#! perl5
#
# wrap.pl Daniel Brockman 20030407 wraps lines of text file
#
# Usage: wrap.pl [nn] <file >outfile
# where nn>0 is the number of characters per line, 
# excluding trailng "\n", defaulting to 60.
#
################################################
# The author grants permission to use for any  #
# purpose, including copying and modification, #
# provided the user acknowledges the author by #
# name.                                        #
# -- Daniel Brockman, author, 20031220         #
################################################
# 
# wrap.pl reads sequentially through the input files.
# wrap.pl trims trailing whitespace from each line it reads.
# wrap.pl replaces consecutive interior whitespace with
# a single space.
# If wrap.pl encounters a word longer than nn chars, then
# wrap.pl breaks it into two words, the first of which is
# the first nn chars. The second word contains the 
# remaining chars. wrap.pl writes the first word on a new 
# line, then considers the 2nd word as a newly encountered
# word. 
# If wrap.pl encounters a line longer than nn chars, then
# wrap.pl finds the last whitespace char before the
# nnth char, and breaks the line just after that 
# whitespace, resulting in two lines. wrap.pl trims 
# leading and trailing whitespace from the first line 
# and writes the first line, then considers the 2nd line 
# as a newly encountered line.
# If wrap.pl encounters a line no longer than nn chars, 
# then wrap.pl writes that line.
#
# Note: To normalize a file, precede wrap.pl with unwrap.pl,
# being aware that special formatting may not survive 
# unwrapping and wrapping.
#
# Example: 
# perl -- unwrap.pl < input > intermediate
# perl -- wrap.pl 45 < intermediate > normalized
#
# 


my(
  $line,       # current line
  $linf,       # next line
  $fracword,   # fractional word on end
  $ll,         # length of line
  $lw,         # length of word
  $pw,         # char pos of lws (last white space) 
  $nn,         # chars per line
  $n,          # number of lines read from file
  @chz,        # chars in line
  $flag,       # whitespace found
  $i,          # aux
);

$n=0;          # init
$linf="";      # init
$nn=60;        # default 60 chars per line
if($#ARGV>=0){
  $nn=$ARGV[0] ;     # chars per line specified
  shift;
}

while($line=<>){     # read a line from file
  $n++;              # incr count
  $line=~s/\s+$// ;  # trim trailing whitespace
  $line=~s/\s+/\ /g ; # replace consecutive interior \s
  if($line=~m/^$/){   # $line empty?
    print "\n";       # print blank line
  } 

  while($line ne ""){  # while nonempty $line
    $line=~s/\s+$// ;  # trim trailing whitespace
    $line=~s/^\s+// ;  # trim leading whitespace
    $ll=length($line); # measure line

    if($ll>$nn){       # long line?
      @chz=split(//,$line);   # get indiv chars

      for($i=$nn-1;$i>=0;$i--){   # loop the chars
	if($chz[$i] eq " "){      # white space?
	  last;                   # leave loop
	}
      }                           # end loop the chars
      $pw=$i ;                    # mark lws

      if($pw<0){                  # long word?
	$pw=$nn ;                 # divide at wraplength
      }
      $linf=substr($line,$pw+1) ; # divide line
      $line=substr($line,0,$pw) ;
      $line=~s/\s+$// ;  # trim trailing whitespace
      $line=~s/^\s+// ;  # trim leading whitespace

    } # end if long line

    print "$line\n";     # write the line
    $line=$linf ;        # remaining unwritten line
    $linf="" ;           # reinit

  }   # end while nonempty line

}     # end while read a line from file

#