Initial import of svn+ssh:://svn.icir.org/bro/trunk/bro as of r7088

COPYING

@@ -0,0 +1,49 @@
+Copyright (c) 1995-2008, The Regents of the University of California,
+through Lawrence Berkeley National Laboratory. All rights reserved.
+
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are met:
+
+(1) Redistributions of source code must retain the above copyright notice,
+    this list of conditions and the following disclaimer.
+
+(2) Redistributions in binary form must reproduce the above copyright
+    notice, this list of conditions and the following disclaimer in the
+    documentation and/or other materials provided with the distribution.
+
+(3) Neither the name of the University of California, Lawrence Berkeley
+    National Laboratory, U.S. Dept. of Energy, International Computer
+    Science Institute, nor the names of contributors may be used to endorse
+    or promote products derived from this software without specific prior
+    written permission.
+
+THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
+LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
+CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
+SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
+INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
+CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
+POSSIBILITY OF SUCH DAMAGE.
+
+Note that some files in the Bro distribution carry their own copyright
+notices.  The above applies to the Bro scripts in policy/ (other than as
+noted below) and the source files in src/ , other than:
+
+	policy/sigs/p0fsyn.osf
+	src/H3.h
+	src/OSFinger.cc
+	src/OSFinger.h
+	src/bsd-getopt-long.c
+	src/bsd-getopt-long.h
+	src/md5.c
+	src/md5.h
+	src/patricia.c
+	src/patricia.h
+
+In addition, the build components such as Makefile.in, acinclude.m4, and
+others have separate copyrights, as do a number of the elements in the
+aux/ subdirectory and in scripts/s2b/snort_rules2.2/ .

Checklist-for-Release

@@ -0,0 +1,42 @@
+- Make sure BroV6 works.
+
+- Make sure --enable-int64 builds w/o warnings.
+
+- Update TODO-For-Next-Release.
+
+- Build distribution on a system with an up-to-date yacc (MacOS suffices).
+
+- make distcheck
+
+- Edit CHANGES to final version.
+
+- Fix VERSION to final value.
+
+- Update version info (and perhaps other stuff) in README.
+
+- Check everything in.
+
+- Make an SVN branch for the release:
+
+    svn cp svn+ssh://svn.icir.org/bro/trunk \
+	    svn+ssh://svn.icir.org/bro/releases/release_1_X
+
+- scp bro-XXX.tar.gz crd.lbl.gov:/ftp/BROIDS/
+  scp CHANGES crd.lbl.gov:/ftp/BROIDS/bro-change-log.txt
+
+- Fix symlinks on crd:
+
+	bro-1.X-release.tar.gz
+	bro-1.X-current.tar.gz
+	bro-1.<n>-release.tar.gz
+
+- Update crd:/www/BROIDS/download.html to reflect new version.  This page
+  is generated from trunk/bro-web/download.xml.  Edit this file, and also
+  update the (web page) version in build.xml, the copyright year in
+  navigation.xml, and create a news entry in news.xml.
+
+  Type 'ant style' on a machine with 'ant' installed, and copy
+  html/*.html to crd.lbl.gov:/www/BROIDS/ .
+
+- Send email to bro@bro-ids.org.  Look for "release now available" in
+  previous messages as a template.

FILES.bin

@@ -0,0 +1,3 @@
+README
+VERSION
+bro

INSTALL

@@ -0,0 +1,104 @@
+Prerequisites
+=============
+
+Bro relies on the following libraries and tools, which need to be installed
+before you begin with the installation:
+
+    * Libpcap
+	   If libpcap is already installed on the system, by default Bro
+	   will use that one.  Otherwise, it falls back to a version shipped
+	   with the Bro distribution.
+
+    * Flex
+	   Flex is already installed on most systems, so with luck you can
+	   skip having to install it yourself.
+
+    * Bison or byacc
+	   These come with many systems, but if you get errors compiling
+	   parse.y, you will need to install them.  bison is available
+	   from GNU sites such as ftp.gnu.org.
+
+    * BIND8 headers and libraries
+           These are usually already installed as well.
+
+    * Autotools
+	   If you have checked the source out from Bro's Subversion
+	   repository, you need the autotools suite installed.  In this
+	   case, run "./autogen.sh" first right after the check out.
+	   Otherwise the installation steps below will fail.
+
+Bro can also make uses of some optional libraries if they are found at
+installation time:
+
+    * OpenSSL
+           For analysis of SSL certificates by the HTTP analyzer, and 
+           for encrypted Bro-to-Bro communication.
+
+    * Libmagic
+           For identifying file types (e.g., in FTP transfers).
+           
+    * LibGeoIP
+           For geo-locating IP addresses.
+           
+    * Libz
+	   For decompressing HTTP bodies by the HTTP analyzer, and for
+	   compressed Bro-to-Bro communication.
+
+    * Endace's DAG tools:
+           For native support of Endace DAG cards.
+
+
+Installation
+============
+
+To build and install into /usr/local:
+
+	> ./configure
+	> make
+	> make install
+
+This will install the Bro binary into /usr/local/bin/bro and the policy
+files into /usr/local/share/bro.
+
+As usual you can specify a different installation directory with
+
+   > ./configure --prefix=<dir>".
+
+Run "./configure --help" for more options. 
+
+
+Running Bro
+===========
+
+Bro is a complex program and it takes a bit of time to get familiar
+with it.  In the following we give a few simple examples.  See
+http://www.bro-ids.org/wiki for more information.
+
+To run a policy file from /usr/local/share/bro, such as mt.bro, on a
+previously captured tcpdump save file named foo:
+
+	bro -r foo mt.bro
+
+To run from interface le0:
+
+	bro -i le0 mt
+
+You can alternatively specify interface and scripts to load in your own
+policy file:
+
+	@load mt
+	redef interfaces = "le0";
+
+and then run
+
+    bro ./my-policy.bro
+
+You can see the BPF filter Bro will use (if not overridden) by executing
+
+	bro mt print-filter
+
+To run interactively (e.g., for playing with expression evaluation):
+
+	bro
+
+"bro -h" lists the various options.

Makefile.am

@@ -0,0 +1,64 @@
+## Process this file with automake to produce Makefile.in
+
+# snag the whole linux-include directory
+EXTRA_DIST = CHANGES README VERSION shtool linux-include \
+	autogen.sh depcomp ylwrap
+
+# When running distcheck, make sure we skip building GtkDoc-based
+# documentation. This applies to Broccoli only, and needs to be
+# duplicated here because DISTCHECK_CONFIGURE_FLAGS isn't otherwise
+# noticed.
+#
+DISTCHECK_CONFIGURE_FLAGS = --disable-gtk-doc
+
+chown = @CHOWN@
+
+# aux before src so we compile the libpcap
+SUBDIRS = aux src scripts policy doc
+
+test:
+	( cd ../testing && $(MAKE) test )
+
+install-broctl:
+	$(MAKE) install
+	( cd aux/broctl && $(MAKE) install-broctl )
+
+# Deprecated. Don't use. 
+install-brolite:
+	$(MAKE) install
+	$(INSTALL) -d $(prefix)/logs
+	$(INSTALL) -d $(prefix)/archive
+	$(INSTALL) -d $(prefix)/var
+	( cd scripts && $(MAKE) install-brolite )
+	( cd aux && $(MAKE) install-brolite )
+	- @CHOWN@ -R `cat scripts/bro_user_id` ${prefix}/
+	@echo "*********************************************************"
+	@echo "Please run \"${prefix}/etc/bro.rc --start\" to start bro"
+	@echo "*********************************************************"
+
+docs:
+	( cd doc && $(MAKE) doc )
+
+doc-install:
+	( cd doc && $(MAKE) doc-install )
+
+update:
+	( cd scripts && $(MAKE) update )
+	( cd policy && $(MAKE) install )
+
+update-sigs:
+	(cd scripts && $(MAKE) update-sigs )
+
+reports:
+	( cd scripts && $(MAKE) reports )
+
+# make sure we don't leak CVS/SVN or private policy files
+dist-hook:
+	rm -rf `find $(distdir) -name CVS`
+	rm -rf `find $(distdir) -name .svn`
+	rm -rf $(distdir)/policy/local 
+
+release:
+	./autogen.sh
+	./configure
+	$(MAKE) distcheck

README

@@ -0,0 +1,36 @@
+This is release 1.5 of Bro, a system for detecting network intruders in
+real-time using passive network monitoring.
+
+Please see the file INSTALL for installation instructions and some examples
+on how to run Bro.  For more documentation, see the Bro Wiki:
+
+  http://www.bro-ids.org/wiki/index.php/User_Manual
+
+Please note that this documentation is preliminary and still missing pieces.
+PDF and HTML versions of older versions of the manuals are also available
+in the doc/ directory.
+
+There's also in doc/misc/conn-logs/ a brief summary of the connection logs
+generated by the sample policy scripts (which are in policy/).
+
+Numerous other Bro-related publications, including a paper describing the
+system, can be found at
+
+  http://www.bro-ids.org/publications.html
+
+Some auxiliary scripts and utilities are available in the aux/ directory.
+Note that these are not installed by default.
+
+Send comments, etc., to the Bro mailing list, bro@bro-ids.org.  However,
+please note that you must first subscribe to the list in order to be able
+to post to it.
+
+- Vern Paxson & Robin Sommer, on behalf of the Bro development team
+
+Lawrence Berkeley National Laboratory
+University of California, Berkeley  USA
+
+ICSI Center for Internet Research (ICIR)
+International Computer Science Institute
+Berkeley, CA  USA
+vern@icir.org / robin@icir.org

TODO-For-Next-Release

@@ -0,0 +1,9 @@
+Plan for 1.6:
+	Originally, with 1.5 we were going to start working with --use-binpac
+	as the default.  However, this has been deferred pending development
+	of BinPAC++.  We might however turn on BinPAC for the SSL analyzer,
+	for which the BinPAC version is more robust.  It, though, doesn't
+	support storing certs to disk, which some folks use operationally.
+
+	Given DPD means we might not filter traffic anyway, we no longer
+	have such a good excuse for not dealing with IPv6 options.

VERSION

@@ -0,0 +1 @@
+1.5.2.7

autogen.sh

@@ -0,0 +1,143 @@
+#!/bin/sh
+
+# Initialization script to set up the initial configuration files etc.
+# shtool usage inspired by the autogen script of the ferite scripting
+# language -- cheers Chris :)
+#
+# This is 'borrowed' from netdude, with minor changes for bro 
+
+BLD_ON=`./shtool echo -n -e %B`
+BLD_OFF=`./shtool echo -n -e %b`
+
+srcdir=`dirname $0`
+NAME=bro
+
+DIE=0
+
+echo
+echo "               "${BLD_ON}"BRO Build Tools Setup"${BLD_OFF}
+echo "===================================================="
+echo
+echo "Checking whether we have all tools available ..."
+
+(autoconf --version) < /dev/null > /dev/null 2>&1 || {
+  echo
+  echo ${BLD_ON}"Error"${BLD_OFF}": You must have \`autoconf' installed to."
+  echo "Download the appropriate package for your distribution,"
+  echo "or get the source tarball at ftp://ftp.gnu.org/pub/gnu/"
+  DIE=1
+}
+
+(automake --version) < /dev/null > /dev/null 2>&1 || {
+  echo
+  echo ${BLD_ON}"Error"${BLD_OFF}": You must have \`automake' installed."
+  echo "Get ftp://ftp.gnu.org/pub/gnu/automake-1.3.tar.gz"
+  echo "(or a newer version if it is available)"
+  DIE=1
+  NO_AUTOMAKE=yes
+}
+
+# if no automake, don't bother testing for aclocal
+test -n "$NO_AUTOMAKE" || (aclocal --version) < /dev/null > /dev/null 2>&1 || {
+  echo
+  echo ${BLD_ON}"Error"${BLD_OFF}": Missing \`aclocal'.  The version of \`automake'"
+  echo "installed doesn't appear recent enough."
+  echo "Get ftp://ftp.gnu.org/pub/gnu/automake-1.3.tar.gz"
+  echo "(or a newer version if it is available)"
+  DIE=1
+}
+
+if test "$DIE" -eq 1; then
+  exit 1
+fi
+
+echo "All necessary tools found."
+echo
+
+if [ -d autom4te.cache ] ; then
+    echo "Removing autom4te.cache ..."
+    rm -rf autom4te.cache
+    #echo
+    #echo ${BLD_ON}"Error"${BLD_OFF}": autom4te.cache directory exists"
+    #echo "please remove it, and rerun this script"
+    #echo 
+    #exit 1
+fi
+
+echo
+echo "running "${BLD_ON}"aclocal"${BLD_OFF}
+echo "----------------------------------------------------"
+aclocal -I . $ACLOCAL_FLAGS
+if [ $? -ne 0 ]; then
+    echo "*** ERROR($NAME), aborting."
+    exit 1
+fi
+
+echo
+echo "running "${BLD_ON}"autoheader"${BLD_OFF}
+echo "----------------------------------------------------"
+autoheader
+if [ $? -ne 0 ]; then
+    echo "*** ERROR($NAME), aborting."
+    exit 1
+fi
+
+echo
+echo "running "${BLD_ON}"automake"${BLD_OFF}
+echo "----------------------------------------------------"
+automake -a -c 
+if [ $? -ne 0 ]; then
+    echo "*** ERROR($NAME), aborting."
+    exit 1
+fi
+
+echo
+echo "running "${BLD_ON}"autoconf"${BLD_OFF}
+echo "----------------------------------------------------"
+autoconf
+if [ $? -ne 0 ]; then
+    echo "*** ERROR($NAME), aborting."
+    exit 1
+fi
+
+echo
+echo
+echo "Running aux/binpac/autogen.sh"
+echo "----------------------------------------------------"
+(cd aux/binpac/ && BROBUILD=yes ./autogen.sh)
+if [ $? -ne 0 ]; then
+    echo "*** ERROR($NAME), aborting."
+    exit 1
+fi
+
+echo
+echo
+echo "Running aux/broccoli/autogen.sh"
+echo "----------------------------------------------------"
+(cd aux/broccoli/ && BROBUILD=yes ./autogen.sh)
+if [ $? -ne 0 ]; then
+    echo "*** ERROR($NAME), aborting."
+    exit 1
+fi
+
+echo
+echo
+echo "Running aux/broctl/aux/capstats/autogen.sh"
+echo "----------------------------------------------------"
+(cd aux/broctl/aux/capstats && ./autogen.sh)
+if [ $? -ne 0 ]; then
+    echo "*** ERROR($NAME), aborting."
+    exit 1
+fi
+
+echo
+echo 
+echo "Setup finished. Now run:"
+echo
+echo "  $ "${BLD_ON}"./configure"${BLD_OFF}" (with options as needed, try --help)"
+echo
+echo "and then"
+echo
+echo "  $ "${BLD_ON}"make"${BLD_OFF}
+echo "  # "${BLD_ON}"make install"${BLD_OFF}
+echo

aux/Makefile.am

@@ -0,0 +1,77 @@
+## Process this file with automake to produce Makefile.in
+
+LIBPCAP_VER = libpcap-0.9.8
+LIBPCAP_LIB = $(LIBPCAP_VER)/libpcap.a
+
+EXTRA_DIST = README $(LIBPCAP_VER).tar.gz
+
+# if we don't have ssl, can't build bdcat
+if USE_SSL
+bdcat_dir = bdcat
+else
+bdcat_dir =
+endif
+
+# don't compile libpcap if they did a '--disable-localpcap' to configure
+if USE_LOCALPCAP
+built_srcs =  $(LIBPCAP_LIB)
+LARGE_FILE =  "-D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE -D_LARGEFILE64_SOURCE"
+else
+built_srcs =  
+endif
+
+if USEV6
+PCAPARGS = --enable-ipv6
+else
+PCAPARGS = 
+endif
+
+if USE_BROCCOLI
+broccoli = broccoli
+else
+broccoli =
+endif
+
+if USE_BROCTL
+broctl = broctl
+else
+broctl =
+endif
+
+BUILT_SOURCES = $(built_srcs)
+
+SUBDIRS = adtrace binpac cf hf nftools rst scripts \
+		$(bdcat_dir) $(broccoli) $(broctl)
+
+DIST_SUBDIRS = adtrace binpac cf hf nftools rst scripts \
+		$(bdcat_dir) $(broccoli) $(broctl)
+
+clean-local:
+	rm -rf $(LIBPCAP_VER)
+
+
+$(LIBPCAP_LIB): $(top_srcdir)/aux/$(LIBPCAP_VER).tar.gz
+	@echo "Unpacking libpcap sources"
+	@gzip -d < $(top_srcdir)/aux/$(LIBPCAP_VER).tar.gz | tar xf -
+	@echo "Building libpcap"
+	( cd $(LIBPCAP_VER) && ./configure --prefix=$(prefix) $(PCAPARGS) CFLAGS=$(LARGE_FILE) && $(MAKE) )
+	@chmod -R 755 $(LIBPCAP_VER)
+
+
+# This is a hack. These are hardcoded here to mimic the previous
+# brolite installation. While these should better go into the 
+# subdirs' Makefile.am, it's not really worth the effort as 
+# we will get rid of all this at some point anyway.
+install-brolite:
+	$(INSTALL) ./hf/hf ${bindir}
+	$(INSTALL) ./hf/nf ${bindir}
+	$(INSTALL) ./hf/pf ${bindir}
+	$(INSTALL) ./cf/cf ${bindir}
+	$(INSTALL) ./rst/rst ${bindir}
+	- install -d $(prefix)/scripts/
+	$(INSTALL) ./scripts/host-to-addrs $(prefix)/scripts
+	$(INSTALL) ./scripts/bro-logchk.pl $(prefix)/scripts
+	$(INSTALL) ./scripts/host-grep $(prefix)/scripts
+	$(INSTALL) ./scripts/mvlog $(prefix)/scripts
+
+

aux/README

@@ -0,0 +1,61 @@
+This directory contains handy auxiliary programs:
+
+adtrace/
+	Makefile and source for the adtrace utility. This program is used
+	in conjunction with the localnetMAC.pl perl script to compute the
+	network address that compose the internal and extern nets that bro
+	is monitoring. This program when run by itself just reads a pcap
+	(tcpcump) file and writes out the src MAC, dst MAC, src IP, dst
+	IP for each packet seen in the file. This output is processed by
+	the localnetMAC.pl script during 'make install'.
+
+bdcat/
+	A utility for decrypting encrypted Bro log files.
+
+binpac/
+	A compiler for generating protocol analyzers from high-level,
+	declarative specifications.  Used extensively for constructing
+	Bro's protocol analyzers, but capable of stand-alone use for
+	building analyzers outside of the Bro system.
+
+broccoli/
+	A C client library for interfacing programs with the Bro system.
+	Enables sending and receiving of Bro values and events.
+
+cf/
+	Makefile and source for the "cf" utility.  cf reads lines from
+	stdin and if the line begins with a number, then it assumes that
+	the number corresponds to a Unix timestamp and replaces it with
+	the corresponding local time in a readable format.  Useful for
+	running on log files.  See cf/cf.man.txt for documentation.
+
+contrib/
+	Unsupported contributions to Bro.
+
+hf/
+	The main utility in this subdirectory is hf, which translates
+	any dotted-quad (in text) appearing on stdin to the corresponding
+	DNS hostname (via a PTR lookup) on stdout.
+
+nftools/
+	Utilities for dealing with Bro's custom file format for storing
+	NetFlow records.  nfcollector reads NetFlow data from a socket
+	and writes it in Bro's format.  ftwire2bro reads NetFlow "wire"
+	format (e.g., as generated by a 'flow-export' directive) and writes
+	it in Bro's format.
+
+rst/
+	Makefile and source for the rst utility. "rst" can be invoked by
+	a Bro script to terminate an established TCP connection by forging
+	RST tear-down packets.  See terminate_connection() in conn.bro.
+
+scripts/
+	A set of utility scripts for munching on Bro connection summaries.
+
+	bro_logchk: orders and scans through FTP and HTTP logs
+	host-grep: greps a summary file for a particular host's activities
+	host-to-addrs: converts a hostname to a list of IP addresses
+	hot-report: formats a summary file in a readable fashion
+	ip-grep: returns a grep pattern for a given IP address
+	mon-report: summarizes a particular host's activity
+	mvlog: compresses and archives log files

aux/adtrace/Makefile.am

@@ -0,0 +1,10 @@
+## Process this file with automake to produce Makefile.in
+
+AM_CFLAGS=@V_INCLS@
+
+# Should use AM_ vars, but automake 1.5 errors out.
+#AM_LDFLAGS = @LDFLAGS@
+LDFLAGS = @LDFLAGS@
+
+noinst_PROGRAMS = adtrace
+adtrace_SOURCES = adtrace.c ether.h ethertype.h ip.h

aux/adtrace/adtrace.c

@@ -0,0 +1,92 @@
+#include <stdio.h>
+#include <unistd.h>
+#include <stdlib.h>
+#include <sys/types.h>
+#include <sys/socket.h>
+
+#include <netinet/in.h>
+#include <arpa/inet.h>
+#include <pcap.h>
+
+#include "../../config.h"
+#include "ip.h"
+#include "ether.h" 
+#include "ethertype.h"
+
+pcap_t *p;
+
+const u_char*  printEAddr(const u_char* pkt, u_char* endp){
+  const struct ether_header *ep;
+  int i=0;
+  ep = (const struct ether_header*) pkt;
+
+  if (pkt+ETHER_HDRLEN > endp ||
+      ntohs(ep->ether_type) != ETHERTYPE_IP){
+    return 0;
+  }
+
+  for (i = 0; i<ETHER_ADDR_LEN; i++){
+    if (i>0) putchar(':');
+    printf("%02x", ep->ether_shost[i]);
+  }
+  putchar (' ');
+  for (i = 0; i<ETHER_ADDR_LEN; i++){
+    if (i>0) putchar(':');
+    printf("%02x", ep->ether_dhost[i]);
+  }
+  putchar(' ');
+  return (pkt+ETHER_HDRLEN);
+}
+
+void printIPAddr(const u_char* pkt, u_char* endp){
+  const struct ip* iph;
+  if (pkt+sizeof(struct ip) > endp) return;
+  iph = (const struct ip*) pkt;
+  fputs ((char*) inet_ntoa(iph->ip_src), stdout);
+  putchar(' ');
+  puts ((char*) inet_ntoa(iph->ip_dst));
+}
+
+void handler(u_char *user, const struct pcap_pkthdr *head, const u_char *packet){
+  u_char* endp;
+
+  endp =(u_char*) packet + head->caplen;
+  packet = printEAddr(packet, endp);
+  if (packet)
+    printIPAddr(packet, endp);
+}
+
+void usage(char *av[])
+{
+	fprintf(stderr,"usage: %s filename \n", av[0]);
+	exit(1);
+}
+
+int main (int argc, char *argv[])
+{
+  char *file;
+  char errbuf[PCAP_ERRBUF_SIZE];
+  u_char* pkt, endp; 
+  struct pcap_pkthdr *head;
+
+  if ( argc != 2 ) 
+	  usage(argv);
+
+  file = argv[1];
+
+  p = pcap_open_offline(file, errbuf);
+  if(p==NULL){
+    fprintf (stderr, "cannot open %s: %s\n", file, errbuf);
+    exit(2);
+  }
+  
+  if (pcap_datalink(p) != DLT_EN10MB){
+    fputs ("sorry, currently only ethernet links supported\n", stderr);
+    exit(1); //if it is not ethernet we are watching we won't have MACs
+  }
+
+  pcap_loop(p, -1, handler, NULL);
+  pcap_close(p);
+  return(0);
+}
+

