#This script is designed to run with Understand - CodeCheck
use base ("Understand::Codecheck");
use strict;


use constant ERR1 => 'Comment appears to contain code';
our @keywords = qw(double not_eq throw and_eq dynamic_cast operator true asm try auto enum or_eq typedef bitand explicit private typeid bitor extern protected typename bool false public union float register unsigned case reinterpret-cast using catch friend virtual char goto short void class signed volatile compl inline sizeof wchar_t const int static const-cast long static_cast xor continue mutable struct xor_eq default namespace switch delete new template);
our @strongKeywordsRegExp = ('for\s*\(' , 'while\s*\(' , 'if\s*\(');
our @preProc = ("#include", "#define", "#endif","#ifndef", "#ifdef");
our @strongLineEnds = (";","{");
our @maybeLineEnds	= (")");
our $multiRatio = .4;

sub register_tr_text() {
  my $check = shift;
  $check->add_tr_text(ERR1);
}

sub name { return "2-7-2 Sections of code should not be \"commented out\" using C-style comments";}

sub description { return "2-7-2 (Required)  Sections of code should not be \"commented out\" using C-style comments.";}

sub detailed_description { return <<"END_DESC"
<p><b>Rationale</b><br>
Using  C-style  start  and  end  comment  markers  for  this  purpose  is  dangerous  because  C-style 
comments do not support nesting, and any comments already existing in the section of code would 
change the effect.</p>
<p>Additionally, comments should only be used to explain aspects of the code that may not be clear 
from the source code itself. Code that is commented-out may become out of date, which may lead 
to confusion when maintaining the code.</p>
<p>A  more  appropriate  method  of  recording  the  history  of  changes  in  source  code  (e.g.  a  Source 
Control System) should be used instead of commenting-out.</p>
<b>Example</b><pre style="margin-top:0;padding-top:0;">
  void fn ( int32_t i ) 
  { 
     /* 
        ++i;      /* We want to increment "i" */ 
     */
     for ( int32_t j = 0 ; j != i ; ++j ) 
     { 
     } 
  }
</pre>
<p><b>See Also</b><br>
Rule 2-7-1, Rule 2-7-3
</p>

END_DESC
}

sub test_language {
  my $language = shift;
  return $language =~ /C\+\+/; #Handles C and C++
}

sub test_entity { return 1;}

sub test_global { return 0;}

sub define_options{
  my $check = shift;
  $check->option->radio_vert("strict","Aggresivness",["More Aggresive - may result in false positives, comments that are not really code", "Less Aggressive - may result in missing instances of commented code"],"Less Aggressive - may result in missing instances of commented code");
}


sub check {
  my $check = shift; 
  my $file = shift; 
  return unless $file->kind->check("file"); 
  
  my $lexer = $file->lexer;
  return unless $lexer;
  
  my $lexeme = $lexer->first;
  return unless $lexeme;
  my $reported = 0;
  LEXEME:foreach my $lexeme($lexer->lexemes()){
    $reported=0 unless $lexeme->token =~ /Comment|Whitespace|Newline/;
    next LEXEME unless $lexeme->token eq "Comment";
    my $comment = $lexeme->text;
    next LEXEME if $comment =~ '^//';
    $comment =~ s'^/\*|^//|\*/$''g; #Strip comment character

	my $isCode = 0;
	my $maybeCode =0;
	my $lineNum=$lexeme->line_begin;
	my $codeLine=0;
  my $strict = $check->option->lookup("strict") =~ /More Aggresive/;
	#consider each line of the comment separately
	LINE:foreach my $line (split('\n',$comment)){
		my ($strongCode, $strongEnd,$weakCode, $weakEnd,$preproc);
		$line =~ s/^\s+//; #trim leading whitespace
		$line =~ s/\s+$//; #trim trailing whitespace
		
		#Test for preprocessors in comments
		foreach my $word (@preProc){
			$preproc = 1 if $line =~ /\Q$word/;
		}
		
		#Test for strong keyword match
		foreach my $test (@strongKeywordsRegExp){
			$strongCode = 1 if $line =~ /(?<!\w)$test/;
		}	
		
		#Test the comment line has a strong line/statement ending
		foreach my $char (@strongLineEnds){
			$strongEnd = 1 if $line =~ /\Q$char\E$/;
		}
		
		#Test the comment line has a weak line/statement ending
		foreach my $char (@maybeLineEnds){
			$weakEnd = 1 if $line =~ /\Q$char\E$/;
		}
		
		#Test that the comment line has weak statements
		foreach my $word (@keywords){		
			$weakCode = 0 if $line =~ /(?<!\w)\Q$word\E(?!\w)/;
		}
		if($strict){
      $isCode = 1 if $preproc || $strongCode || $strongEnd;
      $isCode =1 if $weakCode && $weakEnd;
      $maybeCode ++ if ($weakCode || $weakEnd) && ! $isCode;
		
    }else{
		$isCode = 1 if $preproc;
		$isCode = 1 if $strongCode && ($strongEnd || $weakEnd);
		$isCode = 1 if $weakCode && $strongEnd;
		$maybeCode++ if $weakCode && $weakEnd;
    }
		
		$codeLine= $lineNum if $isCode;
		last LINE if $isCode;
    
    
	}continue{
		$lineNum++;
	}
	
	my $totalLines = ($lineNum - $lexeme->line_begin);
	$isCode = 1 if $maybeCode && $totalLines > 2 && (($maybeCode / $totalLines) > $multiRatio);

	
    if (!$reported && $isCode){
      $check->violation(0,$file,$codeLine,-1,ERR1);
      $reported=1;
    }
  }
  @keywords = ();
  @strongKeywordsRegExp = ();
  @preProc = ();
  @strongLineEnds = ();
  @maybeLineEnds = ();
  $multiRatio = 0;
}