W} $open {\\&}] set closeAsRE [regsub -all {\W} $close {\\&}] # Now remove the blocks using a dynamic non-greedy regular expression regsub -all "$openAsRE.*?$closeAsRE" $string ""
}</lang> Demonstration code: <lang tcl>puts [stripBlockComment " /**
* Some comments * longer comments here that we can parse. * * Rahoo */ function subroutine() { a = /* inline comment */ b + c ; } /*/ <-- triky comments */
/** * Another comment. */ function something() { }
"]</lang> Output:
function subroutine() { a = b + c ; } function something() { }