aux/adtrace/ether.h

@@ -0,0 +1,59 @@
+/* @(#) $Header$ (LBL) */
+/*
+ * Copyright (c) 1982, 1986, 1993
+ *	The Regents of the University of California.  All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. All advertising materials mentioning features or use of this software
+ *    must display the following acknowledgement:
+ *	This product includes software developed by the University of
+ *	California, Berkeley and its contributors.
+ * 4. Neither the name of the University nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
+ * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ *
+ *	@(#)if_ether.h	8.3 (Berkeley) 5/2/95
+ */
+
+#define	ETHERMTU	1500
+
+/*
+ * The number of bytes in an ethernet (MAC) address.
+ */
+#define	ETHER_ADDR_LEN		6
+
+/*
+ * Structure of a DEC/Intel/Xerox or 802.3 Ethernet header.
+ */
+struct	ether_header {
+	u_int8_t	ether_dhost[ETHER_ADDR_LEN];
+	u_int8_t	ether_shost[ETHER_ADDR_LEN];
+	u_int16_t	ether_type;
+};
+
+/*
+ * Length of a DEC/Intel/Xerox or 802.3 Ethernet header; note that some
+ * compilers may pad "struct ether_header" to a multiple of 4 bytes,
+ * for example, so "sizeof (struct ether_header)" may not give the right
+ * answer.
+ */
+#define ETHER_HDRLEN		14

aux/adtrace/ethertype.h

@@ -0,0 +1,122 @@
+/*
+ * Copyright (c) 1993, 1994, 1996
+ *	The Regents of the University of California.  All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that: (1) source code distributions
+ * retain the above copyright notice and this paragraph in its entirety, (2)
+ * distributions including binary code include the above copyright notice and
+ * this paragraph in its entirety in the documentation or other materials
+ * provided with the distribution, and (3) all advertising materials mentioning
+ * features or use of this software display the following acknowledgement:
+ * ``This product includes software developed by the University of California,
+ * Lawrence Berkeley Laboratory and its contributors.'' Neither the name of
+ * the University nor the names of its contributors may be used to endorse
+ * or promote products derived from this software without specific prior
+ * written permission.
+ * THIS SOFTWARE IS PROVIDED ``AS IS'' AND WITHOUT ANY EXPRESS OR IMPLIED
+ * WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED WARRANTIES OF
+ * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.
+ *
+ * @(#) $Header$ (LBL)
+ */
+
+/*
+ * Ethernet types.
+ *
+ * We wrap the declarations with #ifdef, so that if a file includes
+ * <netinet/if_ether.h>, which may declare some of these, we don't
+ * get a bunch of complaints from the C compiler about redefinitions
+ * of these values.
+ *
+ * We declare all of them here so that no file has to include
+ * <netinet/if_ether.h> if all it needs are ETHERTYPE_ values.
+ */
+
+#ifndef ETHERTYPE_PUP
+#define	ETHERTYPE_PUP		0x0200	/* PUP protocol */
+#endif
+#ifndef ETHERTYPE_IP
+#define	ETHERTYPE_IP		0x0800	/* IP protocol */
+#endif
+#ifndef ETHERTYPE_ARP
+#define ETHERTYPE_ARP		0x0806	/* Addr. resolution protocol */
+#endif
+#ifndef ETHERTYPE_REVARP
+#define ETHERTYPE_REVARP	0x8035	/* reverse Addr. resolution protocol */
+#endif
+#ifndef ETHERTYPE_NS
+#define ETHERTYPE_NS		0x0600
+#endif
+#ifndef	ETHERTYPE_SPRITE
+#define	ETHERTYPE_SPRITE	0x0500
+#endif
+#ifndef ETHERTYPE_TRAIL
+#define ETHERTYPE_TRAIL		0x1000
+#endif
+#ifndef	ETHERTYPE_MOPDL
+#define	ETHERTYPE_MOPDL		0x6001
+#endif
+#ifndef	ETHERTYPE_MOPRC
+#define	ETHERTYPE_MOPRC		0x6002
+#endif
+#ifndef	ETHERTYPE_DN
+#define	ETHERTYPE_DN		0x6003
+#endif
+#ifndef	ETHERTYPE_LAT
+#define	ETHERTYPE_LAT		0x6004
+#endif
+#ifndef ETHERTYPE_SCA
+#define ETHERTYPE_SCA		0x6007
+#endif
+#ifndef ETHERTYPE_REVARP
+#define ETHERTYPE_REVARP	0x8035
+#endif
+#ifndef	ETHERTYPE_LANBRIDGE
+#define	ETHERTYPE_LANBRIDGE	0x8038
+#endif
+#ifndef	ETHERTYPE_DECDNS
+#define	ETHERTYPE_DECDNS	0x803c
+#endif
+#ifndef	ETHERTYPE_DECDTS
+#define	ETHERTYPE_DECDTS	0x803e
+#endif
+#ifndef	ETHERTYPE_VEXP
+#define	ETHERTYPE_VEXP		0x805b
+#endif
+#ifndef	ETHERTYPE_VPROD
+#define	ETHERTYPE_VPROD		0x805c
+#endif
+#ifndef ETHERTYPE_ATALK
+#define ETHERTYPE_ATALK		0x809b
+#endif
+#ifndef ETHERTYPE_AARP
+#define ETHERTYPE_AARP		0x80f3
+#endif
+#ifndef	ETHERTYPE_8021Q
+#define	ETHERTYPE_8021Q		0x8100
+#endif
+#ifndef ETHERTYPE_IPX
+#define ETHERTYPE_IPX		0x8137
+#endif
+#ifndef ETHERTYPE_IPV6
+#define ETHERTYPE_IPV6		0x86dd
+#endif
+#ifndef ETHERTYPE_PPP
+#define	ETHERTYPE_PPP		0x880b
+#endif
+#ifndef	ETHERTYPE_MPLS
+#define	ETHERTYPE_MPLS		0x8847
+#endif
+#ifndef	ETHERTYPE_MPLS_MULTI
+#define	ETHERTYPE_MPLS_MULTI	0x8848
+#endif
+#ifndef ETHERTYPE_PPPOED
+#define ETHERTYPE_PPPOED	0x8863
+#endif
+#ifndef ETHERTYPE_PPPOES
+#define ETHERTYPE_PPPOES	0x8864
+#endif
+#ifndef	ETHERTYPE_LOOPBACK
+#define	ETHERTYPE_LOOPBACK	0x9000
+#endif

aux/adtrace/ip.h

@@ -0,0 +1,159 @@
+/* @(#) $Header$ (LBL) */
+/*
+ * Copyright (c) 1982, 1986, 1993
+ *	The Regents of the University of California.  All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. All advertising materials mentioning features or use of this software
+ *    must display the following acknowledgement:
+ *	This product includes software developed by the University of
+ *	California, Berkeley and its contributors.
+ * 4. Neither the name of the University nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
+ * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ *
+ *	@(#)ip.h	8.2 (Berkeley) 6/1/94
+ */
+
+/*
+ * Definitions for internet protocol version 4.
+ * Per RFC 791, September 1981.
+ */
+#define	IPVERSION	4
+
+/*
+ * Structure of an internet header, naked of options.
+ *
+ * We declare ip_len and ip_off to be short, rather than u_short
+ * pragmatically since otherwise unsigned comparisons can result
+ * against negative integers quite easily, and fail in subtle ways.
+ */
+struct ip {
+	u_int8_t	ip_vhl;		/* header length, version */
+#define IP_V(ip)	(((ip)->ip_vhl & 0xf0) >> 4)
+#define IP_HL(ip)	((ip)->ip_vhl & 0x0f)
+	u_int8_t	ip_tos;		/* type of service */
+	u_int16_t	ip_len;		/* total length */
+	u_int16_t	ip_id;		/* identification */
+	u_int16_t	ip_off;		/* fragment offset field */
+#define	IP_DF 0x4000			/* dont fragment flag */
+#define	IP_MF 0x2000			/* more fragments flag */
+#define	IP_OFFMASK 0x1fff		/* mask for fragmenting bits */
+	u_int8_t	ip_ttl;		/* time to live */
+	u_int8_t	ip_p;		/* protocol */
+	u_int16_t	ip_sum;		/* checksum */
+	struct	in_addr ip_src,ip_dst;	/* source and dest address */
+};
+
+#define	IP_MAXPACKET	65535		/* maximum packet size */
+
+/*
+ * Definitions for IP type of service (ip_tos)
+ */
+#define	IPTOS_LOWDELAY		0x10
+#define	IPTOS_THROUGHPUT	0x08
+#define	IPTOS_RELIABILITY	0x04
+
+/*
+ * Definitions for IP precedence (also in ip_tos) (hopefully unused)
+ */
+#define	IPTOS_PREC_NETCONTROL		0xe0
+#define	IPTOS_PREC_INTERNETCONTROL	0xc0
+#define	IPTOS_PREC_CRITIC_ECP		0xa0
+#define	IPTOS_PREC_FLASHOVERRIDE	0x80
+#define	IPTOS_PREC_FLASH		0x60
+#define	IPTOS_PREC_IMMEDIATE		0x40
+#define	IPTOS_PREC_PRIORITY		0x20
+#define	IPTOS_PREC_ROUTINE		0x00
+
+/*
+ * Definitions for options.
+ */
+#define	IPOPT_COPIED(o)		((o)&0x80)
+#define	IPOPT_CLASS(o)		((o)&0x60)
+#define	IPOPT_NUMBER(o)		((o)&0x1f)
+
+#define	IPOPT_CONTROL		0x00
+#define	IPOPT_RESERVED1		0x20
+#define	IPOPT_DEBMEAS		0x40
+#define	IPOPT_RESERVED2		0x60
+
+#define	IPOPT_EOL		0		/* end of option list */
+#define	IPOPT_NOP		1		/* no operation */
+
+#define	IPOPT_RR		7		/* record packet route */
+#define	IPOPT_TS		68		/* timestamp */
+#define	IPOPT_SECURITY		130		/* provide s,c,h,tcc */
+#define	IPOPT_LSRR		131		/* loose source route */
+#define	IPOPT_SATID		136		/* satnet id */
+#define	IPOPT_SSRR		137		/* strict source route */
+
+/*
+ * Offsets to fields in options other than EOL and NOP.
+ */
+#define	IPOPT_OPTVAL		0		/* option ID */
+#define	IPOPT_OLEN		1		/* option length */
+#define IPOPT_OFFSET		2		/* offset within option */
+#define	IPOPT_MINOFF		4		/* min value of above */
+
+/*
+ * Time stamp option structure.
+ */
+struct	ip_timestamp {
+	u_int8_t	ipt_code;	/* IPOPT_TS */
+	u_int8_t	ipt_len;	/* size of structure (variable) */
+	u_int8_t	ipt_ptr;	/* index of current entry */
+	u_int8_t	ipt_oflwflg;	/* flags, overflow counter */
+#define IPTS_OFLW(ip)	(((ipt)->ipt_oflwflg & 0xf0) >> 4)
+#define IPTS_FLG(ip)	((ipt)->ipt_oflwflg & 0x0f)
+	union ipt_timestamp {
+		u_int32_t ipt_time[1];
+		struct	ipt_ta {
+			struct in_addr ipt_addr;
+			u_int32_t ipt_time;
+		} ipt_ta[1];
+	} ipt_timestamp;
+};
+
+/* flag bits for ipt_flg */
+#define	IPOPT_TS_TSONLY		0		/* timestamps only */
+#define	IPOPT_TS_TSANDADDR	1		/* timestamps and addresses */
+#define	IPOPT_TS_PRESPEC	3		/* specified modules only */
+
+/* bits for security (not byte swapped) */
+#define	IPOPT_SECUR_UNCLASS	0x0000
+#define	IPOPT_SECUR_CONFID	0xf135
+#define	IPOPT_SECUR_EFTO	0x789a
+#define	IPOPT_SECUR_MMMM	0xbc4d
+#define	IPOPT_SECUR_RESTR	0xaf13
+#define	IPOPT_SECUR_SECRET	0xd788
+#define	IPOPT_SECUR_TOPSECRET	0x6bc5
+
+/*
+ * Internet implementation parameters.
+ */
+#define	MAXTTL		255		/* maximum time to live (seconds) */
+#define	IPDEFTTL	64		/* default ttl, from RFC 1340 */
+#define	IPFRAGTTL	60		/* time to live for frags, slowhz */
+#define	IPTTLDEC	1		/* subtracted when forwarding */
+
+#define	IP_MSS		576		/* default maximum segment size */

aux/bdcat/Makefile.am

@@ -0,0 +1,4 @@
+## Process this file with automake to produce Makefile.in
+
+noinst_PROGRAMS = bdcat
+bdcat_SOURCES = bdcat.cc

aux/bdcat/bdcat.cc

@@ -0,0 +1,175 @@
+// $Id: bdcat.cc 6 2004-04-30 00:31:26Z jason $
+//
+// Decrypts Bro's log files.
+//
+// Usage: bdcat [-k file-with-secret-rsa-key] [files...]
+//
+// The key file may be alternatively set via the env variable BDCAT_KEY.
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <unistd.h>
+#include <netinet/in.h>
+
+#include "openssl/evp.h"
+#include "openssl/pem.h"
+#include "openssl/err.h"
+
+EVP_PKEY* SecKey = 0;
+EVP_CIPHER* CipherType = 0;
+
+void cryptcat(FILE* f)
+	{
+	unsigned char magic[7];
+	unsigned long secret_len;
+
+	// Read file header.
+	if ( ! (fread(&magic, 7, 1, f) &&
+		fread(&secret_len, sizeof(secret_len), 1, f)) )
+		{
+		fprintf(stderr, "can't read file header: %s\n", strerror(errno));
+		exit(1);
+		}
+
+	if ( memcmp("BROENC1", (const char*) magic, 7) != 0 )
+		{
+		fputs("not a Bro encrypted file\n", stderr);
+		exit(1);
+		}
+
+	secret_len = ntohl(secret_len);
+	int iv_len = EVP_CIPHER_iv_length(CipherType);
+	unsigned char secret[secret_len];
+	unsigned char iv[iv_len];
+
+	if ( ! (fread(&secret, secret_len, 1, f) &&
+		fread(&iv, iv_len, 1, f)) )
+		{
+		fprintf(stderr, "can't read file header: %s\n", strerror(errno));
+		exit(1);
+		}
+
+	// Decrypt data.
+	EVP_CIPHER_CTX cipher_ctx;
+	if ( ! EVP_OpenInit(&cipher_ctx, CipherType,
+				secret, secret_len, iv, SecKey) )
+		{
+		fprintf( stderr, "can't init decryption: %s\n",
+				ERR_error_string(ERR_get_error(), 0));
+		exit(1);
+		return;
+		}
+
+	int block_size = EVP_CIPHER_block_size(CipherType);
+	unsigned char buffer_in[block_size];
+	unsigned char buffer_out[block_size];
+
+	int inl, outl;
+	while ( (inl = fread(buffer_in, 1, block_size, f)) )
+		{
+		if ( ! EVP_OpenUpdate(&cipher_ctx, buffer_out,
+					&outl, buffer_in, inl) )
+			{
+			fprintf( stderr, "can't decrypt: %s\n",
+				 ERR_error_string(ERR_get_error(), 0));
+			exit(1);
+			}
+
+		if ( outl && ! fwrite(buffer_out, outl, 1, stdout) )
+			{
+			fprintf(stderr, "can't write to stdout: %s\n",
+					strerror(errno));
+			exit(1);
+			}
+		}
+
+	if ( ! EVP_OpenFinal(&cipher_ctx, buffer_out, &outl) )
+		{
+		fprintf( stderr, "can't decrypt: %s\n",
+				 ERR_error_string(ERR_get_error(), 0));
+		exit(1);
+		}
+
+	if ( outl && ! fwrite(buffer_out, outl, 1, stdout) )
+		{
+		fprintf(stderr, "can't write to stdout: %s\n", strerror(errno));
+		exit(1);
+		}
+
+	fclose(f);
+	}
+
+void Usage()
+	{
+	fprintf(stderr, "bdcat [-k <sec-key-file>] [files]\n");
+	exit(1);
+	}
+
+int main(int argc, char** argv)
+	{
+	char* keyfile = getenv("BDCAT_KEY");
+
+	// Read options.
+	char op;
+	while ( (op = getopt(argc, argv, "k:")) >= 0 )
+		{
+		if ( op == 'k' )
+			keyfile = optarg;
+		else
+			Usage();
+		}
+
+	if ( ! keyfile )
+		{
+		fputs("no keyfile given\n", stderr);
+		exit(1);
+		}
+
+	// Init crypto.
+
+	ERR_load_crypto_strings();
+	OpenSSL_add_all_algorithms();
+
+	FILE* f = fopen(keyfile, "r");
+	if ( ! f )
+		{
+		fprintf(stderr, "can't open key file %s: %s\n",
+				keyfile, strerror(errno));
+		exit(1);
+		}
+
+	SecKey = PEM_read_PrivateKey(f, 0, 0, 0);
+	if ( ! SecKey )
+		{
+		fprintf(stderr, "can't read key from %s: %s\n", keyfile,
+				ERR_error_string(ERR_get_error(), 0));
+		exit(1);
+		}
+
+	fclose(f);
+
+	// Depending on the OpenSSL version, EVP_*_cbc()
+	// returns a const or a non-const.
+	CipherType = (EVP_CIPHER*) EVP_bf_cbc();
+
+	// Decrypt the files.
+	if ( optind == argc )
+		cryptcat(stdin);
+	else
+		{
+		while ( optind < argc )
+			{
+			FILE* f = fopen(argv[optind], "r");
+			if ( ! f )
+				{
+				fprintf(stderr, "can't open %s: %s\n",
+						argv[optind], strerror(errno));
+				exit(1);
+				}
+
+			cryptcat(f);
+			++optind;
+			}
+		}
+	}

aux/binpac/COPYING

@@ -0,0 +1,18 @@
+// Copyright (c) 1995-2007
+//      The Regents of the University of California.  All rights reserved.
+//
+// Redistribution and use in source and binary forms, with or without
+// modification, are permitted provided that: (1) source code distributions
+// retain the above copyright notice and this paragraph in its entirety, (2)
+// distributions including binary code include the above copyright notice and
+// this paragraph in its entirety in the documentation or other materials
+// provided with the distribution, and (3) all advertising materials mentioning
+// features or use of this software display the following acknowledgement:
+// ``This product includes software developed by the University of California,
+// Lawrence Berkeley Laboratory and its contributors.'' Neither the name of
+// the University nor the names of its contributors may be used to endorse
+// or promote products derived from this software without specific prior
+// written permission.
+// THIS SOFTWARE IS PROVIDED ``AS IS'' AND WITHOUT ANY EXPRESS OR IMPLIED
+// WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED WARRANTIES OF
+// MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.

aux/binpac/INSTALL

@@ -0,0 +1,231 @@
+Installation Instructions
+*************************
+
+Copyright (C) 1994, 1995, 1996, 1999, 2000, 2001, 2002, 2004 Free
+Software Foundation, Inc.
+
+This file is free documentation; the Free Software Foundation gives
+unlimited permission to copy, distribute and modify it.
+
+Basic Installation
+==================
+
+These are generic installation instructions.
+
+   The `configure' shell script attempts to guess correct values for
+various system-dependent variables used during compilation.  It uses
+those values to create a `Makefile' in each directory of the package.
+It may also create one or more `.h' files containing system-dependent
+definitions.  Finally, it creates a shell script `config.status' that
+you can run in the future to recreate the current configuration, and a
+file `config.log' containing compiler output (useful mainly for
+debugging `configure').
+
+   It can also use an optional file (typically called `config.cache'
+and enabled with `--cache-file=config.cache' or simply `-C') that saves
+the results of its tests to speed up reconfiguring.  (Caching is
+disabled by default to prevent problems with accidental use of stale
+cache files.)
+
+   If you need to do unusual things to compile the package, please try
+to figure out how `configure' could check whether to do them, and mail
+diffs or instructions to the address given in the `README' so they can
+be considered for the next release.  If you are using the cache, and at
+some point `config.cache' contains results you don't want to keep, you
+may remove or edit it.
+
+   The file `configure.ac' (or `configure.in') is used to create
+`configure' by a program called `autoconf'.  You only need
+`configure.ac' if you want to change it or regenerate `configure' using
+a newer version of `autoconf'.
+
+The simplest way to compile this package is:
+
+  1. `cd' to the directory containing the package's source code and type
+     `./configure' to configure the package for your system.  If you're
+     using `csh' on an old version of System V, you might need to type
+     `sh ./configure' instead to prevent `csh' from trying to execute
+     `configure' itself.
+
+     Running `configure' takes awhile.  While running, it prints some
+     messages telling which features it is checking for.
+
+  2. Type `make' to compile the package.
+
+  3. Optionally, type `make check' to run any self-tests that come with
+     the package.
+
+  4. Type `make install' to install the programs and any data files and
+     documentation.
+
+  5. You can remove the program binaries and object files from the
+     source code directory by typing `make clean'.  To also remove the
+     files that `configure' created (so you can compile the package for
+     a different kind of computer), type `make distclean'.  There is
+     also a `make maintainer-clean' target, but that is intended mainly
+     for the package's developers.  If you use it, you may have to get
+     all sorts of other programs in order to regenerate files that came
+     with the distribution.
+
+Compilers and Options
+=====================
+
+Some systems require unusual options for compilation or linking that the
+`configure' script does not know about.  Run `./configure --help' for
+details on some of the pertinent environment variables.
+
+   You can give `configure' initial values for configuration parameters
+by setting variables in the command line or in the environment.  Here
+is an example:
+
+     ./configure CC=c89 CFLAGS=-O2 LIBS=-lposix
+
+   *Note Defining Variables::, for more details.
+
+Compiling For Multiple Architectures
+====================================
+
+You can compile the package for more than one kind of computer at the
+same time, by placing the object files for each architecture in their
+own directory.  To do this, you must use a version of `make' that
+supports the `VPATH' variable, such as GNU `make'.  `cd' to the
+directory where you want the object files and executables to go and run
+the `configure' script.  `configure' automatically checks for the
+source code in the directory that `configure' is in and in `..'.
+
+   If you have to use a `make' that does not support the `VPATH'
+variable, you have to compile the package for one architecture at a
+time in the source code directory.  After you have installed the
+package for one architecture, use `make distclean' before reconfiguring
+for another architecture.
+
+Installation Names
+==================
+
+By default, `make install' will install the package's files in
+`/usr/local/bin', `/usr/local/man', etc.  You can specify an
+installation prefix other than `/usr/local' by giving `configure' the
+option `--prefix=PREFIX'.
+
+   You can specify separate installation prefixes for
+architecture-specific files and architecture-independent files.  If you
+give `configure' the option `--exec-prefix=PREFIX', the package will
+use PREFIX as the prefix for installing programs and libraries.
+Documentation and other data files will still use the regular prefix.
+
+   In addition, if you use an unusual directory layout you can give
+options like `--bindir=DIR' to specify different values for particular
+kinds of files.  Run `configure --help' for a list of the directories
+you can set and what kinds of files go in them.
+
+   If the package supports it, you can cause programs to be installed
+with an extra prefix or suffix on their names by giving `configure' the
+option `--program-prefix=PREFIX' or `--program-suffix=SUFFIX'.
+
+Optional Features
+=================
+
+Some packages pay attention to `--enable-FEATURE' options to
+`configure', where FEATURE indicates an optional part of the package.
+They may also pay attention to `--with-PACKAGE' options, where PACKAGE
+is something like `gnu-as' or `x' (for the X Window System).  The
+`README' should mention any `--enable-' and `--with-' options that the
+package recognizes.
+
+   For packages that use the X Window System, `configure' can usually
+find the X include and library files automatically, but if it doesn't,
+you can use the `configure' options `--x-includes=DIR' and
+`--x-libraries=DIR' to specify their locations.
+
+Specifying the System Type
+==========================
+
+There may be some features `configure' cannot figure out automatically,
+but needs to determine by the type of machine the package will run on.
+Usually, assuming the package is built to be run on the _same_
+architectures, `configure' can figure that out, but if it prints a
+message saying it cannot guess the machine type, give it the
+`--build=TYPE' option.  TYPE can either be a short name for the system
+type, such as `sun4', or a canonical name which has the form:
+
+     CPU-COMPANY-SYSTEM
+
+where SYSTEM can have one of these forms:
+
+     OS KERNEL-OS
+
+   See the file `config.sub' for the possible values of each field.  If
+`config.sub' isn't included in this package, then this package doesn't
+need to know the machine type.
+
+   If you are _building_ compiler tools for cross-compiling, you should
+use the `--target=TYPE' option to select the type of system they will
+produce code for.
+
+   If you want to _use_ a cross compiler, that generates code for a
+platform different from the build platform, you should specify the
+"host" platform (i.e., that on which the generated programs will
+eventually be run) with `--host=TYPE'.
+
+Sharing Defaults
+================
+
+If you want to set default values for `configure' scripts to share, you
+can create a site shell script called `config.site' that gives default
+values for variables like `CC', `cache_file', and `prefix'.
+`configure' looks for `PREFIX/share/config.site' if it exists, then
+`PREFIX/etc/config.site' if it exists.  Or, you can set the
+`CONFIG_SITE' environment variable to the location of the site script.
+A warning: not all `configure' scripts look for a site script.
+
+Defining Variables
+==================
+
+Variables not defined in a site shell script can be set in the
+environment passed to `configure'.  However, some packages may run
+configure again during the build, and the customized values of these
+variables may be lost.  In order to avoid this problem, you should set
+them in the `configure' command line, using `VAR=value'.  For example:
+
+     ./configure CC=/usr/local2/bin/gcc
+
+will cause the specified gcc to be used as the C compiler (unless it is
+overridden in the site shell script).
+
+`configure' Invocation
+======================
+
+`configure' recognizes the following options to control how it operates.
+
+`--help'
+`-h'
+     Print a summary of the options to `configure', and exit.
+
+`--version'
+`-V'
+     Print the version of Autoconf used to generate the `configure'
+     script, and exit.
+
+`--cache-file=FILE'
+     Enable the cache: use and save the results of the tests in FILE,
+     traditionally `config.cache'.  FILE defaults to `/dev/null' to
+     disable caching.
+
+`--config-cache'
+`-C'
+     Alias for `--cache-file=config.cache'.
+
+`--quiet'
+`--silent'
+`-q'
+     Do not print messages saying which checks are being made.  To
+     suppress all normal output, redirect it to `/dev/null' (any error
+     messages will still be shown).
+
+`--srcdir=DIR'
+     Look for the package's source code in directory DIR.  Usually
+     `configure' can determine that directory automatically.
+
+`configure' also accepts some other, not widely useful, options.  Run
+`configure --help' for more details.
+

