msg_parser.awk

Sofia SIP is an open-source SIP User-Agent library, compliant with the IETF RFC3261 specification.

#! /usr/bin/env awk
#
# This script recreates C files containing header-specific boilerplate stuff
# using the given list of headers (usually obtained from the master structure).
#
# It can also create a parser table.
#
# --------------------------------------------------------------------
#
# This file is part of the Sofia-SIP package
#
# Copyright (C) 2005 Nokia Corporation.
#
# Contact: Pekka Pessi <[e-mail]>
#
# This library is free software; you can redistribute it and/or
# modify it under the terms of the GNU Lesser General Public License
# as published by the Free Software Foundation; either version 2.1 of
# the License, or (at your option) any later version.
#
# This library is distributed in the hope that it will be useful, but
# WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
# Lesser General Public License for more details.
#
# You should have received a copy of the GNU Lesser General Public
# License along with this library; if not, write to the Free Software
# Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA
# 02110-1301 USA
#
# --------------------------------------------------------------------
#
# Contributor(s): [e-mail].
#
# Created: Fri Apr  6 12:59:59 2001 ppessi
#

BEGIN {
  "date '+%a %b %e %H:%M:%S %Y'" | getline date;

  ascii =			       \
     "                               " \
    " !\"#$%&'()*+,-./0123456789:;<=>?" \
    "@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\\]^_" \
    "`abcdefghijklmnopqrstuvwxyz{|}~";
  lower_case = "abcdefghijklmnopqrstuvwxyz";

  N=0;
  # Initialize these as arrays
  split("", symbols);
  split("", names);
  split("", comments); 
  split("", hashes);
  split("", NAMES);
  split("", Comments);
  split("", COMMENTS);

  # indexed by the C name of the header
  split("", Since);		# Non-NUL if extra
  split("", Extra);		# Offset in extra headers

  total = 0;
  ordinary = 0;
  basic = 0;
  extra = 0;
  without_experimental = 0;

  template="";
  template1="";
  template2="";
  template3="";
  prefix="";
  tprefix="";
  failed=0;
  success=0;

  ERRNO="error";
}

function name_hash (name)
{
  hash = 0;

  len = length(name); 

  for (i = 1; i <= len; i++) {
    c = tolower(substr(name, i, 1));
    hash = (38501 * (hash + index(ascii, c))) % 65536;
  }

  if (hash == 0) {
    print "*** msg_parser.awk: calculating hash failed\n";
    exit(5);
  }

  if (0) {
    # Test that hash algorithm above agrees with the C version
    pipe = ("../msg/msg_name_hash " name);
    pipe | getline;
    close(pipe);
    if (hash != $0) {
      print name ": " hash " != " $0 > "/dev/stderr";
    }
  }

  return hash "";
}

#
# Replace magic patterns in template p with header-specific values
#
function protos (name, comment, hash, since)
{
  NAME=toupper(name);
  sub(/.*[\/][*][*][<][ 	]*/, "", comment); 
  sub(/[ 	]*[*][\/].*/, "", comment);
  sub(/[ 	]+/, " ", comment);

  short = match(comment, /[(][a-z][)]/);
  if (short) {
    short = substr(comment, short + 1, 1);
    sub(/ *[(][a-z][)]/, "", comment);
    shorts[index(lower_case, short)] = name;
  }

  do_hash = hash == 0;

  if (do_hash) {
    split(comment, parts, " ");
    name2 = tolower(parts[1]);
    gsub(/-/, "_", name2);
    if (name2 != name && name2 != tprefix "_" name) {
      print name " mismatch with " comment " (" real ")" > "/dev/stderr";
    }
    
    hash = name_hash(parts[1]);

    hashed[name] = hash;

    if (comment !~ /header/) { 
      comment = comment " header";
    }
  }

  Comment = comment;
  if (!do_hash) {
    comment = tolower(comment);
  }
  COMMENT = toupper(comment);

  # Store the various forms into an array for the footer processing
  N++; 
  hashes[N] = hash; 
  names[N] = name;
  NAMES[N] = NAME; 
  comments[N] = comment; 
  Comments[N] = comment; 
  COMMENTS[N] = COMMENT; 

  symbols[name] = comment;
  if (since) {
    Since[name] = since;
  }

  expr = (without_experimental > 0 && do_hash);
  if (expr) {
    printf "%s is experimental\n", Comment;
  }    
  
  experimental[N] = expr;

