Tag: shell

Shell / sed Program To Remove All C and C++ Comments From Program File

in Categories Academic, Script last updated May 6, 2008

Write the shell / sed program which remove all the comments from a simple C program stored in your current directory. You can assume that the c source code contains only syntactically correct comments:
+ start with // and end with a newline

+ starting with /* and ending */ (can be multiline)

+ nesting of comments is not allowed

Make sure that C source is not changed at all.

How do I use this sed script?

$ ./script.sed < input.c $ ./script.sed < input.c > output.c
$ for c in *.c; do script.sed < $c > /tmp/zyzcc.c; /bin/cp -f /tmp/zyzcc.c $c; done

Sample sed code to remove all comments from sed

#! /bin/sed -nf
 
# Remove C and C++ comments, by Brian Hiles (brian_hiles@rocketmail.com)
 
# Sped up (and bugfixed to some extent) by Paolo Bonzini (bonzini@gnu.org)
# Works its way through the line, copying to hold space the text up to the
# first special character (/, ", ').  The original version went exactly a
# character at a time, hence the greater speed of this one.  But the concept
# and especially the trick of building the line in hold space are entirely
# merit of Brian.
 
:loop
 
# This line is sufficient to remove C++ comments!
/^\/\// s,.*,,
 
/^$/{
  x
  p
  n
  b loop
}
/^"/{
  :double
  /^$/{
    x
    p
    n
    /^"/b break
    b double
  }
 
  H
  x
  s,\n\(.[^\"]*\).*,\1,
  x
  s,.[^\"]*,,
 
  /^"/b break
  /^\\/{
    H
    x
    s,\n\(.\).*,\1,
    x
    s/.//
  }
  b double
}
 
/^'/{
  :single
  /^$/{
    x
    p
    n
    /^'/b break
    b single
  }
  H
  x
  s,\n\(.[^\']*\).*,\1,
  x
  s,.[^\']*,,
 
  /^'/b break
  /^\\/{
    H
    x
    s,\n\(.\).*,\1,
    x
    s/.//
  }
  b single
}
 
/^\/\*/{
  s/.//
  :ccom
  s,^.[^*]*,,
  /^$/ n
  /^\*\//{
    s/..//
    b loop
  }
  b ccom
}
 
:break
H
x
s,\n\(.[^"'/]*\).*,\1,
x
s/.[^"'/]*//
b loop

Please note that this program is copied from public domain.