aux/binpac/Makefile.am

@@ -0,0 +1,4 @@
+## Process this file with automake to produce Makefile.in
+
+EXTRA_DIST = README VERSION CHANGES TODO depcomp shtool autogen.sh
+SUBDIRS = lib src

aux/binpac/TODO

@@ -0,0 +1,34 @@
+Big features
+* Variable context (xid, call in RPC)? -- no variable context
+* Helpers
+* Connection states and actions
+* Case and analyzer redef
+* &also withinput
+* Explicit analyzer context (interface + instantiation) "withcontext"
++ Interface with C++ and Bro (events, extern, weird)
++ Incremental input
++ ASCII protocols
++ Reassembly
+- Dealing with exceptions
+- Dependency analysis to save parsing time on unused fields
+- Performance measurement
+
+Small features
+* Restructure the code: break up pac.{h,cc}
+* ref counting (to keep certain structures)
+* analyzer context as a parameter of class
+* &autolength
+* find a better name for "analyzer_context" ("analcxt", "context", "analyzer") $context
+* &if
+* &autolength (now &restofdata)
+* Use vector<> instead of array<>?
+* set end_of_data when &length = ...
+- make the `default' case mandatory?
+- &inline
+- &warn and &check? (follow &if)
+- typedef?
+
+Binpac 1
+- create a namespace for each .pac file
+- type equivalence
+- byteorder() for every type?

aux/binpac/VERSION

@@ -0,0 +1 @@
+0.1

aux/binpac/autogen.sh

@@ -0,0 +1,114 @@
+#!/bin/sh
+
+# Initialization script to set up the initial configuration files etc.
+# shtool usage inspired by the autogen script of the ferite scripting
+# language -- cheers Chris :)
+#
+# This is 'borrowed' from netdude, with minor changes for bro 
+
+BLD_ON=`./shtool echo -n -e %B`
+BLD_OFF=`./shtool echo -n -e %b`
+
+srcdir=`dirname $0`
+NAME=binpac
+
+DIE=0
+
+echo
+echo "              "${BLD_ON}"Binpac Build Tools Setup"${BLD_OFF}
+echo "===================================================="
+echo
+echo "Checking whether we have all tools available ..."
+
+(autoconf --version) < /dev/null > /dev/null 2>&1 || {
+  echo
+  echo ${BLD_ON}"Error"${BLD_OFF}": You must have \`autoconf' installed to."
+  echo "Download the appropriate package for your distribution,"
+  echo "or get the source tarball at ftp://ftp.gnu.org/pub/gnu/"
+  DIE=1
+}
+
+(automake --version) < /dev/null > /dev/null 2>&1 || {
+  echo
+  echo ${BLD_ON}"Error"${BLD_OFF}": You must have \`automake' installed."
+  echo "Get ftp://ftp.gnu.org/pub/gnu/automake-1.3.tar.gz"
+  echo "(or a newer version if it is available)"
+  DIE=1
+  NO_AUTOMAKE=yes
+}
+
+# if no automake, don't bother testing for aclocal
+test -n "$NO_AUTOMAKE" || (aclocal --version) < /dev/null > /dev/null 2>&1 || {
+  echo
+  echo ${BLD_ON}"Error"${BLD_OFF}": Missing \`aclocal'.  The version of \`automake'"
+  echo "installed doesn't appear recent enough."
+  echo "Get ftp://ftp.gnu.org/pub/gnu/automake-1.3.tar.gz"
+  echo "(or a newer version if it is available)"
+  DIE=1
+}
+
+if test "$DIE" -eq 1; then
+  exit 1
+fi
+
+echo "All necessary tools found."
+echo
+
+if [ -d autom4te.cache ] ; then
+    echo "Removing autom4te.cache ..."
+    rm -rf autom4te.cache
+    #echo
+    #echo ${BLD_ON}"Error"${BLD_OFF}": autom4te.cache directory exists"
+    #echo "please remove it, and rerun this script"
+    #echo 
+    #exit 1
+fi
+
+echo
+echo "running "${BLD_ON}"aclocal"${BLD_OFF}
+echo "----------------------------------------------------"
+aclocal -I . $ACLOCAL_FLAGS
+if [ $? -ne 0 ]; then
+    echo "*** ERROR($NAME), aborting."
+    exit 1
+fi
+
+echo
+echo "running "${BLD_ON}"autoheader"${BLD_OFF}
+echo "----------------------------------------------------"
+autoheader
+if [ $? -ne 0 ]; then
+    echo "*** ERROR($NAME), aborting."
+    exit 1
+fi
+
+echo
+echo "running "${BLD_ON}"automake"${BLD_OFF}
+echo "----------------------------------------------------"
+automake -a -c 
+if [ $? -ne 0 ]; then
+    echo "*** ERROR($NAME), aborting."
+    exit 1
+fi
+
+echo
+echo "running "${BLD_ON}"autoconf"${BLD_OFF}
+echo "----------------------------------------------------"
+autoconf
+if [ $? -ne 0 ]; then
+    echo "*** ERROR($NAME), aborting."
+    exit 1
+fi
+
+if ! test "x$BROBUILD" = xyes; then
+echo 
+echo "Setup finished. Now run:"
+echo
+echo "  $ "${BLD_ON}"./configure"${BLD_OFF}" (with options as needed, try --help)"
+echo
+echo "and then"
+echo
+echo "  $ "${BLD_ON}"make"${BLD_OFF}
+echo "  # "${BLD_ON}"make install"${BLD_OFF}
+echo
+fi

aux/binpac/configure.in

@@ -0,0 +1,55 @@
+AC_INIT
+AC_CONFIG_SRCDIR([src/pac_main.cc])
+
+AC_CANONICAL_SYSTEM
+
+AC_CONFIG_AUX_DIR(.)
+AM_CONFIG_HEADER(config.h)
+AM_INIT_AUTOMAKE(binpac,  esyscmd([tr -d '\n' < VERSION]))
+
+dnl Commands for funkier shell output:
+BLD_ON=`./shtool echo -n -e %B`
+BLD_OFF=`./shtool echo -n -e %b`
+
+dnl ################################################
+dnl # Checks for programs
+dnl ################################################
+AM_PROG_LEX
+AC_PROG_YACC
+AC_PROG_CXX
+AC_PROG_RANLIB
+AC_PROG_INSTALL
+
+m4_ifdef([AC_COMPUTE_INT], [], [AC_DEFUN([AC_COMPUTE_INT], [_AC_COMPUTE_INT([$2],[$1],[$3],[$4])])])
+
+AC_COMPUTE_INT([SIZEOF_UNSIGNED_INT], [sizeof(unsigned int)])
+AC_SUBST(SIZEOF_UNSIGNED_INT)
+
+AC_ARG_ENABLE(debug,
+    [  --enable-debug          no compiler optimizations],
+    debug="yes"
+    CFLAGS="-DDEBUG `echo $CFLAGS | sed -e 's/-O2//'`"
+    CXXFLAGS="-DDEBUG `echo $CXXFLAGS | sed -e 's/-O2//'`",
+    debug="no")
+
+AC_C_BIGENDIAN(
+	AC_DEFINE(WORDS_BIGENDIAN,1,[whether words are stored with the most significant byte first])
+	dnl This is intentionally named differently so as to not collide with WORDS_BIGENDIAN
+	HOST_BIGENDIAN="#define HOST_BIGENDIAN 1"
+	AC_SUBST(HOST_BIGENDIAN))
+
+AC_CONFIG_FILES([
+Makefile 
+src/Makefile 
+lib/Makefile
+lib/binpac.h
+])
+AC_OUTPUT
+
+echo
+echo "                "${BLD_ON}"Binpac Configuration Summary"${BLD_OFF}
+echo "=========================================================="
+echo
+echo "  - Debugging enabled:      "${BLD_ON}$debug${BLD_OFF}
+echo
+exit 0

aux/binpac/lib/Makefile.am

@@ -0,0 +1,8 @@
+## Process this file with automake to produce Makefile.in
+
+noinst_LIBRARIES = libbinpac.a
+
+libbinpac_a_SOURCES = \
+	binpac_buffer.cc binpac_bytestring.cc \
+	binpac.h binpac_analyzer.h binpac_buffer.h \
+	binpac_bytestring.h binpac_exception.h binpac_regex.h

aux/binpac/lib/README

@@ -0,0 +1,3 @@
+This directory contains a library needed by generated C++ code from
+binpac. Note that the library is not needed by the binpac compiler
+itself.

aux/binpac/lib/binpac.h.in

@@ -0,0 +1,142 @@
+// $Id: binpac.h,v 1.1.4.2 2006/06/02 15:13:13 rpang Exp $
+// Do not edit binpac.h, edit binpac.h.in instead!
+
+#ifndef binpac_h
+#define binpac_h
+
+#include <sys/param.h>
+
+@HOST_BIGENDIAN@
+#ifdef HOST_BIGENDIAN
+#  define HOST_BYTEORDER	bigendian
+#else
+#  define HOST_BYTEORDER	littleendian
+#endif
+
+#include <assert.h>
+#include <stdarg.h>
+#include <stdio.h>
+#include <string>
+#include <memory>
+
+#define BINPAC_ASSERT(x)	assert(x)
+
+using namespace std;
+
+namespace binpac {
+
+const int bigendian = 0;
+const int littleendian = 1;
+const int unspecified_byteorder = -1;
+
+#ifndef pac_type_defs
+#define pac_type_defs
+
+typedef char 		int8;
+typedef short 		int16;
+typedef long 		int32;
+typedef unsigned char 	uint8;
+typedef unsigned short 	uint16;
+typedef unsigned int 	uint32;
+typedef void		*nullptr;
+typedef void		*voidptr;
+typedef uint8		*byteptr;
+typedef const uint8	*const_byteptr;
+typedef const char	*const_charptr;
+
+#if @SIZEOF_UNSIGNED_INT@ != 4
+#error "unexpected size of unsigned int"
+#endif
+
+#endif /* pac_type_defs */
+
+/* Handling byte order */
+
+namespace {
+
+inline int16 pac_swap(int16 x)
+	{
+	return (x >> 8) | ((x & 0xff) << 8);
+	}
+
+inline uint16 pac_swap(uint16 x)
+	{
+	return (x >> 8) | ((x & 0xff) << 8);
+	}
+
+inline int32 pac_swap(int32 x)
+	{
+	return 	(x >> 24) | 
+		((x & 0xff0000) >> 8) | 
+		((x & 0xff00) << 8) | 
+		((x & 0xff) << 24);
+	}
+
+inline uint32 pac_swap(uint32 x)
+	{
+	return 	(x >> 24) | 
+		((x & 0xff0000) >> 8) | 
+		((x & 0xff00) << 8) | 
+		((x & 0xff) << 24);
+	}
+
+#define FixByteOrder(byteorder, x)	(byteorder == HOST_BYTEORDER ? (x) : pac_swap(x))
+
+template <class T>
+inline T UnMarshall(const u_char *data, int byteorder)
+	{
+	T result = 0;
+	for ( int i = 0; i < (int) sizeof(T); ++i )
+		result = ( result << 8 ) | 
+			data[byteorder == bigendian ? i : sizeof(T) - 1 - i];
+	return result;
+	}
+
+inline const char* do_fmt(const char* format, va_list ap)
+	{
+	static char buf[1024];
+	vsnprintf(buf, sizeof(buf), format, ap);
+	return buf;
+	}
+
+inline string strfmt(const char* format, ...)
+	{
+	va_list ap;
+	va_start(ap, format);
+	const char* r = do_fmt(format, ap);
+	va_end(ap);
+	return string(r);
+	}
+
+} // anonymous namespace
+
+#define binpac_fmt(x...) strfmt(x).c_str()
+
+class RefCount
+{
+public:
+	RefCount() 	{ count = 1; }
+	void Ref() 	{ ++count; }
+	int Unref() 	{ BINPAC_ASSERT(count > 0); return --count; }
+
+private:
+	int count;
+};
+
+namespace {
+	inline void Unref(RefCount *x)
+		{
+		if ( x && x->Unref() <= 0 )
+			delete x;
+		}
+}  // anonymous namespace
+
+} // namespace binpac
+
+#include "binpac_analyzer.h"
+#include "binpac_buffer.h"
+#include "binpac_bytestring.h"
+#include "binpac_exception.h"
+#include "binpac_regex.h"
+
+#endif /* binpac_h */

aux/binpac/lib/binpac_analyzer.h

@@ -0,0 +1,27 @@
+#ifndef binpac_an_h
+#define binpac_an_h
+
+namespace binpac {
+
+// TODO: Add the Done() function
+
+// The interface for a connection analyzer
+class ConnectionAnalyzer {
+public:
+	virtual ~ConnectionAnalyzer() {}
+	virtual void NewData(bool is_orig,
+	                     const u_char *begin_of_data, 
+	                     const u_char *end_of_data) = 0;
+};
+
+// The interface for a flow analyzer
+class FlowAnalyzer {
+public:
+	virtual ~FlowAnalyzer() {}
+	virtual void NewData(const u_char *begin_of_data, 
+	                     const u_char *end_of_data) = 0;
+};
+
+}  // namespace binpac
+
+#endif  // binpac_an_h

aux/binpac/lib/binpac_buffer.cc

@@ -0,0 +1,465 @@
+#include <stdlib.h>
+#include <stdio.h>
+#include <string.h>	// for memcpy
+
+#define binpac_regex_h
+
+#include "binpac.h"
+#include "binpac_buffer.h"
+
+namespace binpac {
+
+extern double network_time();
+
+namespace {
+	const u_char CR = '\r';
+	const u_char LF = '\n';
+}
+
+FlowBuffer::FlowBuffer(LineBreakStyle linebreak_style)
+	{
+	buffer_length_ = 0;
+	buffer_ = 0;
+
+	orig_data_begin_ = 0;
+	orig_data_end_ = 0;
+
+	linebreak_style_ = linebreak_style;
+	ResetLineState();
+
+	mode_ = UNKNOWN_MODE;
+	frame_length_ = 0;
+	chunked_ = false;
+
+	data_seq_at_orig_data_end_ = 0;
+	eof_ = false;
+	have_pending_request_ = false;
+
+	buffer_n_ = 0;
+
+	NewMessage();
+	}
+
+FlowBuffer::~FlowBuffer()
+	{
+	if ( buffer_ )
+		free(buffer_);
+	}
+
+void FlowBuffer::NewMessage()
+	{
+	BINPAC_ASSERT(frame_length_ >= 0);
+
+	int bytes_to_advance = 0;
+	if ( buffer_n_ == 0 ) 
+		{
+		switch ( mode_ ) 
+			{
+			case LINE_MODE:
+				bytes_to_advance = (frame_length_ + 
+					(linebreak_style_ == STRICT_CRLF ? 
+						2 : 1));
+				break;
+			case FRAME_MODE:
+				bytes_to_advance = frame_length_;
+				break;
+			case UNKNOWN_MODE:
+				break;
+			}
+		}
+
+	orig_data_begin_ += bytes_to_advance;
+	BINPAC_ASSERT(orig_data_begin_ <= orig_data_end_);
+
+	buffer_n_ = 0;
+	message_complete_ = false;
+	}
+
+void FlowBuffer::ResetLineState()
+	{
+	switch ( linebreak_style_ )
+		{
+		case CR_OR_LF:
+			state_ = CR_OR_LF_0;
+			break;
+		case STRICT_CRLF:
+			state_ = STRICT_CRLF_0;
+			break;
+		default:
+			BINPAC_ASSERT(0);
+			break;
+		}
+	}
+
+void FlowBuffer::ExpandBuffer(int length)
+	{
+	if ( buffer_length_ >= length )
+		return;
+	// So length > 0
+	if ( length < 512 )
+		length = 512;
+	
+	if ( length < buffer_length_ * 2 )
+		length = buffer_length_ * 2;
+
+	// Allocate a new buffer and copy the existing contents
+	buffer_length_ = length;
+	u_char *new_buf = (u_char *) realloc(buffer_, buffer_length_);
+	BINPAC_ASSERT(new_buf);
+#if 0
+	u_char* new_buf = new u_char[buffer_length_];
+	if ( buffer_ && buffer_n_ > 0 )
+		memcpy(new_buf, buffer_, buffer_n_);
+	delete [] buffer_;
+#endif
+	buffer_ = new_buf;
+	}
+
+void FlowBuffer::NewLine()
+	{
+	FlowBuffer::NewMessage();
+	mode_ = LINE_MODE;
+	frame_length_ = 0;
+	chunked_ = false;
+	have_pending_request_ = true;
+	if ( state_ == FRAME_0 )
+		ResetLineState();
+	MarkOrCopyLine();
+	}
+
+void FlowBuffer::NewFrame(int frame_length, bool chunked)
+	{
+	FlowBuffer::NewMessage();
+	mode_ = FRAME_MODE;
+	frame_length_ = frame_length;
+	chunked_ = chunked;
+	have_pending_request_ = true;
+	MarkOrCopyFrame();
+	}
+
+void FlowBuffer::GrowFrame(int length)
+	{
+	BINPAC_ASSERT(frame_length_ >= 0);
+	if ( length <= frame_length_ )
+		return;
+	BINPAC_ASSERT(! chunked_ || frame_length_ == 0);
+	mode_ = FRAME_MODE;
+	frame_length_ = length;
+	MarkOrCopyFrame();
+	}
+
+void FlowBuffer::DiscardData()
+	{
+	mode_ = UNKNOWN_MODE;
+	message_complete_ = false;
+	have_pending_request_ = false;
+	orig_data_begin_ = orig_data_end_ = 0;
+
+	buffer_n_ = 0;
+	frame_length_ = 0;
+	}
+
+void FlowBuffer::set_eof()
+	{
+	// fprintf(stderr, "EOF\n");
+	eof_ = true;
+	if ( chunked_ )
+		frame_length_ = orig_data_end_ - orig_data_begin_;
+	if ( frame_length_ < 0 )
+		frame_length_ = 0;
+	}
+
+void FlowBuffer::NewData(const_byteptr begin, const_byteptr end)
+	{
+	BINPAC_ASSERT(begin <= end);
+
+	ClearPreviousData();
+
+	BINPAC_ASSERT((buffer_n_ == 0 && message_complete_) || 
+		orig_data_begin_ == orig_data_end_);
+
+	orig_data_begin_ = begin;
+	orig_data_end_ = end;
+	data_seq_at_orig_data_end_ += (end - begin);
+
+	MarkOrCopy();
+	}
+
+void FlowBuffer::MarkOrCopy()
+	{
+	if ( ! message_complete_ )
+		{
+		switch ( mode_ ) 
+			{
+			case LINE_MODE:
+				MarkOrCopyLine();
+				break;
+
+			case FRAME_MODE:
+				MarkOrCopyFrame();
+				break;
+		
+			default:
+				break;
+			}
+		}
+	}
+
+void FlowBuffer::ClearPreviousData()
+	{
+	// All previous data must have been processed or buffered already
+	if ( orig_data_begin_ < orig_data_end_ )
+		{
+		BINPAC_ASSERT(buffer_n_ == 0);
+		if ( chunked_ )
+			{
+			if ( frame_length_ > 0 )
+				{
+				frame_length_ -= 
+					(orig_data_end_ - orig_data_begin_);
+				}
+			orig_data_begin_ = orig_data_end_;
+			}
+		}
+	}
+
+void FlowBuffer::NewGap(int length)
+	{
+	ClearPreviousData();
+
+	if ( chunked_ && frame_length_ >= 0 )
+		{
+		frame_length_ -= length;
+		if ( frame_length_ < 0 )
+			frame_length_ = 0;
+		}
+
+	orig_data_begin_ = orig_data_end_ = 0;
+	MarkOrCopy();
+	}
+
+void FlowBuffer::MarkOrCopyLine()
+	{
+	switch ( linebreak_style_ )
+		{
+		case CR_OR_LF:
+			MarkOrCopyLine_CR_OR_LF();
+			break;
+		case STRICT_CRLF:
+			MarkOrCopyLine_STRICT_CRLF();
+			break;
+		default:
+			BINPAC_ASSERT(0);
+			break;
+		}
+	}
+
+/*
+Finite state automaton for CR_OR_LF:
+(!--line is complete, *--add to buffer)
+
+CR_OR_LF_0:
+	CR:	CR_OR_LF_1 !
+	LF:	CR_OR_LF_0 !
+	.:	CR_OR_LF_0 *
+
+CR_OR_LF_1:
+	CR:	CR_OR_LF_1 !
+	LF:	CR_OR_LF_0
+	.:	CR_OR_LF_0 *
+*/
+
+void FlowBuffer::MarkOrCopyLine_CR_OR_LF()
+	{
+	if ( state_ == CR_OR_LF_1 && 
+	     orig_data_begin_ < orig_data_end_ && *orig_data_begin_ == LF )
+		{
+		state_ = CR_OR_LF_0;
+		++orig_data_begin_;
+		}
+
+	const_byteptr data;
+	for ( data = orig_data_begin_; data < orig_data_end_; ++data )
+		{
+		switch ( *data )
+			{
+			case CR:
+				state_ = CR_OR_LF_1;
+				goto found_end_of_line;
+
+			case LF:
+				// state_ = CR_OR_LF_0;
+				goto found_end_of_line;
+
+			default:
+				// state_ = CR_OR_LF_0;
+				break;
+			}
+		}
+
+	AppendToBuffer(orig_data_begin_, orig_data_end_ - orig_data_begin_);
+	return;
+
+found_end_of_line:
+	if ( buffer_n_ == 0 )
+		{
+		frame_length_ = data - orig_data_begin_;
+		}
+	else
+		{
+		AppendToBuffer(orig_data_begin_, data + 1 - orig_data_begin_);
+		// But eliminate the last CR or LF
+		--buffer_n_;
+		}
+	message_complete_ = true;
+
+#if DEBUG_FLOW_BUFFER
+	fprintf(stderr, "%.6f Line complete: [%s]\n",
+		network_time(),
+		string((const char *) begin(), (const char *) end()).c_str());
+#endif
+	}
+
+/*
+Finite state automaton and STRICT_CRLF:
+(!--line is complete, *--add to buffer)
+
+STRICT_CRLF_0:
+	CR:	STRICT_CRLF_1 *
+	LF:	STRICT_CRLF_0 *
+	.:	STRICT_CRLF_0 *
+
+STRICT_CRLF_1:
+	CR:	STRICT_CRLF_1 *
+	LF:	STRICT_CRLF_0 ! (--buffer_n_)
+	.:	STRICT_CRLF_0 *
+*/
+
+void FlowBuffer::MarkOrCopyLine_STRICT_CRLF()
+	{
+	const_byteptr data;
+	for ( data = orig_data_begin_; data < orig_data_end_; ++data )
+		{
+		switch ( *data )
+			{
+			case CR:
+				state_ = STRICT_CRLF_1;
+				break;
+
+			case LF:
+				if ( state_ == STRICT_CRLF_1 )
+					{
+					state_ = STRICT_CRLF_0;
+					goto found_end_of_line;
+					}
+				break;
+
+			default:
+				state_ = STRICT_CRLF_0;
+				break;
+			}
+		}
+
+	AppendToBuffer(orig_data_begin_, orig_data_end_ - orig_data_begin_);
+	return;
+
+found_end_of_line:
+	if ( buffer_n_ == 0 )
+		{
+		frame_length_ = data - 1 - orig_data_begin_;
+		}
+	else
+		{
+		AppendToBuffer(orig_data_begin_, data + 1 - orig_data_begin_);
+		// Pop the preceding CR and LF from the buffer
+		buffer_n_ -= 2;
+		}
+
+	message_complete_ = true;
+
+#if DEBUG_FLOW_BUFFER
+	fprintf(stderr, "%.6f Line complete: [%s]\n",
+		network_time(),
+		string((const char *) begin(), (const char *) end()).c_str());
+#endif
+	}
+
+// Invariants:
+// 
+// When buffer_n_ == 0:
+// Frame = [orig_data_begin_..(orig_data_begin_ + frame_length_)]
+// 
+// When buffer_n_ > 0:
+// Frame = [0..buffer_n_][orig_data_begin_..]
+
+void FlowBuffer::MarkOrCopyFrame()
+	{
+	if ( mode_ == FRAME_MODE && state_ == CR_OR_LF_1 && 
+	     orig_data_begin_ < orig_data_end_ )
+		{
+		// Skip the lingering LF
+		if ( *orig_data_begin_ == LF )
+			{
+			++orig_data_begin_;
+			}
+		state_ = FRAME_0;
+		}
+
+	if ( buffer_n_ == 0 )
+		{
+		// If there is enough data
+		if ( frame_length_ >= 0 &&
+		     orig_data_end_ - orig_data_begin_ >= frame_length_ )
+			{
+			// Do nothing except setting the message complete flag
+			message_complete_ = true;
+			}
+		else 
+			{
+			if ( ! chunked_ )
+				{
+				AppendToBuffer(orig_data_begin_, 
+					orig_data_end_ - orig_data_begin_);
+				}
+			message_complete_ = false;
+			}
+		}
+	else
+		{
+		BINPAC_ASSERT(!chunked_);
+		int bytes_to_copy = orig_data_end_ - orig_data_begin_;
+		message_complete_ = false;
+		if ( frame_length_ >= 0 && buffer_n_ + bytes_to_copy >= frame_length_ ) 
+			{
+			bytes_to_copy = frame_length_ - buffer_n_;
+			message_complete_ = true;
+			}
+		AppendToBuffer(orig_data_begin_, bytes_to_copy);
+		}
+
+#if DEBUG_FLOW_BUFFER
+	if ( message_complete_ )
+		{
+		fprintf(stderr, "%.6f frame complete: [%s]\n",
+			network_time(),
+			string((const char *) begin(), 
+				(const char *) end()).c_str());
+		}
+#endif
+	}
+
+void FlowBuffer::AppendToBuffer(const_byteptr data, int len)
+	{
+	if ( len <= 0 )
+		return;
+
+	BINPAC_ASSERT(! chunked_);
+	ExpandBuffer(buffer_n_ + len);
+	memcpy(buffer_ + buffer_n_, data, len);
+	buffer_n_ += len;
+
+	orig_data_begin_ += len;
+	BINPAC_ASSERT(orig_data_begin_ <= orig_data_end_);
+	}
+
+}  // namespace binpac