  if (PR) {
    if (expr) {
      print "#if SU_HAVE_EXPERIMENTAL" > PR;
    }
    replace(template, hash, name, NAME, comment, Comment, COMMENT, since);
    replace(template1, hash, name, NAME, comment, Comment, COMMENT, since);
    replace(template2, hash, name, NAME, comment, Comment, COMMENT, since);
    replace(template3, hash, name, NAME, comment, Comment, COMMENT, since);
    if (expr) {
      print "#endif /* SU_HAVE_EXPERIMENTAL */" > PR;
    }
  }
}

function replace (p, hash, name, NAME, comment, Comment, COMMENT, since)
{
  #
  # Replace various forms of header name in template, print it out
  #
  if (p) {
    gsub(/#hash#/, hash, p);
    gsub(/#xxxxxx#/, name, p); 
    gsub(/#XXXXXX#/, NAME, p);
    gsub(/#xxxxxxx_xxxxxxx#/, comment, p);
    gsub(/#Xxxxxxx_Xxxxxxx#/, Comment, p);
    gsub(/#XXXXXXX_XXXXXXX#/, COMMENT, p);

    if (since) {
      gsub(/#version#/, since, p);
    }
    else {
      # Remove line with #version#
      gsub(/\n[^#\n]*#version#[^\n]*/, "", p);
    }
    	    
    print p > PR;
  }
}

# 
# Repeat each line in the footer containing the magic replacement
# pattern with an instance of all headers
#
function process_footer (text)
{ 
  if (!match(tolower(text), /#(xxxxxx(x_xxxxxxx)?|hash)#/)) {
    n = length(text);
    while (substr(text, n) == "\n") {
      n = n - 1;
      text = substr(text, 1, n);
    }
    if (n > 0)
      print text > PR;
    return;
  }
  
  n = split(text, lines, RS);

  for (i = 1; i <= n; i++) {
    l = lines[i];
    if (match(tolower(l), /#(xxxxxx(x_xxxxxxx)?|hash)#/)) {
      expr = 0;

      for (j = 1; j <= N; j++) {
	l = lines[i];
	if (expr != experimental[j]) {
	  expr = experimental[j];
	  if (expr) {
	    print "#if SU_HAVE_EXPERIMENTAL" > PR;
	  }
	  else {
	    print "#endif /* SU_HAVE_EXPERIMENTAL */" > PR;
	  }
	}
	gsub(/#hash#/, hashes[j], l);
	gsub(/#xxxxxxx_xxxxxxx#/, comments[j], l);
	gsub(/#Xxxxxxx_Xxxxxxx#/, Comments[j], l);
	gsub(/#XXXXXXX_XXXXXXX#/, COMMENTS[j], l);
	gsub(/#xxxxxx#/, names[j], l); 
	gsub(/#XXXXXX#/, NAMES[j], l);
	print l > PR;
      }

      if (expr) {
	print "#endif /* SU_HAVE_EXPERIMENTAL */" > PR;
      }
    } else {
      print l > PR;
    }
  }
}

#
# Read flags used with headers
#
function read_header_flags (flagfile,    line, tokens, name, value)
{
  while ((getline line < flagfile) > 0) {
    sub(/^[ \t]+/, "", line);
    sub(/[ \t]+$/, "", line);
    if (line ~ /^#/ || line ~ /^$/)
      continue;

    split(line, tokens,  /[ \t]*=[ \t]*/); 
    name = tolower(tokens[1]);
    gsub(/-/, "_", name);
    gsub(/,/, " ", name);
    # print "FLAG: " name " = " tokens[2]

    if (header_flags[name]) {
      print flagfile ": already defined " tokens[1];
    }
    else if (!symbols[name]) {
      print flagfile ": unknown header \"" tokens[1] "\"";
    }
    else {
      header_flags[name] = tokens[2];
    }
  }
  close(flagfile);
}

#
# Read in templates
#
function templates ()
{
  if (!auto) {
    auto = FILENAME; 

    if (!prefix) { prefix = module; }
    if (!tprefix) { tprefix = prefix; }

    sub(/.*\//, "", auto);
    auto = "This file is automatically generated from <" auto "> by msg_parser.awk.";

    if (PR) {
      if (TEMPLATE == "") { TEMPLATE = PR ".in"; }
      RS0=RS; RS="\f\n";
      if ((getline theader < TEMPLATE) < 0) {
	print ( TEMPLATE ": " ERRNO );
	failed=1;
        exit(1);
      }
      getline header < TEMPLATE;
      getline template < TEMPLATE;
      getline footer < TEMPLATE;

      if (TEMPLATE1) {
	if ((getline dummy < TEMPLATE1) < 0) {
	  print(TEMPLATE1 ": " ERRNO);
	  failed=1;
          exit(1);
        }
	getline dummy < TEMPLATE1;
	getline template1 < TEMPLATE1;
	getline dummy < TEMPLATE1;
      }

      if (TEMPLATE2) {