aux/binpac/lib/binpac_buffer.h

@@ -0,0 +1,148 @@
+#ifndef binpac_buffer_h
+#define binpac_buffer_h
+
+#include <sys/types.h>
+#include "binpac.h"
+
+namespace binpac {
+
+class FlowBuffer {
+public:
+	enum LineBreakStyle { 
+		CR_OR_LF, 	// CR or LF or CRLF
+		STRICT_CRLF, 	// CR followed by LF
+		CR_LF_NUL,	// CR or LF or CR-LF or CR-NUL
+	};
+
+	FlowBuffer(LineBreakStyle linebreak_style = CR_OR_LF);
+	virtual ~FlowBuffer();
+
+	void NewData(const_byteptr begin, const_byteptr end);
+	void NewGap(int length);
+
+	// Discard unprocessed data
+	void DiscardData();
+
+	// Whether there is enough data for the frame
+	bool ready() const{ return message_complete_ || mode_ == UNKNOWN_MODE; }
+
+	inline const_byteptr begin() const
+		{
+		BINPAC_ASSERT(ready());
+		return ( buffer_n_ == 0 ) ? 
+			orig_data_begin_ : buffer_;
+		}
+
+	inline const_byteptr end() const
+		{
+		BINPAC_ASSERT(ready());
+		if ( buffer_n_ == 0 )
+			{
+			BINPAC_ASSERT(frame_length_ >= 0);
+			const_byteptr end = orig_data_begin_ + frame_length_;
+			BINPAC_ASSERT(end <= orig_data_end_);
+			return end;
+			}
+		else
+			return buffer_ + buffer_n_;
+		}
+
+	inline int data_length() const
+		{
+		if ( buffer_n_ > 0 ) 
+			return buffer_n_;
+
+		if ( frame_length_ < 0 ||
+		     orig_data_begin_ + frame_length_ > orig_data_end_ )
+			return orig_data_end_ - orig_data_begin_;
+		else
+			return frame_length_;
+		}
+
+	inline bool data_available() const
+	        {
+		return buffer_n_ > 0 || orig_data_end_ > orig_data_begin_;
+		}
+
+	void NewLine();
+	// A negative frame_length represents a frame till EOF
+	void NewFrame(int frame_length, bool chunked_);
+	void GrowFrame(int new_frame_length);
+
+	int data_seq() const
+		{ 
+		int data_seq_at_orig_data_begin = 
+			data_seq_at_orig_data_end_ - 
+			(orig_data_end_ - orig_data_begin_);
+		if ( buffer_n_ > 0 )
+			return data_seq_at_orig_data_begin;
+		else
+			return data_seq_at_orig_data_begin + data_length(); 
+		}
+	bool eof() const	{ return eof_; }
+	void set_eof();
+
+	bool have_pending_request() const { return have_pending_request_; }
+
+protected:
+	// Reset the buffer for a new message
+	void NewMessage();
+
+	void ClearPreviousData();
+
+	// Expand the buffer to at least <length> bytes. If there 
+	// are contents in the existing buffer, copy them to the new
+	// buffer.
+	void ExpandBuffer(int length);
+
+	// Reset line state when transit from frame mode to line mode.
+	void ResetLineState();
+
+	void AppendToBuffer(const_byteptr data, int len);
+
+	// MarkOrCopy{Line,Frame} sets message_complete_ and
+	// marks begin/end pointers if a line/frame is complete, 
+	// otherwise it clears message_complete_ and copies all 
+	// the original data to the buffer.
+	//
+	void MarkOrCopy();
+	void MarkOrCopyLine();
+	void MarkOrCopyFrame();
+
+	void MarkOrCopyLine_CR_OR_LF();
+	void MarkOrCopyLine_STRICT_CRLF();
+
+	int 	buffer_n_;		// number of bytes in the buffer
+	int 	buffer_length_;	// size of the buffer
+	u_char	*buffer_;
+	bool 	message_complete_;
+	int 	frame_length_;
+	bool	chunked_;
+	const_byteptr orig_data_begin_, orig_data_end_;
+
+	LineBreakStyle linebreak_style_;
+
+	enum {
+		UNKNOWN_MODE,
+		LINE_MODE,
+		FRAME_MODE,
+	} mode_;
+
+	enum {
+		CR_OR_LF_0,
+		CR_OR_LF_1,
+		STRICT_CRLF_0,
+		STRICT_CRLF_1,
+		FRAME_0,
+	} state_;
+
+	int 	data_seq_at_orig_data_end_;
+	bool 	eof_;
+	bool    have_pending_request_;
+};
+
+typedef FlowBuffer *flow_buffer_t;
+
+}  // namespace binpac
+
+#endif // binpac_buffer_h

aux/binpac/lib/binpac_bytestring.cc

@@ -0,0 +1,24 @@
+#define binpac_regex_h
+
+#include <stdlib.h>
+#include "binpac_bytestring.h"
+
+namespace binpac 
+{
+
+std::string std_string(bytestring const *s)
+	{
+	return std::string((const char *) s->begin(), (const char *) s->end());
+	}
+
+int bytestring_to_int(bytestring const *s)
+	{
+	return atoi((const char *) s->begin());
+	}
+
+double bytestring_to_double(bytestring const *s)
+	{
+	return atof((const char *) s->begin());
+	}
+
+}  // namespace binpac

aux/binpac/lib/binpac_bytestring.h

@@ -0,0 +1,199 @@
+#ifndef binpac_bytestring_h
+#define binpac_bytestring_h
+
+#include <string.h>
+#include <string>
+#include "binpac.h"
+
+namespace binpac
+{
+
+template<class T> class datastring;
+
+template <class T>
+class const_datastring
+{
+public:
+	const_datastring()
+		: begin_(0), end_(0)
+		{ 
+		}
+
+	const_datastring(T const *data, int length)
+		: begin_(data), end_(data + length)
+		{
+		}
+
+	const_datastring(const T *begin, const T *end)
+		: begin_(begin), end_(end)
+		{
+		}
+
+	const_datastring(datastring<T> const &s)
+		: begin_(s.begin()), end_(s.end())
+		{
+		}
+
+	void init(const T *data, int length)
+		{
+		begin_ = data;
+		end_ = data + length;
+		}
+
+	T const *begin() const	{ return begin_; }
+	T const *end() const	{ return end_; }
+	int length() const	{ return end_ - begin_; }
+
+	T const &operator[](int index) const
+		{
+		return begin()[index];
+		}
+
+	bool operator==(const_datastring<T> const &s)
+		{
+		if ( length() != s.length() )
+			return false;
+		return memcmp((const void *) begin(), (const void *) s.begin(),
+			sizeof(T) * length()) == 0;
+		}
+
+	void set_begin(T const *begin)	{ begin_ = begin; }
+	void set_end(T const *end)	{ end_ = end; }
+
+private:
+	T const *begin_; 
+	T const *end_;
+};
+
+typedef const_datastring<uint8>	const_bytestring;
+
+template<class T>
+class datastring
+{
+public:
+	datastring()
+		{ 
+		clear();
+		}
+
+	datastring(T *data, int len)
+		{
+		set(data, len);
+		}
+
+	datastring(T const *begin, T const *end)
+		{
+		set_const(begin, end - begin);
+		}
+
+	datastring(datastring<T> const &x)
+		: data_(x.data()), length_(x.length())
+		{
+		}
+
+	explicit datastring(const_datastring<T> const &x)
+		{
+		set_const(x.begin(), x.length());
+		}
+
+	datastring const &operator=(datastring<T> const &x)
+		{
+		BINPAC_ASSERT(!data_);
+		set(x.data(), x.length());
+		return *this;
+		}
+
+	void init(T const *begin, int length)
+		{
+		BINPAC_ASSERT(!data_);
+		set_const(begin, length);
+		}
+
+	void clear()
+		{
+		data_ = 0; length_ = 0;
+		}
+
+	void free()
+		{
+		if ( data_ )
+			delete [] data_;
+		clear();
+		}
+
+	void clone()
+		{
+		set_const(begin(), length());
+		}
+
+	datastring const &operator=(const_datastring<T> const &x)
+		{
+		BINPAC_ASSERT(!data_);
+		set_const(x.begin(), x.length());
+		return *this;
+		}
+
+	T const &operator[](int index) const
+		{
+		return begin()[index];
+		}
+
+	T *data() const		{ return data_; }
+	int length() const	{ return length_; }
+
+	T const *begin() const	{ return data_; }
+	T const *end() const	{ return data_ + length_; }
+
+private:
+	void set(T *data, int len)
+		{
+		data_ = data;
+		length_ = len;
+		}
+
+	void set_const(T const *data, int len)
+		{
+		length_ = len;
+		data_ = new T[len + 1];
+		memcpy(data_, data, sizeof(T) * len);
+		data_[len] = 0;
+		}
+
+	T * data_;
+	int length_;
+};
+
+typedef datastring<uint8> bytestring;
+
+inline const char *c_str(bytestring const &s)
+	{
+	return (const char *) s.begin();
+	}
+
+inline std::string std_str(const_bytestring const &s) 
+	{
+	return std::string((const char *) s.begin(), (const char *) s.end());
+	}
+
+inline bool operator==(bytestring const &s1, const char *s2)
+	{
+	return strcmp(c_str(s1), s2) == 0;
+	}
+
+inline void get_pointers(const_bytestring const &s, 
+		uint8 const **pbegin, uint8 const **pend)
+	{
+	*pbegin = s.begin();
+	*pend = s.end();
+	}
+
+inline void get_pointers(bytestring const *s, 
+		uint8 const **pbegin, uint8 const **pend)
+	{
+	*pbegin = s->begin();
+	*pend = s->end();
+	}
+
+} // namespace binpac
+
+#endif  // binpac_bytestring_h

aux/binpac/lib/binpac_exception.h

@@ -0,0 +1,112 @@
+#ifndef binpac_exception_h
+#define binpac_exception_h
+
+namespace binpac {
+
+class Exception
+{
+public:
+	Exception(const char* m = 0)
+		: msg_("binpac exception: ")
+		{
+		if ( m )
+			append(m);
+		// abort();
+		}
+
+	void append(string m) 		{ msg_ += m; }
+	string msg() const		{ return msg_; }
+	const char* c_msg() const 	{ return msg().c_str(); }
+
+protected:
+	string msg_;
+};
+
+class ExceptionOutOfBound : public Exception
+{
+public:
+	ExceptionOutOfBound(const char* where, int len_needed, int len_given)
+		{
+		append(binpac_fmt("out_of_bound: %s: %d > %d", 
+			where, len_needed, len_given));
+		}
+};
+
+class ExceptionInvalidCase : public Exception
+{
+public:
+	ExceptionInvalidCase(const char* location, 
+			int index,
+			const char *expected)
+		: location_(location), 
+		  index_(index), 
+		  expected_(expected)
+		{
+		append(binpac_fmt("invalid case: %s: %d (%s)",
+			location, index, expected));
+		}
+
+protected:
+	const char* location_;
+	int index_;
+	string expected_;
+};
+
+class ExceptionInvalidCaseIndex : public Exception
+{
+public:
+	ExceptionInvalidCaseIndex(const char* location, 
+			int index)
+		: location_(location), 
+		  index_(index)
+		{
+		append(binpac_fmt("invalid index for case: %s: %d",
+			location, index));
+		}
+
+protected:
+	const char* location_;
+	int index_;
+};
+
+class ExceptionInvalidOffset : public Exception
+{
+public:
+	ExceptionInvalidOffset(const char* location, 
+			int min_offset, int offset)
+		: location_(location), 
+		  min_offset_(min_offset), offset_(offset)
+		{
+		append(binpac_fmt("invalid offset: %s: min_offset = %d, offset = %d",
+			location, min_offset, offset));
+		}
+
+protected:
+	const char* location_;
+	int min_offset_, offset_;
+};
+
+class ExceptionStringMismatch : public Exception
+{
+public:
+	ExceptionStringMismatch(const char* location, 
+			const char *expected, const char *actual_data)
+		{
+		append(binpac_fmt("string mismatch at %s: \nexpected pattern: \"%s\"\nactual data: \"%s\"",
+			location, expected, actual_data));
+		}
+};
+
+class ExceptionInvalidStringLength : public Exception
+{
+public:
+	ExceptionInvalidStringLength(const char* location, int len)
+		{
+		append(binpac_fmt("invalid length string: %s: %d",
+			location, len));
+		}
+};
+
+}
+
+#endif  // binpac_exception_h

aux/binpac/lib/binpac_regex.h

@@ -0,0 +1,43 @@
+#ifndef binpac_regex_h
+#define binpac_regex_h
+
+#include "binpac.h"
+#include "RE.h"
+
+class RE_Matcher;
+
+namespace binpac
+{
+
+class RegExMatcher {
+public:
+	RegExMatcher(const char *pattern)
+		: pattern_(pattern)
+		{
+		re_matcher_ = 0;
+		}
+
+	~RegExMatcher()
+		{
+		delete re_matcher_;
+		}
+
+	// Returns the length of longest match, or -1 on mismatch.
+	int MatchPrefix(const_byteptr data, int len)
+		{
+		if ( ! re_matcher_ )
+			{
+			re_matcher_ = new RE_Matcher(pattern_.c_str());
+			re_matcher_->Compile();
+			}
+		return re_matcher_->MatchPrefix(data, len);
+		}
+
+private:
+	string pattern_;
+	RE_Matcher *re_matcher_;
+};
+
+}  // namespace binpac
+
+#endif  // binpac_regex_h

aux/binpac/patches/binpac-1.patch

@@ -0,0 +1,31 @@
+diff -urN bro-1.2.1-orig/src/binpac/pac_expr.cc bro-1.2.1-ssl-binpac/src/binpac/pac_expr.cc
+--- bro-1.2.1-orig/src/binpac/pac_expr.cc	2006-07-26 15:02:40.000000000 -0700
++++ bro-1.2.1-ssl-binpac/src/binpac/pac_expr.cc	2007-05-04 14:31:11.728494000 -0700
+@@ -776,6 +776,27 @@
+ 			}
+ 			break;
+ 
++		case EXPR_CALLARGS:
++		        {
++		        mhs = 0;
++			if ( args_ )
++			        for ( uint i = 0; i < args_->size(); ++i )
++				        mhs = mhs_max(mhs, args_->at(i)->MinimalHeaderSize(env));
++			}
++		        break;
++		case EXPR_CASE:
++		        {
++		        mhs = operand_[0]->MinimalHeaderSize(env);
++			for ( uint i = 0; i < cases_->size(); ++i )
++			        {
++				CaseExpr * ce = cases_->at(i);
++				if ( ce->index() )
++				        for ( uint j = 0; j < ce->index()->size(); ++j )
++						mhs = mhs_max(mhs, ce->index()->at(j)->MinimalHeaderSize(env));
++				mhs = mhs_max(mhs, ce->value()->MinimalHeaderSize(env));
++				}
++			}
++			break;
+ 		default:
+ 			// Evaluate every operand by default
+ 			mhs = 0;

aux/binpac/patches/binpac-2.patch

@@ -0,0 +1,97 @@
+diff -urN bro-1.2.1-orig/src/binpac/pac_expr.cc bro-1.2.1-ssl-binpac/src/binpac/pac_expr.cc
+--- bro-1.2.1-orig/src/binpac/pac_expr.cc	2006-07-26 15:02:40.000000000 -0700
++++ bro-1.2.1-ssl-binpac/src/binpac/pac_expr.cc	2007-05-04 14:31:11.728494000 -0700
+@@ -245,6 +245,12 @@
+ 	out_cc->println("%s %s;", 
+ 	                val_type->DataTypeStr().c_str(), 
+ 	                env->LValue(val_var));
++
++	// force evaluation of IDs appearing in case stmt
++       operand_[0]->ForceIDEval(out_cc, env);
++	foreach(i, CaseExprList, cases_)
++		(*i)->value()->ForceIDEval(out_cc, env);
++
+ 	out_cc->println("switch ( %s )", operand_[0]->EvalExpr(out_cc, env));
+ 
+ 	out_cc->inc_indent();
+@@ -386,6 +392,49 @@
+ 		}
+ 	}
+ 
++void Expr::ForceIDEval(Output* out_cc, Env* env)
++        {
++	switch ( expr_type_ )
++		{
++		case EXPR_NUM:
++		case EXPR_SIZEOF:
++		case EXPR_OFFSETOF:
++			break;
++
++		case EXPR_ID:
++			if ( ! env->Evaluated(id_) )
++				env->Evaluate(out_cc, id_);
++			break;
++
++		case EXPR_MEMBER:
++			operand_[0]->ForceIDEval(out_cc, env);
++			break;
++
++		case EXPR_CALLARGS:
++		        {
++		        foreach(i, ExprList, args_)
++			        (*i)->ForceIDEval(out_cc, env);
++			}
++			break;
++
++		case EXPR_CASE:
++		        {
++		        operand_[0]->ForceIDEval(out_cc, env);
++			foreach(i, CaseExprList, cases_)
++				(*i)->value()->ForceIDEval(out_cc, env);
++		        }
++			break;
++
++		default:
++			// Evaluate every operand by default
++			for ( int i = 0; i < 3; ++i )
++				if ( operand_[i] )
++					operand_[i]->ForceIDEval(out_cc, env);
++			break;
++		}
++	}
++
++
+ const char* Expr::EvalExpr(Output* out_cc, Env* env)
+ 	{
+ 	GenEval(out_cc, env);
+diff -urN bro-1.2.1-orig/src/binpac/pac_expr.h bro-1.2.1-ssl-binpac/src/binpac/pac_expr.h
+--- bro-1.2.1-orig/src/binpac/pac_expr.h	2006-07-26 15:02:39.000000000 -0700
++++ bro-1.2.1-ssl-binpac/src/binpac/pac_expr.h	2007-05-04 14:16:31.624287000 -0700
+@@ -56,6 +56,11 @@
+ 	//
+ 	const char *EvalExpr(Output *out, Env *env);
+ 
++	// force evaulation of IDs contained in this expression;
++	// necessary with case expr and conditional let fields (&if)
++	// for correct parsing of fields
++	void ForceIDEval(Output *out_cc, Env *env);
++
+ 	// Returns the set_* function of the expression. 
+ 	// The expression must be of form ID or x.ID.
+ 	string SetFunc(Output *out, Env *env);
+diff -urN bro-1.2.1-orig/src/binpac/pac_let.cc bro-1.2.1-ssl-binpac/src/binpac/pac_let.cc
+--- bro-1.2.1-orig/src/binpac/pac_let.cc	2006-07-26 15:02:39.000000000 -0700
++++ bro-1.2.1-ssl-binpac/src/binpac/pac_let.cc	2007-05-04 15:32:09.695568000 -0700
+@@ -80,7 +80,12 @@
+ 	if ( type_->attr_if_expr() )
+ 		{
+ 		// A conditional field
++
+ 		env->Evaluate(out_cc, type_->has_value_var());
++
++		// force evaluation of IDs contained in this expr
++		expr()->ForceIDEval(out_cc, env);
++
+ 		out_cc->println("if ( %s )", 
+ 			env->RValue(type_->has_value_var()));
+ 		out_cc->inc_indent();

aux/binpac/patches/binpac-3.patch

@@ -0,0 +1,37 @@
+diff -urN bro-1.2.1-orig/src/binpac/pac_let.cc bro-1.2.1-ssl-binpac/src/binpac/pac_let.cc
+--- bro-1.2.1-orig/src/binpac/pac_let.cc	2006-07-26 15:02:39.000000000 -0700
++++ bro-1.2.1-ssl-binpac/src/binpac/pac_let.cc	2007-05-04 15:32:09.695568000 -0700
+@@ -108,11 +108,6 @@
+ void LetField::GenEval(Output* out_cc, Env* env)
+ 	{
+ 	GenParseCode(out_cc, env);
+-	if ( type_->attr_if_expr() )
+-		{
+-		out_cc->println("BINPAC_ASSERT(%s);", 
+-			env->RValue(type_->has_value_var()));
+-		}
+ 	}
+ 
+ LetDecl::LetDecl(ID *id, Type *type, Expr *expr)
+diff -urN bro-1.2.1-orig/src/binpac/pac_type.cc bro-1.2.1-ssl-binpac/src/binpac/pac_type.cc
+--- bro-1.2.1-orig/src/binpac/pac_type.cc	2006-07-26 15:02:40.000000000 -0700
++++ bro-1.2.1-ssl-binpac/src/binpac/pac_type.cc	2007-05-24 10:56:42.140658000 -0700
+@@ -316,9 +316,15 @@
+ 	{
+ 	if ( DefineValueVar() )
+ 		{
+-		out_h->println("%s %s const	{ return %s; }", 
+-			DataTypeConstRefStr().c_str(), 
+-			env->RValue(value_var()), lvalue());
++		if ( attr_if_expr_ )
++		        out_h->println("%s %s const { BINPAC_ASSERT(%s); return %s; }",
++			        DataTypeConstRefStr().c_str(), 
++			        env->RValue(value_var()),
++			        env->RValue(has_value_var()), lvalue());
++		else
++		        out_h->println("%s %s const { return %s; }",
++			        DataTypeConstRefStr().c_str(), 
++			        env->RValue(value_var()), lvalue());
+ 		}
+ 
+ 	foreach (i, FieldList, fields_)

aux/binpac/patches/binpac-4.patch

@@ -0,0 +1,12 @@
+diff -urN bro-1.2.1-orig/src/binpac/pac_record.cc bro-1.2.1-ssl-binpac/src/binpac/pac_record.cc
+--- bro-1.2.1-orig/src/binpac/pac_record.cc	2006-07-26 15:02:40.000000000 -0700
++++ bro-1.2.1-ssl-binpac/src/binpac/pac_record.cc	2007-05-08 16:13:33.373850000 -0700
+@@ -123,7 +123,7 @@
+ void RecordType::DoGenParseCode(Output* out_cc, Env* env, 
+ 		const DataPtr& data, int flags)
+ 	{
+-	if ( StaticSize(env) >= 0 )
++	if ( !incremental_input() && StaticSize(env) >= 0 )
+ 		GenBoundaryCheck(out_cc, env, data);
+ 
+ 	if ( incremental_parsing() )

aux/binpac/patches/binpac-5.patch

@@ -0,0 +1,66 @@
+diff -urN bro-1.2.1-orig/src/binpac/pac_paramtype.cc bro-1.2.1-ssl-binpac/src/binpac/pac_paramtype.cc
+--- bro-1.2.1-orig/src/binpac/pac_paramtype.cc	2006-07-26 15:02:40.000000000 -0700
++++ bro-1.2.1-ssl-binpac/src/binpac/pac_paramtype.cc	2007-05-10 15:09:47.470104000 -0700
+@@ -208,7 +208,13 @@
+ 	const char *parse_func;
+ 	string parse_params;
+ 
+-	if ( ref_type->incremental_input() )
++	if ( buffer_mode() == BUFFER_NOTHING )
++	        {
++		ASSERT(!ref_type->incremental_input());
++		parse_func = kParseFuncWithoutBuffer;
++		parse_params = "0, 0";
++		}
++	else if ( ref_type->incremental_input() )
+ 		{
+ 		parse_func = kParseFuncWithBuffer;
+ 		parse_params = env->RValue(flow_buffer_id);
+@@ -239,15 +245,24 @@
+ 
+ 	if ( incremental_input() )
+ 		{
+-		ASSERT(parsing_complete_var());
+-		out_cc->println("%s = %s;",
+-			env->LValue(parsing_complete_var()),
+-			call_parse_func.c_str());
+-
+-		// parsing_complete_var might have been already
+-		// evaluated when set to false
+-		if ( ! env->Evaluated(parsing_complete_var()) )
+-			env->SetEvaluated(parsing_complete_var());
++		if ( buffer_mode() == BUFFER_NOTHING )
++		        {
++		        out_cc->println("%s;", call_parse_func.c_str());
++			out_cc->println("%s = true;", 
++				env->LValue(parsing_complete_var()));
++			}
++		else
++		        {
++			ASSERT(parsing_complete_var());
++			out_cc->println("%s = %s;",
++				env->LValue(parsing_complete_var()),
++				call_parse_func.c_str());
++
++			// parsing_complete_var might have been already
++			// evaluated when set to false
++			if ( ! env->Evaluated(parsing_complete_var()) )
++			        env->SetEvaluated(parsing_complete_var());
++			}
+ 		}
+ 	else
+ 		{
+diff -urN bro-1.2.1-orig/src/binpac/pac_type.cc bro-1.2.1-ssl-binpac/src/binpac/pac_type.cc
+--- bro-1.2.1-orig/src/binpac/pac_type.cc	2006-07-26 15:02:40.000000000 -0700
++++ bro-1.2.1-ssl-binpac/src/binpac/pac_type.cc	2007-05-24 10:56:42.140658000 -0700
+@@ -501,8 +501,8 @@
+ 
+ 		if ( buffer_mode() == BUFFER_NOTHING )
+ 			{
+-			out_cc->println("%s = true;", 
+-				env->LValue(parsing_complete_var()));
++			// this is the empty type
++			DoGenParseCode(out_cc, env, data, flags);
+ 			}
+ 		else if ( buffer_input() )
+ 			{

aux/binpac/patches/binpac-6.patch

@@ -0,0 +1,28 @@
+diff -urN bro-1.2.1-orig/src/binpac/lib/binpac_buffer.h bro-1.2.1-ssl-binpac/src/binpac/lib/binpac_buffer.h
+--- bro-1.2.1-orig/src/binpac/lib/binpac_buffer.h	2006-07-26 15:02:38.000000000 -0700
++++ bro-1.2.1-ssl-binpac/src/binpac/lib/binpac_buffer.h	2007-05-09 16:14:54.501656000 -0700
+@@ -59,6 +59,11 @@
+ 			return frame_length_;
+ 		}
+ 
++	inline bool data_available() const
++	        {
++		return buffer_n_ > 0 || orig_data_end_ > orig_data_begin_;
++		}
++
+ 	void NewLine();
+ 	// A negative frame_length represents a frame till EOF
+ 	void NewFrame(int frame_length, bool chunked_);
+diff -urN bro-1.2.1-orig/src/binpac/pac_flow.cc bro-1.2.1-ssl-binpac/src/binpac/pac_flow.cc
+--- bro-1.2.1-orig/src/binpac/pac_flow.cc	2006-10-12 14:13:12.000000000 -0700
++++ bro-1.2.1-ssl-binpac/src/binpac/pac_flow.cc	2007-05-22 16:43:55.997562000 -0700
+@@ -272,7 +272,8 @@
+ 		env_->RValue(begin_of_data), 
+ 		env_->RValue(end_of_data)); 
+ 
+-	out_cc->println("while ( true )");
++	out_cc->println("while ( %s->data_available() )",
++		env_->LValue(flow_buffer_id));
+ 	out_cc->inc_indent();
+ 	out_cc->println("{");
+ 

aux/binpac/patches/binpac-7.patch

@@ -0,0 +1,21 @@
+diff -urN bro-1.2.1-orig/src/binpac/pac_type.cc bro-1.2.1-ssl-binpac/src/binpac/pac_type.cc
+--- bro-1.2.1-orig/src/binpac/pac_type.cc	2006-07-26 15:02:40.000000000 -0700
++++ bro-1.2.1-ssl-binpac/src/binpac/pac_type.cc	2007-05-24 10:56:42.140658000 -0700
+@@ -393,7 +393,7 @@
+ 			break;
+ 
+ 		case BUFFER_BY_LENGTH:
+-			if ( buffering_state_var_field_ )
++		        if ( env->GetDataType(buffering_state_id) )
+ 				{
+ 				out_cc->println("if ( %s == 0 )",
+ 					env->RValue(buffering_state_id));
+@@ -421,7 +421,7 @@
+ 				frame_buffer_arg.c_str(),
+ 				attr_chunked() ? "true" : "false");
+ 
+-			if ( buffering_state_var_field_ )
++			if ( env->GetDataType(buffering_state_id) )
+ 				{
+ 				out_cc->println("%s = 1;",
+ 					env->LValue(buffering_state_id));

aux/binpac/patches/binpac-8.patch

@@ -0,0 +1,190 @@
+diff -urN bro-1.2.1-orig/src/binpac/pac_analyzer.cc bro-1.2.1-ssl-binpac/src/binpac/pac_analyzer.cc
+--- bro-1.2.1-orig/src/binpac/pac_analyzer.cc	2006-07-26 15:02:40.000000000 -0700
++++ bro-1.2.1-ssl-binpac/src/binpac/pac_analyzer.cc	2007-05-22 17:00:10.091531000 -0700
+@@ -26,8 +26,9 @@
+ 	helpers_ = new AnalyzerHelperList();
+ 	functions_ = new FunctionList();
+ 
+-	constructor_helper_ = 0;
+-	destructor_helper_ = 0;
++	constructor_helpers_ = new AnalyzerHelperList();
++	destructor_helpers_ = new AnalyzerHelperList();
++	eof_helpers_ = new AnalyzerHelperList();
+ 
+ 	SetAnalyzerContext();
+ 
+@@ -41,6 +42,9 @@
+ 	delete_list(AnalyzerHelperList, helpers_);
+ 	delete_list(FunctionList, functions_);
+ 	delete_list(ParamList, params_);
++	delete_list(AnalyzerHelperList, constructor_helpers_);
++	delete_list(AnalyzerHelperList, destructor_helpers_);
++	delete_list(AnalyzerHelperList, eof_helpers_);
+ 	}
+ 
+ void AnalyzerDecl::AddElements(AnalyzerElementList *elemlist)
+@@ -75,28 +79,20 @@
+ 				AnalyzerHelper *helper_elem = 
+ 					(AnalyzerHelper *) elem;
+ 
+-				if ( helper_elem->helper_type() == 
+-				     AnalyzerHelper::INIT_CODE)
+-					{
+-					if ( constructor_helper_ )
+-						{
+-						throw Exception(elem,
+-							"Repeated definition of %init code");
+-						}
+-					constructor_helper_ = helper_elem;
++				switch ( helper_elem->helper_type() )
++				        {
++					case AnalyzerHelper::INIT_CODE:
++					        constructor_helpers_->push_back(helper_elem);
++						break;
++					case AnalyzerHelper::CLEANUP_CODE:
++ 				                destructor_helpers_->push_back(helper_elem);
++						break;
++					case AnalyzerHelper::EOF_CODE:
++					        eof_helpers_->push_back(helper_elem);
++						break;
++					default:
++					        helpers_->push_back(helper_elem);
+ 					}
+-				else if ( helper_elem->helper_type() == 
+-				          AnalyzerHelper::CLEANUP_CODE)
+-					{
+-					if ( destructor_helper_ )
+-						{
+-						throw Exception(elem,
+-							"Repeated definition of %cleanup code");
+-						}
+-					destructor_helper_ = helper_elem;
+-					}
+-				else
+-					helpers_->push_back(helper_elem);
+ 				}
+ 				break;
+ 			case AnalyzerElement::FUNCTION:
+@@ -217,15 +213,19 @@
+ void AnalyzerDecl::GenInitCode(Output *out_cc)
+ 	{
+ 	TypeDecl::GenInitCode(out_cc);
+-	if ( constructor_helper_ )
+-		constructor_helper_->GenCode(0, out_cc, this);
++	foreach(i, AnalyzerHelperList, constructor_helpers_)
++		{
++		(*i)->GenCode(0, out_cc, this);
++		}
+ 	}
+ 
+ void AnalyzerDecl::GenCleanUpCode(Output *out_cc)
+ 	{
+ 	TypeDecl::GenCleanUpCode(out_cc);
+-	if ( destructor_helper_ )
+-		destructor_helper_->GenCode(0, out_cc, this);
++	foreach(i, AnalyzerHelperList, destructor_helpers_)
++		{
++		(*i)->GenCode(0, out_cc, this);
++		}
+ 	}
+ 
+ void AnalyzerDecl::GenStateVarDecls(Output *out_h)
+@@ -295,6 +295,7 @@
+ 			break;
+ 		case INIT_CODE:
+ 		case CLEANUP_CODE:
++		case EOF_CODE:
+ 			out = out_cc;
+ 			break;
+ 		}
+diff -urN bro-1.2.1-orig/src/binpac/pac_analyzer.h bro-1.2.1-ssl-binpac/src/binpac/pac_analyzer.h
+--- bro-1.2.1-orig/src/binpac/pac_analyzer.h	2006-07-26 15:02:39.000000000 -0700
++++ bro-1.2.1-ssl-binpac/src/binpac/pac_analyzer.h	2007-05-22 16:32:08.397926000 -0700
+@@ -76,8 +76,9 @@
+ 	AnalyzerHelperList *helpers_;
+ 	FunctionList *functions_;
+ 
+-	AnalyzerHelper *constructor_helper_;
+-	AnalyzerHelper *destructor_helper_;
++	AnalyzerHelperList *constructor_helpers_;
++	AnalyzerHelperList *destructor_helpers_;
++	AnalyzerHelperList *eof_helpers_;
+ };
+ 
+ class AnalyzerElement : public Object
+@@ -117,6 +118,7 @@
+ 		MEMBER_DECLS,
+ 		INIT_CODE,
+ 		CLEANUP_CODE,
++		EOF_CODE,
+ 	};
+ 	AnalyzerHelper(Type helper_type, EmbeddedCode *code)
+ 		: AnalyzerElement(HELPER),
+diff -urN bro-1.2.1-orig/src/binpac/pac_conn.cc bro-1.2.1-ssl-binpac/src/binpac/pac_conn.cc
+--- bro-1.2.1-orig/src/binpac/pac_conn.cc	2006-07-26 15:02:40.000000000 -0700
++++ bro-1.2.1-ssl-binpac/src/binpac/pac_conn.cc	2007-05-22 16:42:35.406135000 -0700
+@@ -97,6 +97,12 @@
+ 	out_cc->println("%s->%s();",
+ 	                env_->LValue(downflow_id),
+ 	                kFlowEOF);
++
++	foreach(i, AnalyzerHelperList, eof_helpers_)
++		{
++		(*i)->GenCode(0, out_cc, this);
++		}
++
+ 	out_cc->dec_indent();
+ 	
+ 	out_cc->println("}");
+diff -urN bro-1.2.1-orig/src/binpac/pac_flow.cc bro-1.2.1-ssl-binpac/src/binpac/pac_flow.cc
+--- bro-1.2.1-orig/src/binpac/pac_flow.cc	2006-10-12 14:13:12.000000000 -0700
++++ bro-1.2.1-ssl-binpac/src/binpac/pac_flow.cc	2007-05-22 16:43:55.997562000 -0700
+@@ -151,6 +151,11 @@
+ 	out_cc->inc_indent();
+ 	out_cc->println("{");
+ 
++	foreach(i, AnalyzerHelperList, eof_helpers_)
++		{
++		(*i)->GenCode(0, out_cc, this);
++		}
++
+ 	if ( dataunit_->type() == AnalyzerDataUnit::FLOWUNIT )
+ 		{
+ 		out_cc->println("%s->set_eof();", 
+diff -urN bro-1.2.1-orig/src/binpac/pac_parse.yy bro-1.2.1-ssl-binpac/src/binpac/pac_parse.yy
+--- bro-1.2.1-orig/src/binpac/pac_parse.yy	2006-10-12 14:13:12.000000000 -0700
++++ bro-1.2.1-ssl-binpac/src/binpac/pac_parse.yy	2007-05-22 16:56:09.280526000 -0700
+@@ -22,7 +22,7 @@
+ %token TOK_STATE TOK_ACTION TOK_WHEN TOK_HELPER 
+ %token TOK_DATAUNIT TOK_FLOWDIR TOK_WITHCONTEXT
+ %token TOK_LPB_EXTERN TOK_LPB_HEADER TOK_LPB_CODE 
+-%token TOK_LPB_MEMBER TOK_LPB_INIT TOK_LPB_CLEANUP 
++%token TOK_LPB_MEMBER TOK_LPB_INIT TOK_LPB_CLEANUP TOK_LPB_EOF
+ %token TOK_LPB TOK_RPB 
+ %token TOK_EMBEDDED_ATOM TOK_EMBEDDED_STRING
+ %token TOK_PAC_VAL TOK_PAC_SET TOK_PAC_TYPE TOK_PAC_TYPEOF TOK_PAC_CONST_DEF 
+@@ -795,6 +795,10 @@
+ 				{
+ 				$$ = new AnalyzerHelper(AnalyzerHelper::CLEANUP_CODE, $2);
+ 				}
++		|	TOK_LPB_EOF embedded_code TOK_RPB
++				{
++				$$ = new AnalyzerHelper(AnalyzerHelper::EOF_CODE, $2);
++				}
+ 		|	TOK_FLOWDIR '=' tok_id optargs ';'
+ 				{
+ 				$$ = new AnalyzerFlow((AnalyzerFlow::Direction) $1, $3, $4);
+diff -urN bro-1.2.1-orig/src/binpac/pac_scan.ll bro-1.2.1-ssl-binpac/src/binpac/pac_scan.ll
+--- bro-1.2.1-orig/src/binpac/pac_scan.ll	2006-07-26 15:02:40.000000000 -0700
++++ bro-1.2.1-ssl-binpac/src/binpac/pac_scan.ll	2007-05-22 16:55:19.349644000 -0700
+@@ -96,6 +96,10 @@
+ 				BEGIN(EC);
+ 				return TOK_LPB_MEMBER;
+ 				}
++<INITIAL>"%eof{"		{
++				BEGIN(EC);
++				return TOK_LPB_EOF;
++				}
+ <INITIAL>"%{"			{
+ 				BEGIN(EC);
+ 				return TOK_LPB;

aux/binpac/patches/binpac-patch-doc.txt

@@ -0,0 +1,87 @@
+binpac fixes
+----------------
+
+numbers of issues below correspond to the patch numbers
+
+(1) correct calculation of minimal header size in pac_expr.cc
+- problem: EXPR_CALLARGS and EXPR_CASE not considered for the calculation
+  of minimal header size
+- solution: added two cases in switch stmt of Expr::MinimalHeaderSize
+  for EXPR_CALLARGS and EXPR_CASE
+
+
+(2) ensure parsing of fields first referenced in a case expression or
+    let field with an &if attribute
+- problem: in cases where the if expression evaluates to false or the
+  proper case does not occur, fields get not parsed at all
+- solution: force evaluation of all IDs referenced in a let field with
+  if attribute or a case expression before the body of the corresponding
+  switch stmt or the if stmt
+- added public method Expr::ForceIDEval, properly called before
+  generating the code of a field with if attribute or the case expression
+
+
+(3) properly assert the use of fields with an if attribute
+- problem: the use of fields with an if attribute was not asserted in all
+  cases and asserted in the wrong way in some others due to the
+  corresponding BINPAC_ASSERT only called upon parsing the field
+- solution: perform BINPAC_ASSERT upon calling the fields accessor
+  function
+- moved BINPAC_ASSERT statement from LetField::GenEval to
+  Type::GenPubDecls
+
+
+(4) incremental input with records with a non-negative StaticSize
+- problem: incremental input with records with a StaticSize >= 0
+  cannot be performed due to necessary length attribute, leading to
+  an invalid call of GenBoundaryCheck in RecordType::DoGenParseCode
+- solution: added a check for incremental input in
+  RecordType::DoGenParseCode before calling GenBoundaryCheck
+
+
+(5) empty type with incremental input
+- problem: with an empty type and incremental input, although the
+  Parse function is created, it is never called, leading to problems,
+  if additional actions are to be performed when encountering that
+  empty type
+- solution: generate call to Parse of empty type in Type::GenParseBuffer
+
+
+(6) parsing loop in flow ParseBuffer (while(true))
+- problem: while(true) leads to problems after parsing of a type is
+  complete; at this time, it is unexpected that parsing continues, even
+  if no data is available in the flow buffer
+- solution: check if data is available before starting a new parsing
+  cycle
+- added a method data_available to FlowBuffer
+- changed while(true) in FlowDecl::GenCodeFlowUnit to
+  while(flow_buffer_->data_available())
+
+
+(7) initialization of flow buffer in CaseType with bufferable fields
+    in cases
+- problem: initialization of buffer occurs in every Parse call,
+  regardless if it was initialized before or not; initialization
+  is correct only on first such occurence
+- solution: check to buffer_state is to be created always when
+  buffering_state_id is in environment in Type::GenBufferConfig
+- changed condition from buffering_state_var_field_ to
+  env->GetDataType(buffering_state_id)
+
+
+(8) allowing init and cleanup code to be redefined, as well as addition
+    of code to FlowEOF calls in analyzer and flow
+- problem 1: when refining an analyzer or flow definition, additional
+  init and cleanup code was not allowed, if these were already defined
+  before; this leads to problems when adding new members, as these
+  cannot be initialized and destroyed properly
+- solution: allow init and cleanup code to be specified more than once
+- changed deifnitions and usage of constructor_helper and
+  destructor_helper to allow for lists of constructor and destructor
+  helpers (similar to member declarations) in pac_analyzer.h and
+  pac_analyzer.cc
+- problem 2: in some cases, it is desirable to execute code when
+  encountering the end of the input stream, which is not possible in
+  binpac
+- solution: added a %eof binpac primitive similar to %init, which adds
+  code to the FlowEOF function of an analyzer or a flow

aux/binpac/patches/nadi-bittorrent.patch

@@ -0,0 +1,172 @@
+Index: pac_type.h
+===================================================================
+--- pac_type.h	(revision 4130)
++++ pac_type.h	(working copy)
+@@ -78,12 +78,6 @@
+ 	string EvalByteOrder(Output *out_cc, Env *env) const;
+ 
+ 	virtual string EvalMember(const ID *member_id) const;
+-#if 0
+-	// member_env() is used for finding a member of the type.
+-	// Thus member_env() of a ParameterizedType should return 
+-	// ReferredDataType()->env()
+-	// virtual Env *member_env() const;
+-#endif
+ 
+ 	// The variable defined by the type
+ 	const ID *value_var() const		{ return value_var_; }
+@@ -223,6 +217,8 @@
+ 
+ 	virtual bool ByteOrderSensitive() const = 0;
+ 
++	bool NeedsBufferingStateVar() const;
++
+ 	void GenBufferingLoop(Output* out_cc, Env* env, int flags);
+ 	void GenParseBuffer(Output* out_cc, Env* env, int flags);
+ 	void GenParseCode2(Output* out_cc, Env* env, const DataPtr& data, int flags);
+Index: lib/binpac_buffer.h
+===================================================================
+--- lib/binpac_buffer.h	(revision 4130)
++++ lib/binpac_buffer.h	(working copy)
+@@ -24,18 +24,18 @@
+ 	void DiscardData();
+ 
+ 	// Whether there is enough data for the frame
+-	bool ready() const{ return message_complete_; }
++	bool ready() const{ return message_complete_ || mode_ == UNKNOWN_MODE; }
+ 
+ 	inline const_byteptr begin() const
+ 		{
+-		BINPAC_ASSERT(message_complete_);
++		BINPAC_ASSERT(ready());
+ 		return ( buffer_n_ == 0 ) ? 
+ 			orig_data_begin_ : buffer_;
+ 		}
+ 
+ 	inline const_byteptr end() const
+ 		{
+-		BINPAC_ASSERT(message_complete_);
++		BINPAC_ASSERT(ready());
+ 		if ( buffer_n_ == 0 )
+ 			{
+ 			BINPAC_ASSERT(frame_length_ >= 0);
+Index: pac_type.cc
+===================================================================
+--- pac_type.cc	(revision 4130)
++++ pac_type.cc	(working copy)
+@@ -285,9 +285,8 @@
+ 			parsing_complete_var, extern_type_bool->Clone());
+ 		parsing_complete_var_field_->Prepare(env);
+ 
+-		if ( ( buffer_mode() == BUFFER_BY_LENGTH ||
+-		       buffer_mode() == BUFFER_BY_LINE ) &&
+-		       ! env->GetDataType(buffering_state_id) )
++		if ( NeedsBufferingStateVar() && 
++		     !env->GetDataType(buffering_state_id) )
+ 			{
+ 			buffering_state_var_field_ = new PrivVarField(
+ 				buffering_state_id->clone(), 
+@@ -387,17 +386,17 @@
+ 			break;
+ 
+ 		case BUFFER_BY_LENGTH:
+-			if ( buffering_state_var_field_ )
+-				{
+-				out_cc->println("if ( %s == 0 )",
+-					env->RValue(buffering_state_id));
+-				out_cc->inc_indent();
+-				out_cc->println("{");
+-				}
++			if ( !NeedsBufferingStateVar() )
++				break;
+ 
++			ASSERT(env->GetDataType(buffering_state_id));
++			out_cc->println("if ( %s == 0 )",
++				env->RValue(buffering_state_id));
++			out_cc->inc_indent();
++			out_cc->println("{");
++
+ 			if ( attr_length_expr_ )
+ 				{
+-				// frame_buffer_arg = attr_length_expr_->EvalExpr(out_cc, env);
+ 				frame_buffer_arg = strfmt("%d", InitialBufferLength());
+ 				}
+ 			else if ( attr_restofflow_ )
+@@ -407,7 +406,7 @@
+ 				}
+ 			else
+ 				{
+-				frame_buffer_arg = strfmt("%d", InitialBufferLength());
++				ASSERT(0);
+ 				}
+ 
+ 			out_cc->println("%s->NewFrame(%s, %s);",
+@@ -415,16 +414,14 @@
+ 				frame_buffer_arg.c_str(),
+ 				attr_chunked() ? "true" : "false");
+ 
+-			if ( buffering_state_var_field_ )
+-				{
+-				out_cc->println("%s = 1;",
+-					env->LValue(buffering_state_id));
+-				out_cc->println("}");
+-				out_cc->dec_indent();
+-				}
++			out_cc->println("%s = 1;",
++				env->LValue(buffering_state_id));
++			out_cc->println("}");
++			out_cc->dec_indent();
+ 			break;
+ 
+ 		case BUFFER_BY_LINE:
++			ASSERT(env->GetDataType(buffering_state_id));
+ 			out_cc->println("if ( %s == 0 )",
+ 				env->RValue(buffering_state_id));
+ 			out_cc->inc_indent();
+@@ -890,6 +887,25 @@
+ 	return ! attr_byteorder_expr() && ByteOrderSensitive(); 
+ 	}
+ 
++bool Type::NeedsBufferingStateVar() const
++	{
++	if ( !incremental_input() )
++		return false;
++	switch ( buffer_mode() )
++		{
++		case BUFFER_NOTHING:
++		case NOT_BUFFERABLE:
++			return false;
++		case BUFFER_BY_LINE:
++			return true;
++		case BUFFER_BY_LENGTH:
++			return ( attr_length_expr_ || attr_restofflow_ );
++		default:
++			ASSERT(0);
++			return false;
++		}
++	}
++
+ bool Type::DoTraverse(DataDepVisitor *visitor)
+ 	{
+ 	foreach (i, FieldList, fields_)
+Index: pac_flow.cc
+===================================================================
+--- pac_flow.cc	(revision 4130)
++++ pac_flow.cc	(working copy)
+@@ -224,15 +224,13 @@
+ 	out_cc->println("catch ( Exception const &e )");
+ 	out_cc->inc_indent();
+ 	out_cc->println("{");
+-	out_cc->println("DEBUG_MSG(\"%%.6f binpac exception: %%s\\n\", network_time(), e.c_msg());");
+ 	GenCleanUpCode(out_cc);
+ 	if ( dataunit_->type() == AnalyzerDataUnit::FLOWUNIT )
+ 		{
+ 		out_cc->println("%s->DiscardData();",
+ 			env_->LValue(flow_buffer_id));
+-		out_cc->println("BINPAC_ASSERT(!%s->ready());",
+-			env_->RValue(flow_buffer_id));
+ 		}
++	out_cc->println("throw e;");
+ 	out_cc->println("}");
+ 	out_cc->dec_indent();
+ 

aux/binpac/shtool

@@ -0,0 +1,716 @@
+#!/bin/sh
+##
+##  GNU shtool -- The GNU Portable Shell Tool
+##  Copyright (c) 1994-2000 Ralf S. Engelschall <rse@engelschall.com>
+##
+##  See http://www.gnu.org/software/shtool/ for more information.
+##  See ftp://ftp.gnu.org/gnu/shtool/ for latest version.
+##
+##  Version 1.4.9 (16-Apr-2000)
+##  Ingredients: 3/17 available modules
+##
+
+##
+##  This program is free software; you can redistribute it and/or modify
+##  it under the terms of the GNU General Public License as published by
+##  the Free Software Foundation; either version 2 of the License, or
+##  (at your option) any later version.
+##
+##  This program 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
+##  General Public License for more details.
+##
+##  You should have received a copy of the GNU General Public License
+##  along with this program; if not, write to the Free Software
+##  Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307,
+##  USA, or contact Ralf S. Engelschall <rse@engelschall.com>.
+##
+##  Notice: Given that you include this file verbatim into your own
+##  source tree, you are justified in saying that it remains separate
+##  from your package, and that this way you are simply just using GNU
+##  shtool. So, in this situation, there is no requirement that your
+##  package itself is licensed under the GNU General Public License in
+##  order to take advantage of GNU shtool.
+##
+
+##
+##  Usage: shtool [<options>] [<cmd-name> [<cmd-options>] [<cmd-args>]]
+##
+##  Available commands:
+##    echo       Print string with optional construct expansion
+##    install    Install a program, script or datafile
+##    mkdir      Make one or more directories
+##
+##  Not available commands (because module was not built-in):
+##    mdate      Pretty-print modification time of a file or dir
+##    table      Pretty-print a field-separated list as a table
+##    prop       Display progress with a running propeller
+##    move       Move files with simultaneous substitution
+##    mkln       Make link with calculation of relative paths
+##    mkshadow   Make a shadow tree through symbolic links
+##    fixperm    Fix file permissions inside a source tree
+##    tarball    Roll distribution tarballs
+##    guessos    Simple operating system guesser
+##    arx        Extended archive command
+##    slo        Separate linker options by library class
+##    scpp       Sharing C Pre-Processor
+##    version    Generate and maintain a version information file
+##    path       Deal with program paths
+##
+
+if [ $# -eq 0 ]; then
+    echo "$0:Error: invalid command line" 1>&2
+    echo "$0:Hint:  run \`$0 -h' for usage" 1>&2
+    exit 1
+fi
+if [ ".$1" = ".-h" -o ".$1" = ".--help" ]; then
+    echo "This is GNU shtool, version 1.4.9 (16-Apr-2000)"
+    echo "Copyright (c) 1994-2000 Ralf S. Engelschall <rse@engelschall.com>"
+    echo "Report bugs to <bug-shtool@gnu.org>"
+    echo ''
+    echo "Usage: shtool [<options>] [<cmd-name> [<cmd-options>] [<cmd-args>]]" 
+    echo ''
+    echo 'Available global <options>:'
+    echo '  -v, --version   display shtool version information'
+    echo '  -h, --help      display shtool usage help page (this one)'
+    echo '  -d, --debug     display shell trace information'
+    echo ''
+    echo 'Available <cmd-name> [<cmd-options>] [<cmd-args>]:'
+    echo '  echo     [-n] [-e] [<str> ...]'
+    echo '  install  [-v] [-t] [-c] [-C] [-s] [-m<mode>] [-o<owner>] [-g<group>]'
+    echo '           [-e<ext>] <file> <path>'
+    echo '  mkdir    [-t] [-f] [-p] [-m<mode>] <dir> [<dir> ...]'
+    echo ''
+    echo 'Not available <cmd-name> (because module was not built-in):'
+    echo '  mdate    [-n] [-z] [-s] [-d] [-f<str>] [-o<spec>] <path>'
+    echo '  table    [-F<sep>] [-w<width>] [-c<cols>] [-s<strip>] <str><sep><str>...'
+    echo '  prop     [-p<str>]'
+    echo '  move     [-v] [-t] [-e] [-p] <src-file> <dst-file>'
+    echo '  mkln     [-t] [-f] [-s] <src-path> [<src-path> ...] <dst-path>'
+    echo '  mkshadow [-v] [-t] [-a] <src-dir> <dst-dir>'
+    echo '  fixperm  [-v] [-t] <path> [<path> ...]'
+    echo '  tarball  [-t] [-v] [-o <tarball>] [-c <prog>] [-d <dir>] [-u'
+    echo '           <user>] [-g <group>] [-e <pattern>] <path> [<path> ...]'
+    echo '  guessos  '
+    echo '  arx      [-t] [-C<cmd>] <op> <archive> [<file> ...]'
+    echo '  slo      [-p<str>] -- -L<dir> -l<lib> [-L<dir> -l<lib> ...]'
+    echo '  scpp     [-v] [-p] [-f<filter>] [-o<ofile>] [-t<tfile>] [-M<mark>]'
+    echo '           [-D<dname>] [-C<cname>] <file> [<file> ...]'
+    echo '  version  [-l<lang>] [-n<name>] [-p<prefix>] [-s<version>] [-i<knob>]'
+    echo '           [-d<type>] <file>'
+    echo '  path     [-s] [-r] [-d] [-b] [-m] [-p<path>] <str> [<str> ...]'
+    echo ''
+    exit 0
+fi
+if [ ".$1" = ".-v" -o ".$1" = ."--version" ]; then
+    echo "GNU shtool 1.4.9 (16-Apr-2000)"
+    exit 0
+fi
+if [ ".$1" = ".-d" -o ".$1" = ."--debug" ]; then
+    shift
+    set -x
+fi
+name=`echo "$0" | sed -e 's;.*/\([^/]*\)$;\1;' -e 's;-sh$;;' -e 's;\.sh$;;'`
+case "$name" in
+    echo|install|mkdir )
+        #   implicit tool command selection
+        tool="$name"
+        ;;
+    * )
+        #   explicit tool command selection
+        tool="$1"
+        shift
+        ;;
+esac
+arg_spec=""
+opt_spec=""
+gen_tmpfile=no
+
+##
+##  DISPATCH INTO SCRIPT PROLOG
+##
+
+case $tool in
+    echo )
+        str_tool="echo"
+        str_usage="[-n] [-e] [<str> ...]"
+        arg_spec="0+"
+        opt_spec="n.e."
+        opt_n=no
+        opt_e=no
+        ;;
+    install )
+        str_tool="install"
+        str_usage="[-v] [-t] [-c] [-C] [-s] [-m<mode>] [-o<owner>] [-g<group>] [-e<ext>] <file> <path>"
+        arg_spec="2="
+        opt_spec="v.t.c.C.s.m:o:g:e:"
+        opt_v=no
+        opt_t=no
+        opt_c=no
+        opt_C=no
+        opt_s=no
+        opt_m=""
+        opt_o=""
+        opt_g=""
+        opt_e=""
+        ;;
+    mkdir )
+        str_tool="mkdir"
+        str_usage="[-t] [-f] [-p] [-m<mode>] <dir> [<dir> ...]"
+        arg_spec="1+"
+        opt_spec="t.f.p.m:"
+        opt_t=no
+        opt_f=no
+        opt_p=no
+        opt_m=""
+        ;;
+    -* )
+        echo "$0:Error: unknown option \`$tool'" 2>&1
+        echo "$0:Hint:  run \`$0 -h' for usage" 2>&1
+        exit 1
+        ;;
+    * )
+        echo "$0:Error: unknown command \`$tool'" 2>&1
+        echo "$0:Hint:  run \`$0 -h' for usage" 2>&1
+        exit 1
+        ;;
+esac
+
+##
+##  COMMON UTILITY CODE
+##
+
+#   determine name of tool
+if [ ".$tool" != . ]; then
+    #   used inside shtool script
+    toolcmd="$0 $tool"
+    toolcmdhelp="shtool $tool"
+    msgprefix="shtool:$tool"
+else
+    #   used as standalone script
+    toolcmd="$0"
+    toolcmdhelp="sh $0"
+    msgprefix="$str_tool"
+fi
+
+#   parse argument specification string
+eval `echo $arg_spec |\
+      sed -e 's/^\([0-9]*\)\([+=]\)/arg_NUMS=\1; arg_MODE=\2/'`
+
+#   parse option specification string
+eval `echo h.$opt_spec |\
+      sed -e 's/\([a-zA-Z0-9]\)\([.:+]\)/opt_MODE_\1=\2;/g'`
+
+#   interate over argument line
+opt_PREV=''
+while [ $# -gt 0 ]; do
+    #   special option stops processing
+    if [ ".$1" = ".--" ]; then
+        shift
+        break
+    fi
+
+    #   determine option and argument
+    opt_ARG_OK=no
+    if [ ".$opt_PREV" != . ]; then
+        #   merge previous seen option with argument
+        opt_OPT="$opt_PREV"
+        opt_ARG="$1"
+        opt_ARG_OK=yes
+        opt_PREV=''
+    else
+        #   split argument into option and argument
+        case "$1" in
+            -[a-zA-Z0-9]*)
+                eval `echo "x$1" |\
+                      sed -e 's/^x-\([a-zA-Z0-9]\)/opt_OPT="\1";/' \
+                          -e 's/";\(.*\)$/"; opt_ARG="\1"/'`
+                ;;
+            -[a-zA-Z0-9])
+                opt_OPT=`echo "x$1" | cut -c3-`
+                opt_ARG=''
+                ;;
+            *)
+                break
+                ;;
+        esac
+    fi
+
+    #   eat up option
+    shift
+
+    #   determine whether option needs an argument
+    eval "opt_MODE=\$opt_MODE_${opt_OPT}"
+    if [ ".$opt_ARG" = . -a ".$opt_ARG_OK" != .yes ]; then
+        if [ ".$opt_MODE" = ".:" -o ".$opt_MODE" = ".+" ]; then
+            opt_PREV="$opt_OPT"
+            continue
+        fi
+    fi
+
+    #   process option
+    case $opt_MODE in
+        '.' )
+            #   boolean option
+            eval "opt_${opt_OPT}=yes"
+            ;;
+        ':' )
+            #   option with argument (multiple occurances override)
+            eval "opt_${opt_OPT}=\"\$opt_ARG\""
+            ;;
+        '+' )
+            #   option with argument (multiple occurances append)
+            eval "opt_${opt_OPT}=\"\$opt_${opt_OPT} \$opt_ARG\""
+            ;;
+        * )
+            echo "$msgprefix:Error: unknown option: \`-$opt_OPT'" 1>&2
+            echo "$msgprefix:Hint:  run \`$toolcmdhelp -h' or \`man shtool' for details" 1>&2
+            exit 1
+            ;;
+    esac
+done
+if [ ".$opt_PREV" != . ]; then
+    echo "$msgprefix:Error: missing argument to option \`-$opt_PREV'" 1>&2
+    echo "$msgprefix:Hint:  run \`$toolcmdhelp -h' or \`man shtool' for details" 1>&2
+    exit 1
+fi
+
+#   process help option
+if [ ".$opt_h" = .yes ]; then
+    echo "Usage: $toolcmdhelp $str_usage"
+    exit 0
+fi
+
+#   complain about incorrect number of arguments
+case $arg_MODE in
+    '=' )
+        if [ $# -ne $arg_NUMS ]; then
+            echo "$msgprefix:Error: invalid number of arguments (exactly $arg_NUMS expected)" 1>&2
+            echo "$msgprefix:Hint:  run \`$toolcmd -h' or \`man shtool' for details" 1>&2
+            exit 1
+        fi
+        ;;
+    '+' )
+        if [ $# -lt $arg_NUMS ]; then
+            echo "$msgprefix:Error: invalid number of arguments (at least $arg_NUMS expected)" 1>&2
+            echo "$msgprefix:Hint:  run \`$toolcmd -h' or \`man shtool' for details" 1>&2
+            exit 1
+        fi
+        ;;
+esac
+
+#   establish a temporary file on request
+if [ ".$gen_tmpfile" = .yes ]; then
+    if [ ".$TMPDIR" != . ]; then
+        tmpdir="$TMPDIR"
+    elif [ ".$TEMPDIR" != . ]; then
+        tmpdir="$TEMPDIR"
+    else
+        tmpdir="/tmp"
+    fi
+    tmpfile="$tmpdir/.shtool.$$"
+    rm -f $tmpfile >/dev/null 2>&1
+    touch $tmpfile
+fi
+
+##
+##  DISPATCH INTO SCRIPT BODY
+##
+
+case $tool in
+
+echo )
+    ##
+    ##  echo -- Print string with optional construct expansion
+    ##  Copyright (c) 1998-2000 Ralf S. Engelschall <rse@engelschall.com>
+    ##  Originally written for WML as buildinfo
+    ##
+    
+    text="$*"
+    
+    #   check for broken escape sequence expansion
+    seo=''
+    bytes=`echo '\1' | wc -c | awk '{ printf("%s", $1); }'`
+    if [ ".$bytes" != .3 ]; then
+        bytes=`echo -E '\1' | wc -c | awk '{ printf("%s", $1); }'`
+        if [ ".$bytes" = .3 ]; then
+            seo='-E'
+        fi
+    fi
+    
+    #   check for existing -n option (to suppress newline)
+    minusn=''
+    bytes=`echo -n 123 2>/dev/null | wc -c | awk '{ printf("%s", $1); }'`
+    if [ ".$bytes" = .3 ]; then
+        minusn='-n'
+    fi
+    
+    #   determine terminal bold sequence
+    term_bold='' 
+    term_norm=''
+    if [ ".$opt_e" = .yes -a ".`echo $text | egrep '%[Bb]'`" != . ]; then
+        case $TERM in
+            #   for the most important terminal types we directly know the sequences
+            xterm|xterm*|vt220|vt220*)
+                term_bold=`awk 'BEGIN { printf("%c%c%c%c", 27, 91, 49, 109); }' </dev/null 2>/dev/null`
+                term_norm=`awk 'BEGIN { printf("%c%c%c", 27, 91, 109); }' </dev/null 2>/dev/null`
+                ;;
+            vt100|vt100*)
+                term_bold=`awk 'BEGIN { printf("%c%c%c%c%c%c", 27, 91, 49, 109, 0, 0); }' </dev/null 2>/dev/null`
+                term_norm=`awk 'BEGIN { printf("%c%c%c%c%c", 27, 91, 109, 0, 0); }' </dev/null 2>/dev/null`
+                ;;
+            #   for all others, we try to use a possibly existing `tput' or `tcout' utility
+            * )
+                paths=`echo $PATH | sed -e 's/:/ /g'`
+                for tool in tput tcout; do
+                    for dir in $paths; do
+                        if [ -r "$dir/$tool" ]; then
+                            for seq in bold md smso; do # 'smso' is last
+                                bold="`$dir/$tool $seq 2>/dev/null`"
+                                if [ ".$bold" != . ]; then
+                                    term_bold="$bold"
+                                    break
+                                fi
+                            done
+                            if [ ".$term_bold" != . ]; then
+                                for seq in sgr0 me rmso reset; do # 'reset' is last
+                                    norm="`$dir/$tool $seq 2>/dev/null`"
+                                    if [ ".$norm" != . ]; then
+                                        term_norm="$norm"
+                                        break
+                                    fi
+                                done
+                            fi
+                            break
+                        fi
+                    done
+                    if [ ".$term_bold" != . -a ".$term_norm" != . ]; then
+                        break;
+                    fi
+                done
+                ;;
+        esac
+        if [ ".$term_bold" = . -o ".$term_norm" = . ]; then
+            echo "$msgprefix:Warning: unable to determine terminal sequence for bold mode" 1>&2
+        fi
+    fi
+    
+    #   determine user name
+    username=''
+    if [ ".$opt_e" = .yes -a ".`echo $text | egrep '%[uU]'`" != . ]; then
+        username="$LOGNAME"
+        if [ ".$username" = . ]; then
+            username="$USER"
+            if [ ".$username" = . ]; then
+                username="`(whoami) 2>/dev/null |\
+                           awk '{ printf("%s", $1); }'`"
+                if [ ".$username" = . ]; then
+                    username="`(who am i) 2>/dev/null |\
+                               awk '{ printf("%s", $1); }'`"
+                    if [ ".$username" = . ]; then
+                        username='unknown'
+                    fi
+                fi
+            fi
+        fi
+    fi
+    
+    #   determine user id
+    userid=''
+    if [ ".$opt_e" = .yes -a ".`echo $text | egrep '%U'`" != . ]; then
+        userid="`(id -u) 2>/dev/null`"
+        if [ ".$userid" = . ]; then
+            str="`(id) 2>/dev/null`"
+            if [ ".`echo $str | grep '^uid[ 	]*=[ 	]*[0-9]*('`" != . ]; then
+                userid=`echo $str | sed -e 's/^uid[ 	]*=[ 	]*//' -e 's/(.*//'`
+            fi
+            if [ ".$userid" = . ]; then
+                userid=`egrep "^${username}:" /etc/passwd 2>/dev/null | \
+                        sed -e 's/[^:]*:[^:]*://' -e 's/:.*$//'`
+                if [ ".$userid" = . ]; then
+                    userid=`(ypcat passwd) 2>/dev/null |
+                            egrep "^${username}:" | \
+                            sed -e 's/[^:]*:[^:]*://' -e 's/:.*$//'`
+                    if [ ".$userid" = . ]; then
+                        userid='?'
+                    fi
+                fi
+            fi
+        fi
+    fi
+    
+    #   determine host name
+    hostname=''
+    if [ ".$opt_e" = .yes -a ".`echo $text | egrep '%h'`" != . ]; then
+        hostname="`(uname -n) 2>/dev/null |\
+                   awk '{ printf("%s", $1); }'`"
+        if [ ".$hostname" = . ]; then
+            hostname="`(hostname) 2>/dev/null |\
+                       awk '{ printf("%s", $1); }'`"
+            if [ ".$hostname" = . ]; then
+                hostname='unknown'
+            fi
+        fi
+        case $hostname in
+            *.* )
+                domainname=".`echo $hostname | cut -d. -f2-`"
+                hostname="`echo $hostname | cut -d. -f1`"
+                ;;
+        esac
+    fi
+    
+    #   determine domain name
+    domainname=''
+    if [ ".$opt_e" = .yes -a ".`echo $text | egrep '%d'`" != . ]; then
+        if [ ".$domainname" = . ]; then
+            if [ -f /etc/resolv.conf ]; then
+                domainname="`egrep '^[ 	]*domain' /etc/resolv.conf | head -1 |\
+                             sed -e 's/.*domain//' \
+                                 -e 's/^[ 	]*//' -e 's/^ *//' -e 's/^	*//' \
+                                 -e 's/^\.//' -e 's/^/./' |\
+                             awk '{ printf("%s", $1); }'`"
+                if [ ".$domainname" = . ]; then
+                    domainname="`egrep '^[ 	]*search' /etc/resolv.conf | head -1 |\
+                                 sed -e 's/.*search//' \
+                                     -e 's/^[ 	]*//' -e 's/^ *//' -e 's/^	*//' \
+                                     -e 's/ .*//' -e 's/	.*//' \
+                                     -e 's/^\.//' -e 's/^/./' |\
+                                 awk '{ printf("%s", $1); }'`"
+                fi
+            fi
+        fi
+    fi
+    
+    #   determine current time
+    time_day=''
+    time_month=''
+    time_year=''
+    time_monthname=''
+    if [ ".$opt_e" = .yes -a ".`echo $text | egrep '%[DMYm]'`" != . ]; then
+        time_day=`date '+%d'`
+        time_month=`date '+%m'`
+        time_year=`date '+%Y' 2>/dev/null`
+        if [ ".$time_year" = . ]; then
+            time_year=`date '+%y'`
+            case $time_year in
+                [5-9][0-9]) time_year="19$time_year" ;;
+                [0-4][0-9]) time_year="20$time_year" ;;
+            esac
+        fi
+        case $time_month in
+            1|01) time_monthname='Jan' ;;
+            2|02) time_monthname='Feb' ;;
+            3|03) time_monthname='Mar' ;;
+            4|04) time_monthname='Apr' ;;
+            5|05) time_monthname='May' ;;
+            6|06) time_monthname='Jun' ;;
+            7|07) time_monthname='Jul' ;;
+            8|08) time_monthname='Aug' ;;
+            9|09) time_monthname='Sep' ;;
+              10) time_monthname='Oct' ;;
+              11) time_monthname='Nov' ;;
+              12) time_monthname='Dec' ;;
+        esac
+    fi
+    
+    #   expand special ``%x'' constructs
+    if [ ".$opt_e" = .yes ]; then
+        text=`echo $seo "$text" |\
+              sed -e "s/%B/${term_bold}/g" \
+                  -e "s/%b/${term_norm}/g" \
+                  -e "s/%u/${username}/g" \
+                  -e "s/%U/${userid}/g" \
+                  -e "s/%h/${hostname}/g" \
+                  -e "s/%d/${domainname}/g" \
+                  -e "s/%D/${time_day}/g" \
+                  -e "s/%M/${time_month}/g" \
+                  -e "s/%Y/${time_year}/g" \
+                  -e "s/%m/${time_monthname}/g" 2>/dev/null`
+    fi
+    
+    #   create output
+    if [ .$opt_n = .no ]; then
+        echo $seo "$text"
+    else
+        #   the harder part: echo -n is best, because
+        #   awk may complain about some \xx sequences.
+        if [ ".$minusn" != . ]; then
+            echo $seo $minusn "$text"
+        else
+            echo dummy | awk '{ printf("%s", TEXT); }' TEXT="$text"
+        fi
+    fi
+    ;;
+
+install )
+    ##
+    ##  install -- Install a program, script or datafile
+    ##  Copyright (c) 1997-2000 Ralf S. Engelschall <rse@engelschall.com>
+    ##  Originally written for shtool
+    ##
+    
+    src="$1"
+    dst="$2"
+    
+    #  If destination is a directory, append the input filename
+    if [ -d $dst ]; then
+        dst=`echo "$dst" | sed -e 's:/$::'`
+        dstfile=`echo "$src" | sed -e 's;.*/\([^/]*\)$;\1;'`
+        dst="$dst/$dstfile"
+    fi
+    
+    #  Add a possible extension to src and dst
+    if [ ".$opt_e" != . ]; then
+        src="$src$opt_e"
+        dst="$dst$opt_e"
+    fi
+    
+    #  Check for correct arguments
+    if [ ".$src" = ".$dst" ]; then
+        echo "$msgprefix:Error: source and destination are the same" 1>&2
+        exit 1
+    fi
+    
+    #  Make a temp file name in the destination directory
+    dstdir=`echo $dst | sed -e 's;[^/]*$;;' -e 's;\(.\)/$;\1;' -e 's;^$;.;'`
+    dsttmp="$dstdir/#INST@$$#"
+    
+    #  Verbosity
+    if [ ".$opt_v" = .yes ]; then
+        echo "$src -> $dst" 1>&2
+    fi
+    
+    #  Copy or move the file name to the temp name
+    #  (because we might be not allowed to change the source)
+    if [ ".$opt_C" = .yes ]; then
+        opt_c=yes
+    fi
+    if [ ".$opt_c" = .yes ]; then
+        if [ ".$opt_t" = .yes ]; then
+            echo "cp $src $dsttmp" 1>&2
+        fi
+        cp $src $dsttmp || exit $?
+    else
+        if [ ".$opt_t" = .yes ]; then
+            echo "mv $src $dsttmp" 1>&2
+        fi
+        mv $src $dsttmp || exit $?
+    fi
+    
+    #  Adjust the target file
+    #  (we do chmod last to preserve setuid bits)
+    if [ ".$opt_s" = .yes ]; then
+        if [ ".$opt_t" = .yes ]; then
+            echo "strip $dsttmp" 1>&2
+        fi
+        strip $dsttmp || exit $?
+    fi
+    if [ ".$opt_o" != . ]; then
+        if [ ".$opt_t" = .yes ]; then
+            echo "chown $opt_o $dsttmp" 1>&2
+        fi
+        chown $opt_o $dsttmp || exit $?
+    fi
+    if [ ".$opt_g" != . ]; then
+        if [ ".$opt_t" = .yes ]; then
+            echo "chgrp $opt_g $dsttmp" 1>&2
+        fi
+        chgrp $opt_g $dsttmp || exit $?
+    fi
+    if [ ".$opt_m" != . ]; then
+        if [ ".$opt_t" = .yes ]; then
+            echo "chmod $opt_m $dsttmp" 1>&2
+        fi
+        chmod $opt_m $dsttmp || exit $?
+    fi
+    
+    #   Determine whether to do a quick install
+    #   (has to be done _after_ the strip was already done)
+    quick=no
+    if [ ".$opt_C" = .yes ]; then
+        if [ -r $dst ]; then
+            if cmp -s $src $dst; then
+                quick=yes
+            fi
+        fi
+    fi
+    
+    #   Finally install the file to the real destination
+    if [ $quick = yes ]; then
+        if [ ".$opt_t" = .yes ]; then
+            echo "rm -f $dsttmp" 1>&2
+        fi
+        rm -f $dsttmp
+    else
+        if [ ".$opt_t" = .yes ]; then
+            echo "rm -f $dst && mv $dsttmp $dst" 1>&2
+        fi
+        rm -f $dst && mv $dsttmp $dst
+    fi
+    ;;
+
+mkdir )
+    ##
+    ##  mkdir -- Make one or more directories
+    ##  Copyright (c) 1996-2000 Ralf S. Engelschall <rse@engelschall.com>
+    ##  Originally written for public domain by Noah Friedman <friedman@prep.ai.mit.edu>
+    ##  Cleaned up and enhanced for shtool
+    ##
+    
+    errstatus=0
+    for p in ${1+"$@"}; do
+        #   if the directory already exists...
+        if [ -d "$p" ]; then
+            if [ ".$opt_f" = .no ] && [ ".$opt_p" = .no ]; then
+                echo "$msgprefix:Error: directory already exists: $p" 1>&2
+                errstatus=1
+                break
+            else
+                continue
+            fi
+        fi
+        #   if the directory has to be created...
+        if [ ".$opt_p" = .no ]; then
+            if [ ".$opt_t" = .yes ]; then
+                echo "mkdir $p" 1>&2
+            fi
+            mkdir $p || errstatus=$?
+        else
+            #   the smart situation
+            set fnord `echo ":$p" |\
+                       sed -e 's/^:\//%/' \
+                           -e 's/^://' \
+                           -e 's/\// /g' \
+                           -e 's/^%/\//'`
+            shift
+            pathcomp=''
+            for d in ${1+"$@"}; do
+                pathcomp="$pathcomp$d"
+                case "$pathcomp" in
+                    -* ) pathcomp="./$pathcomp" ;;
+                esac
+                if [ ! -d "$pathcomp" ]; then
+                    if [ ".$opt_t" = .yes ]; then
+                        echo "mkdir $pathcomp" 1>&2
+                    fi
+                    mkdir $pathcomp || errstatus=$?
+                    if [ ".$opt_m" != . ]; then
+                        if [ ".$opt_t" = .yes ]; then
+                            echo "chmod $opt_m $pathcomp" 1>&2
+                        fi
+                        chmod $opt_m $pathcomp || errstatus=$?
+                    fi
+                fi
+                pathcomp="$pathcomp/"
+            done
+        fi
+    done
+    exit $errstatus
+    ;;
+
+esac
+
+exit 0
+
+##EOF##

aux/binpac/src/Makefile.am

@@ -0,0 +1,62 @@
+## Process this file with automake to produce Makefile.in
+
+AM_YFLAGS = -d -t -v
+AM_CPPFLAGS = -W -Wall -Wno-unused
+
+noinst_PROGRAMS = binpac
+
+binpac_SOURCES = \
+	pac_scan.ll pac_parse.yy \
+	pac_action.cc \
+	pac_analyzer.cc \
+	pac_array.cc \
+	pac_attr.cc \
+	pac_btype.cc \
+	pac_case.cc \
+	pac_conn.cc \
+	pac_context.cc \
+	pac_cstr.cc \
+	pac_datadep.cc \
+	pac_dataptr.cc \
+	pac_dataunit.cc \
+	pac_decl.cc \
+	pac_embedded.cc \
+	pac_enum.cc \
+	pac_expr.cc \
+	pac_exttype.cc \
+	pac_field.cc \
+	pac_flow.cc \
+	pac_func.cc \
+	pac_id.cc \
+	pac_inputbuf.cc \
+	pac_let.cc \
+	pac_param.cc \
+	pac_paramtype.cc \
+	pac_primitive.cc \
+	pac_record.cc \
+	pac_redef.cc \
+	pac_regex.cc \
+	pac_state.cc \
+	pac_strtype.cc \
+	pac_type.cc \
+	pac_typedecl.cc \
+	pac_withinput.cc \
+	pac_output.cc pac_utils.cc pac_exception.cc \
+	pac_main.cc \
+	pac_action.h pac_analyzer.h pac_array.h pac_attr.h pac_btype.h \
+	pac_case.h pac_cclass.h pac_common.h pac_conn.h pac_context.h \
+	pac_cstr.h pac_ctype.h pac_datadep.h pac_dataptr.h pac_dataunit.h \
+	pac_dbg.h pac_decl-inl.h pac_decl.h pac_embedded.h pac_enum.h \
+	pac_exception.h pac_expr.h pac_exttype.h pac_field.h pac_flow.h \
+	pac_func.h pac_id.h pac_inputbuf.h pac_let.h pac_number.h \
+	pac_output.h pac_param.h pac_paramtype.h pac_parse.h pac_primitive.h \
+	pac_record.h pac_redef.h pac_regex.h pac_state.h pac_strtype.h \
+	pac_type.h pac_typedecl.h pac_utils.h pac_varfield.h pac_withinput.h
+
+EXTRA_DIST = pac_expr.def pac_type.def pac_externtype.def
+
+DISTCLEANFILES = pac_parse.cc pac_parse.h pac_scan.cc y.output
+
+# Manual rules below:
+
+pac_scan.o:	pac_parse.h

aux/binpac/src/pac_action.cc

@@ -0,0 +1,119 @@
+#include "pac_embedded.h"
+#include "pac_exception.h"
+#include "pac_id.h"
+#include "pac_output.h"
+#include "pac_type.h"
+#include "pac_typedecl.h"
+#include "pac_utils.h"
+
+#include "pac_action.h"
+
+AnalyzerAction::AnalyzerAction(ID *action_id, 
+	                       When when, 
+	                       ActionParam *param, 
+	                       EmbeddedCode *code)
+	: AnalyzerElement(ACTION),
+	  action_id_(action_id),
+	  when_(when),
+	  param_(param),
+	  code_(code),
+	  analyzer_(0) 
+	{
+	}
+
+AnalyzerAction::~AnalyzerAction()
+	{
+	delete action_id_;
+	delete param_;
+	delete code_;
+	}
+
+string AnalyzerAction::action_function() const
+	{
+	return strfmt("Action_%s", action_id_->Name());
+	}
+
+void AnalyzerAction::InstallHook(AnalyzerDecl *analyzer)
+	{
+	ASSERT(0);
+	analyzer_ = analyzer;
+	// param_->MainDataType()->InstallAction(this);
+	}
+
+void AnalyzerAction::GenCode(Output *out_h, Output *out_cc, AnalyzerDecl *decl)
+	{
+	Env action_func_env(decl->env(), this);
+	action_func_env.AddID(param_->id(), 
+	                      TEMP_VAR, 
+	                      param_->DataType());
+	action_func_env.SetEvaluated(param_->id());
+
+	string action_func_proto = 
+		strfmt("%s(%s)", 
+		       action_function().c_str(), 
+		       ParamDecls(&action_func_env).c_str());
+
+	out_h->println("void %s;", action_func_proto.c_str());
+
+	out_cc->println("void %s::%s", 
+	                decl->class_name().c_str(), 
+	                action_func_proto.c_str());
+	out_cc->inc_indent();
+	out_cc->println("{");
+
+	code_->GenCode(out_cc, &action_func_env);
+
+	out_cc->println("");
+	out_cc->println("}");
+	out_cc->dec_indent();
+	out_cc->println("");
+	}
+
+string AnalyzerAction::ParamDecls(Env *env) const
+	{
+	return param_->DeclStr(env);
+	}
+
+Type *ActionParam::MainDataType() const
+	{
+	// Note: this is not equal to DataType()
+	Type *main_type = TypeDecl::LookUpType(type()->type_id());
+
+	if ( ! main_type )
+		{
+		throw Exception(type()->type_id(), 
+		                "type not defined");
+		}
+
+	return main_type;
+	}
+
+Type *ActionParam::DataType() const
+	{
+	Type *main_type = MainDataType();
+
+	if ( ! type()->field_id() )
+		{
+		return main_type;
+		}
+	else
+		{
+		Type *member_type = 
+			main_type->MemberDataType(type()->field_id());
+		if ( ! member_type )
+			{
+			throw Exception(type()->field_id(),
+				fmt("cannot find member type for `%s.%s'",
+				    type()->type_id()->Name(),
+				    type()->field_id()->Name()));
+			}
+		return member_type;
+		}
+	}
+
+string ActionParam::DeclStr(Env *env) const
+	{
+	return strfmt("%s %s", 
+	              DataType()->DataTypeStr().c_str(), 
+	              env->LValue(id()));
+	}

aux/binpac/src/pac_action.h

@@ -0,0 +1,74 @@
+#ifndef pac_action_h
+#define pac_action_h
+
+// Classes representing analyzer actions.
+
+#include "pac_common.h"
+#include "pac_analyzer.h"
+
+class AnalyzerAction : public AnalyzerElement
+{
+public:
+	enum When { BEFORE, AFTER };
+
+	AnalyzerAction(ID *action_id, 
+	               When when, 
+	               ActionParam *param, 
+	               EmbeddedCode *code);
+
+	~AnalyzerAction();
+
+	When when() const		{ return when_; }
+	ActionParam *param() const	{ return param_; }
+	AnalyzerDecl *analyzer() const	{ return analyzer_; }
+	string action_function() const;
+
+	// Generate function prototype and code for the action
+	void GenCode(Output *out_h, Output *out_cc, AnalyzerDecl *decl);
+
+	// Install the hook at the corresponding data type parsing
+	// function to invoke the action.
+	void InstallHook(AnalyzerDecl *analyzer);
+
+private:
+	string ParamDecls(Env *env) const;
+
+	ID *action_id_;
+	When when_;
+	ActionParam *param_;
+	EmbeddedCode *code_;
+	AnalyzerDecl *analyzer_;
+};
+
+class ActionParam
+{
+public:
+	ActionParam(const ID *id, ActionParamType *type)
+		: id_(id), type_(type) {}
+
+	const ID *id() const		{ return id_; }
+	ActionParamType *type() const	{ return type_; }
+
+	Type *MainDataType() const;
+	Type *DataType() const;
+	string DeclStr(Env *env) const;
+
+private:
+	const ID *id_;
+	ActionParamType *type_;
+};
+
+class ActionParamType
+{
+public:
+	ActionParamType(const ID *type_id, const ID *field_id = 0)
+		: type_id_(type_id), field_id_(field_id) {}
+
+	const ID *type_id() const	{ return type_id_; }
+	const ID *field_id() const	{ return field_id_; }
+
+protected:
+	const ID *type_id_, *field_id_;
+};
+
+#endif  // pac_action_h

aux/binpac/src/pac_analyzer.cc

@@ -0,0 +1,358 @@
+#include "pac_action.h"
+#include "pac_context.h"
+#include "pac_embedded.h"
+#include "pac_exception.h"
+#include "pac_expr.h"
+#include "pac_flow.h"
+#include "pac_func.h"
+#include "pac_output.h"
+#include "pac_param.h"
+#include "pac_paramtype.h"
+#include "pac_state.h"
+#include "pac_type.h"
+#include "pac_varfield.h"
+
+#include "pac_analyzer.h"
+
+AnalyzerDecl::AnalyzerDecl(ID *id, 
+                           DeclType decl_type,
+                           ParamList *params)
+	: TypeDecl(id, params, new DummyType())
+	{
+	decl_type_ = decl_type;
+
+	statevars_ = new StateVarList();
+	actions_ = new AnalyzerActionList();
+	helpers_ = new AnalyzerHelperList();
+	functions_ = new FunctionList();
+
+	constructor_helpers_ = new AnalyzerHelperList();
+	destructor_helpers_ = new AnalyzerHelperList();
+	eof_helpers_ = new AnalyzerHelperList();
+
+	SetAnalyzerContext();
+
+	env_ = 0;
+	}
+
+AnalyzerDecl::~AnalyzerDecl()
+	{
+	delete_list(StateVarList, statevars_);
+	delete_list(AnalyzerActionList, actions_);
+	delete_list(AnalyzerHelperList, helpers_);
+	delete_list(FunctionList, functions_);
+	delete_list(ParamList, params_);
+	delete_list(AnalyzerHelperList, constructor_helpers_);
+	delete_list(AnalyzerHelperList, destructor_helpers_);
+	delete_list(AnalyzerHelperList, eof_helpers_);
+	}
+
+void AnalyzerDecl::AddElements(AnalyzerElementList *elemlist)
+	{
+	ASSERT(! env_);
+	foreach(i, AnalyzerElementList, elemlist)
+		{
+		AnalyzerElement *elem = *i;
+		switch ( elem->type() )
+			{
+			case AnalyzerElement::STATE:
+				{
+				ASSERT(0);
+				AnalyzerState *state_elem = 
+					(AnalyzerState *) elem;
+				statevars_->insert(
+					statevars_->end(),
+					state_elem->statevars()->begin(),
+					state_elem->statevars()->end());
+				}
+				break;
+			case AnalyzerElement::ACTION:
+				{
+				ASSERT(0);
+				AnalyzerAction *action_elem = 
+					(AnalyzerAction *) elem;
+				actions_->push_back(action_elem);
+				}
+				break;
+			case AnalyzerElement::HELPER:
+				{
+				AnalyzerHelper *helper_elem = 
+					(AnalyzerHelper *) elem;
+
+				switch ( helper_elem->helper_type() )
+				        {
+					case AnalyzerHelper::INIT_CODE:
+					        constructor_helpers_->push_back(helper_elem);
+						break;
+					case AnalyzerHelper::CLEANUP_CODE:
+ 				                destructor_helpers_->push_back(helper_elem);
+						break;
+					case AnalyzerHelper::EOF_CODE:
+					        eof_helpers_->push_back(helper_elem);
+						break;
+					default:
+					        helpers_->push_back(helper_elem);
+					}
+				}
+				break;
+			case AnalyzerElement::FUNCTION:
+				{
+				AnalyzerFunction *func_elem = 
+					(AnalyzerFunction *) elem;
+				Function *func = func_elem->function();
+				func->set_analyzer_decl(this);
+				functions_->push_back(func);
+				}
+				break;
+			case AnalyzerElement::FLOW:
+				{
+				AnalyzerFlow *flow_elem = 
+					(AnalyzerFlow *) elem;
+				ProcessFlowElement(flow_elem);
+				}
+				break;
+			case AnalyzerElement::DATAUNIT:
+				{
+				AnalyzerDataUnit *dataunit_elem = 
+					(AnalyzerDataUnit *) elem;
+				ProcessDataUnitElement(dataunit_elem);
+				}
+				break;
+			}
+		}
+	}
+
+string AnalyzerDecl::class_name() const
+	{ 
+	return id_->Name(); 
+	}
+
+void AnalyzerDecl::Prepare()
+	{
+	TypeDecl::Prepare();
+
+	ASSERT(statevars_->empty());
+	ASSERT(actions_->empty());
+
+	foreach(i, FunctionList, functions_)
+		{
+		Function *function = *i;
+		function->Prepare(env_);
+		}
+	foreach(i, StateVarList, statevars_)
+		{
+		StateVar *statevar = *i;
+		env_->AddID(statevar->id(), STATE_VAR, statevar->type());
+		}
+	foreach(i, AnalyzerActionList, actions_)
+		{
+		AnalyzerAction *action = *i;
+		action->InstallHook(this);
+		}
+	}
+
+void AnalyzerDecl::GenForwardDeclaration(Output* out_h)
+	{
+	out_h->println("class %s;", class_name().c_str());
+	foreach(i, FunctionList, functions_)
+		{
+		Function *function = *i;
+		function->GenForwardDeclaration(out_h);
+		}
+	}
+
+void AnalyzerDecl::GenActions(Output *out_h, Output *out_cc)
+	{
+	foreach(i, AnalyzerActionList, actions_)
+		{
+		(*i)->GenCode(out_h, out_cc, this);
+		}
+	}
+
+void AnalyzerDecl::GenHelpers(Output *out_h, Output *out_cc)
+	{
+	foreach(i, AnalyzerHelperList, helpers_)
+		{
+		(*i)->GenCode(out_h, out_cc, this);
+		}
+	}
+
+void AnalyzerDecl::GenPubDecls(Output *out_h, Output *out_cc)
+	{
+	TypeDecl::GenPubDecls(out_h, out_cc);
+
+	GenProcessFunc(out_h, out_cc);
+	GenGapFunc(out_h, out_cc);
+	GenEOFFunc(out_h, out_cc);
+	out_h->println("");
+
+	if ( ! functions_->empty() )
+		{
+		out_h->println("// Functions");
+		GenFunctions(out_h, out_cc);
+		out_h->println("");
+		}
+
+	// TODO: export public state variables
+	}
+
+void AnalyzerDecl::GenPrivDecls(Output *out_h, Output *out_cc)
+	{
+	TypeDecl::GenPrivDecls(out_h, out_cc);
+
+	if ( ! helpers_->empty() )
+		{
+		out_h->println("");
+		out_h->println("// Additional members");
+		GenHelpers(out_h, out_cc);
+		}
+
+	// TODO: declare state variables
+	}
+
+void AnalyzerDecl::GenInitCode(Output *out_cc)
+	{
+	TypeDecl::GenInitCode(out_cc);
+	foreach(i, AnalyzerHelperList, constructor_helpers_)
+		{
+		(*i)->GenCode(0, out_cc, this);
+		}
+	}
+
+void AnalyzerDecl::GenCleanUpCode(Output *out_cc)
+	{
+	TypeDecl::GenCleanUpCode(out_cc);
+	foreach(i, AnalyzerHelperList, destructor_helpers_)
+		{
+		(*i)->GenCode(0, out_cc, this);
+		}
+	}
+
+void AnalyzerDecl::GenStateVarDecls(Output *out_h)
+	{
+	foreach(i, StateVarList, statevars_)
+		{
+		StateVar *var = *i;
+		var->GenDecl(out_h, env_);
+		}
+	}
+
+void AnalyzerDecl::GenStateVarSetFunctions(Output *out_h)
+	{
+	foreach(i, StateVarList, statevars_)
+		{
+		StateVar *var = *i;
+		var->GenSetFunction(out_h, env_);
+		}
+	}
+
+void AnalyzerDecl::GenStateVarInitCode(Output *out_cc)
+	{
+	foreach(i, StateVarList, statevars_)
+		{
+		StateVar *var = *i;
+		var->GenInitCode(out_cc, env_);
+		}
+	}
+
+void AnalyzerDecl::GenStateVarCleanUpCode(Output *out_cc)
+	{
+	foreach(i, StateVarList, statevars_)
+		{
+		StateVar *var = *i;
+		var->GenCleanUpCode(out_cc, env_);
+		}
+	}
+
+void AnalyzerDecl::GenFunctions(Output *out_h, Output *out_cc)
+	{
+	foreach(i, FunctionList, functions_)
+		{
+		Function *function = *i;
+		function->GenCode(out_h, out_cc);
+		}
+	}
+
+AnalyzerState::~AnalyzerState()
+	{
+	// Note: do not delete elements of statevars_, because they
+	// are referenced by the AnalyzerDecl.
+	delete statevars_;
+	}
+
+AnalyzerHelper::~AnalyzerHelper()
+	{
+	delete code_;
+	}
+
+void AnalyzerHelper::GenCode(Output *out_h, Output *out_cc, AnalyzerDecl *decl)
+	{
+	Output *out = 0;
+	switch ( helper_type_ )
+		{
+		case MEMBER_DECLS:
+			out = out_h;
+			break;
+		case INIT_CODE:
+		case CLEANUP_CODE:
+		case EOF_CODE:
+			out = out_cc;
+			break;
+		}
+	ASSERT(out);
+	code()->GenCode(out, decl->env());
+	}
+
+FlowField::FlowField(ID *flow_id, ParameterizedType *flow_type)
+	: Field(FLOW_FIELD, 
+		TYPE_NOT_TO_BE_PARSED | CLASS_MEMBER | PUBLIC_READABLE, 
+		flow_id, flow_type)
+	{
+	}
+
+void FlowField::GenInitCode(Output *out_cc, Env *env)
+	{
+	type_->GenPreParsing(out_cc, env);
+	}
+
+AnalyzerFlow::AnalyzerFlow(Direction dir, ID *type_id, ExprList *params)
+	: AnalyzerElement(FLOW),
+	  dir_(dir),
+	  type_id_(type_id)
+	{
+	if ( ! params )
+		params = new ExprList();
+
+	// Add "this" to the list of params
+	params->insert(params->begin(), new Expr(this_id->clone()));
+
+	ID *flow_id = ((dir == UP) ? upflow_id : downflow_id)->clone();
+
+	ParameterizedType *flow_type = new ParameterizedType(type_id_, params);
+
+	flow_field_ = new FlowField(flow_id, flow_type);
+
+	flow_decl_ = 0;
+	}
+
+AnalyzerFlow::~AnalyzerFlow()
+	{
+	delete flow_field_;
+	}
+
+FlowDecl *AnalyzerFlow::flow_decl()
+	{
+	DEBUG_MSG("Getting flow_decl for %s\n", type_id_->Name());
+	if ( ! flow_decl_ )
+		{
+		Decl *decl = Decl::LookUpDecl(type_id_);
+		if ( decl && decl->decl_type() == Decl::FLOW )
+			flow_decl_ = static_cast<FlowDecl *>(decl);
+		if ( ! flow_decl_ )
+			{
+			throw Exception(this, 
+			                "cannot find the flow declaration");
+			}
+		}
+	return flow_decl_;
+	}

aux/binpac/src/pac_analyzer.h

@@ -0,0 +1,168 @@
+#ifndef pac_analyzer_h
+#define pac_analyzer_h
+
+#include "pac_common.h"
+#include "pac_field.h"
+#include "pac_typedecl.h"
+
+class AnalyzerElement;
+class AnalyzerState;
+class AnalyzerAction;	// defined in pac_action.h
+class AnalyzerHelper;
+class AnalyzerFlow;
+class AnalyzerDataUnit;
+class AnalyzerFunction;
+class ConnDecl;
+class FlowDecl;
+typedef vector<AnalyzerHelper *> AnalyzerHelperList;
+typedef vector<Function *> FunctionList;
+
+class AnalyzerDecl : public TypeDecl
+{
+public:
+	AnalyzerDecl(ID *id, DeclType decl_type, ParamList *params);
+	~AnalyzerDecl();
+
+	void AddElements(AnalyzerElementList *elemlist);
+
+	void Prepare();
+	void GenForwardDeclaration(Output *out_h);
+	// void GenCode(Output *out_h, Output *out_cc);
+
+	void GenInitCode(Output *out_cc);
+	void GenCleanUpCode(Output *out_cc);
+
+	string class_name() const;
+	// string cookie_name() const;
+
+protected:
+	virtual void ProcessFlowElement(AnalyzerFlow *flow_elem) = 0;
+	virtual void ProcessDataUnitElement(AnalyzerDataUnit *dataunit_elem) = 0;
+
+	// Generate public/private declarations for member functions and 
+	// variables
+	void GenPubDecls(Output *out_h, Output *out_cc);
+	void GenPrivDecls(Output *out_h, Output *out_cc);
+
+	// Generate the NewData() function
+	virtual void GenProcessFunc(Output *out_h, Output *out_cc) = 0;
+
+	// Generate the NewGap() function
+	virtual void GenGapFunc(Output *out_h, Output *out_cc) = 0;
+
+	// Generate the FlowEOF() function
+	virtual void GenEOFFunc(Output *out_h, Output *out_cc) = 0;
+
+	// Generate the functions
+	void GenFunctions(Output *out_h, Output *out_cc);
+
+	// Generate the action functions
+	void GenActions(Output *out_h, Output *out_cc);
+
+	// Generate the helper code segments
+	void GenHelpers(Output *out_h, Output *out_cc);
+
+	// Generate declarations for state variables and their set functions
+	void GenStateVarDecls(Output *out_h);
+	void GenStateVarSetFunctions(Output *out_h);
+
+	// Generate code for initializing and cleaning up (including
+	// memory de-allocating) state variables
+	void GenStateVarInitCode(Output *out_cc);
+	void GenStateVarCleanUpCode(Output *out_cc);
+
+	StateVarList *statevars_;
+	AnalyzerActionList *actions_;
+	AnalyzerHelperList *helpers_;
+	FunctionList *functions_;
+
+	AnalyzerHelperList *constructor_helpers_;
+	AnalyzerHelperList *destructor_helpers_;
+	AnalyzerHelperList *eof_helpers_;
+};
+
+class AnalyzerElement : public Object
+{
+public:
+	enum ElementType { STATE, ACTION, FUNCTION, HELPER, FLOW, DATAUNIT };
+	AnalyzerElement(ElementType type)
+		: type_(type) {}
+	virtual ~AnalyzerElement() {}
+
+	ElementType type() const	{ return type_; }
+
+private:
+	ElementType type_;
+};
+
+// A collection of variables representing analyzer states.
+class AnalyzerState : public AnalyzerElement
+{
+public:
+	AnalyzerState(StateVarList *statevars)
+		: AnalyzerElement(STATE),
+		  statevars_(statevars) {}
+	~AnalyzerState();
+
+	StateVarList *statevars() const	{ return statevars_; }
+
+private:
+	StateVarList *statevars_;
+};
+
+// A collection of embedded C++ code
+class AnalyzerHelper : public AnalyzerElement
+{
+public:
+	enum Type {
+		MEMBER_DECLS,
+		INIT_CODE,
+		CLEANUP_CODE,
+		EOF_CODE,
+	};
+	AnalyzerHelper(Type helper_type, EmbeddedCode *code)
+		: AnalyzerElement(HELPER),
+		  helper_type_(helper_type), 
+		  code_(code) {}
+	~AnalyzerHelper();
+
+	Type helper_type() const	{ return helper_type_; }
+
+	void GenCode(Output *out_h, Output *out_cc, AnalyzerDecl *decl);
+
+	EmbeddedCode *code() const	{ return code_; }
+
+private:
+	Type helper_type_;
+	EmbeddedCode *code_;
+};
+
+// The type and parameters of (uni-directional) flows of a connection.
+
+class FlowField : public Field
+{
+public:
+	FlowField(ID *flow_id, ParameterizedType *flow_type);
+	void GenInitCode(Output *out, Env *env);
+};
+
+class AnalyzerFlow : public AnalyzerElement
+{
+public:
+	enum Direction { UP, DOWN };
+	AnalyzerFlow(Direction dir, ID *type_id, ExprList *params);
+	~AnalyzerFlow();
+
+	Direction dir() const		{ return dir_; }
+	FlowField *flow_field() const	{ return flow_field_; }
+
+	FlowDecl *flow_decl();
+
+private:
+	Direction dir_;
+	ID *type_id_;
+	FlowField *flow_field_;
+	FlowDecl *flow_decl_;
+};
+
+#endif  // pac_analyzer_h

aux/binpac/src/pac_array.cc

@@ -0,0 +1,700 @@
+#include "pac_attr.h"
+#include "pac_dataptr.h"
+#include "pac_exception.h"
+#include "pac_expr.h"
+#include "pac_exttype.h"
+#include "pac_id.h"
+#include "pac_number.h"
+#include "pac_output.h"
+#include "pac_utils.h"
+#include "pac_varfield.h"
+
+#include "pac_array.h"
+
+ArrayType::ArrayType(Type *elemtype, Expr *length)
+	: Type(ARRAY), elemtype_(elemtype), length_(length)
+	{
+	init();
+
+	switch ( elemtype_->tot() )
+		{
+		case BUILTIN:
+		case PARAMETERIZED:
+		case STRING:
+		case EXTERN:
+			break;
+
+		case ARRAY:
+		case CASE:
+		case DUMMY:
+		case EMPTY:
+		case RECORD:
+		case UNDEF:
+			ASSERT(0);
+			break;
+		}
+	}
+
+void ArrayType::init()
+	{
+	arraylength_var_field_ = 0;
+	elem_it_var_field_ = 0;
+	elem_var_field_ = 0;
+	elem_dataptr_var_field_ = 0;
+	elem_input_var_field_ = 0;
+
+	elem_dataptr_until_expr_ = 0;
+
+	end_of_array_loop_label_ = "@@@";
+
+	vector_str_ = strfmt("vector<%s>", elemtype_->DataTypeStr().c_str());
+
+	datatype_str_ = strfmt("%s *", vector_str_.c_str());
+
+	attr_generic_until_expr_ = 0;
+	attr_until_element_expr_ = 0;
+	attr_until_input_expr_ = 0;
+	}
+
+ArrayType::~ArrayType()
+	{
+	delete arraylength_var_field_;
+	delete elem_it_var_field_;
+	delete elem_var_field_;
+	delete elem_dataptr_var_field_;
+	delete elem_input_var_field_;
+
+	delete elem_dataptr_until_expr_;
+	}
+
+Type *ArrayType::DoClone() const
+	{
+	Type *elemtype = elemtype_->Clone();
+	if ( ! elemtype )
+		return 0;
+	return new ArrayType(elemtype, length_);
+	}
+
+bool ArrayType::DefineValueVar() const
+	{
+	return true;
+	}
+
+string ArrayType::DataTypeStr() const
+	{
+	return datatype_str_;
+	}
+
+Type *ArrayType::ElementDataType() const
+	{
+	return elemtype_;
+	}
+
+string ArrayType::EvalElement(const string &array, const string &index) const
+	{
+	return strfmt("(*(%s))[%s]", array.c_str(), index.c_str());
+	}
+
+const ID *ArrayType::arraylength_var() const
+	{
+	return arraylength_var_field_ ? arraylength_var_field_->id() : 0;
+	}
+
+const ID *ArrayType::elem_it_var() const
+	{
+	return elem_it_var_field_ ? elem_it_var_field_->id() : 0;
+	}
+
+const ID *ArrayType::elem_var() const
+	{
+	return elem_var_field_ ? elem_var_field_->id() : 0;
+	}
+
+const ID *ArrayType::elem_dataptr_var() const
+	{
+	return elem_dataptr_var_field_ ? elem_dataptr_var_field_->id() : 0;
+	}
+
+const ID *ArrayType::elem_input_var() const
+	{
+	return elem_input_var_field_ ? elem_input_var_field_->id() : 0;
+	}
+
+void ArrayType::ProcessAttr(Attr *a)
+	{
+	Type::ProcessAttr(a);
+
+	switch ( a->type() )
+		{
+		case ATTR_RESTOFDATA:
+			{
+			if ( elemtype_->StaticSize(env()) != 1 )
+				{
+				throw Exception(elemtype_, 
+					"&restofdata can be applied"
+					" to only byte arrays");
+				}
+			if ( length_ )
+				{
+				throw Exception(length_, 
+					"&restofdata cannot be applied"
+					" to arrays with specified length");
+				}
+			attr_restofdata_ = true;
+			// As the array automatically extends to the end of 
+			// data, we do not have to check boundary.
+			SetBoundaryChecked();
+			}
+			break;
+
+		case ATTR_RESTOFFLOW:
+			attr_restofflow_ = true;
+			// TODO: handle &restofflow
+			break;
+
+		case ATTR_UNTIL:
+			{
+			bool ref_element = a->expr()->HasReference(element_macro_id);
+			bool ref_input = a->expr()->HasReference(input_macro_id);
+			if ( ref_element && ref_input )
+				{
+				throw Exception(a->expr(), 
+					"cannot reference both $element and $input "
+					"in the same &until---please separate them.");
+				}
+
+			if ( ref_element )
+				{
+				if ( attr_until_element_expr_ )
+					{
+					throw Exception(a->expr(), 
+						"multiple &until on $element");
+					}
+				attr_until_element_expr_ = a->expr();
+				}
+			else if ( ref_input )
+				{
+				if ( attr_until_input_expr_ )
+					{
+					throw Exception(a->expr(), 
+						"multiple &until on $input");
+					}
+				attr_until_input_expr_ = a->expr();
+				}
+			else
+				{
+				if ( attr_generic_until_expr_ )
+					{
+					throw Exception(a->expr(), 
+						"multiple &until condition");
+					}
+				attr_generic_until_expr_ = a->expr();
+				}
+			}
+			break;
+
+		default:
+			break;
+		}
+	}
+
+void ArrayType::Prepare(Env *env, int flags)
+	{
+	if ( flags & TO_BE_PARSED )
+		{
+		ID *arraylength_var = new ID(fmt("%s__arraylength", value_var()->Name()));
+		ID *elem_var = new ID(fmt("%s__elem", value_var()->Name()));
+		ID *elem_it_var = new ID(fmt("%s__it", elem_var->Name()));
+
+		elem_var_field_ = 
+			new ParseVarField(Field::CLASS_MEMBER, elem_var, elemtype_);
+		AddField(elem_var_field_);
+
+		if ( incremental_parsing() )
+			{
+			arraylength_var_field_ = 
+				new PrivVarField(arraylength_var, extern_type_int->Clone());
+			elem_it_var_field_ = 
+				new PrivVarField(elem_it_var, extern_type_int->Clone());
+
+			AddField(arraylength_var_field_);
+			AddField(elem_it_var_field_);
+			}
+		else
+			{
+			arraylength_var_field_ = 
+				new TempVarField(arraylength_var, extern_type_int->Clone());
+			elem_it_var_field_ = 
+				new TempVarField(elem_it_var, extern_type_int->Clone());
+
+			arraylength_var_field_->Prepare(env);
+			elem_it_var_field_->Prepare(env);
+
+			// Add elem_dataptr_var only when not parsing incrementally
+			ID *elem_dataptr_var = 
+				new ID(fmt("%s__dataptr", elem_var->Name()));
+			elem_dataptr_var_field_ = new TempVarField(
+				elem_dataptr_var, 
+				extern_type_const_byteptr->Clone());
+			elem_dataptr_var_field_->Prepare(env);
+
+			// until(dataptr >= end_of_data)
+			elem_dataptr_until_expr_ = new Expr(
+				Expr::EXPR_GE, 
+				new Expr(elem_dataptr_var->clone()),
+				new Expr(end_of_data->clone()));
+			}
+
+		if ( attr_until_input_expr_ )
+			{
+			elemtype_->SetUntilCheck(this);
+			}
+
+		end_of_array_loop_label_ = strfmt("end_of_%s", value_var()->Name());
+		}
+
+	Type::Prepare(env, flags);
+	}
+
+void ArrayType::GenArrayLength(Output *out_cc, Env *env, const DataPtr& data)
+	{
+	if ( env->Evaluated(arraylength_var()) )
+		return;
+
+	if ( ! incremental_parsing() )
+		{
+		arraylength_var_field_->GenTempDecls(out_cc, env);
+		arraylength_var_field_->GenInitCode(out_cc, env);
+		}
+
+	if ( length_ )
+		{
+		out_cc->println("%s = %s;", 
+			env->LValue(arraylength_var()), 
+			length_->EvalExpr(out_cc, env));
+
+		env->SetEvaluated(arraylength_var());
+
+		// Check for overlong array length. We cap it at the
+		// maximum data size as we won't store more elements.
+		out_cc->println("if ( t_begin_of_data + %s > t_end_of_data + 1 )",
+			env->LValue(arraylength_var()));
+		out_cc->inc_indent();
+		out_cc->println("{");
+		out_cc->println("%s = t_end_of_data - t_begin_of_data + 1;",
+			env->LValue(arraylength_var()));
+		out_cc->println("}");
+		out_cc->dec_indent();
+		
+		// Check negative array length
+		out_cc->println("if ( %s < 0 )",
+			env->LValue(arraylength_var())); 
+		out_cc->inc_indent();
+		out_cc->println("{");
+		out_cc->println("%s = 0;",
+			env->LValue(arraylength_var())); 
+		out_cc->println("}");
+		out_cc->dec_indent();
+		}
+	else if ( attr_restofdata_ )
+		{
+		ASSERT(elemtype_->StaticSize(env) == 1);
+		out_cc->println("%s = (%s) - (%s);", 
+			env->LValue(arraylength_var()), 
+			env->RValue(end_of_data),
+			data.ptr_expr());
+		env->SetEvaluated(arraylength_var());
+		}
+	}
+
+void ArrayType::GenPubDecls(Output *out_h, Env *env)
+	{
+	Type::GenPubDecls(out_h, env);
+
+	if ( declared_as_type() )
+		{
+		out_h->println("int size() const	{ return %s ? %s->size() : 0; }",
+			env->RValue(value_var()),
+			env->RValue(value_var()));
+		out_h->println("%s operator[](int index) const { BINPAC_ASSERT(%s); return (*%s)[index]; }",
+			elemtype_->DataTypeConstRefStr().c_str(),
+			env->RValue(value_var()),
+			env->RValue(value_var()));
+		}
+	}
+
+void ArrayType::GenPrivDecls(Output *out_h, Env *env)
+	{
+	ASSERT(elem_var_field_->type() == elemtype_);
+	ASSERT(elemtype_->value_var());
+	Type::GenPrivDecls(out_h, env);
+	}
+
+void ArrayType::GenInitCode(Output *out_cc, Env *env)
+	{
+	// Do not initiate the array here
+	// out_cc->println("%s = new %s;", lvalue(), vector_str_.c_str());
+	out_cc->println("%s = 0;", lvalue());
+
+	Type::GenInitCode(out_cc, env);
+	if ( incremental_parsing() )
+		{
+		out_cc->println("%s = -1;", 
+			env->LValue(elem_it_var()));
+		}
+	}
+
+void ArrayType::GenCleanUpCode(Output *out_cc, Env *env)
+	{
+	Type::GenCleanUpCode(out_cc, env);
+	if ( elemtype_->NeedsCleanUp() )
+		{
+		if ( ! elem_var_field_ )
+			{
+			ID *elem_var = new ID(fmt("%s__elem", value_var()->Name()));
+			elem_var_field_ = 
+				new ParseVarField(
+					Field::NOT_CLASS_MEMBER, 
+					elem_var, 
+					elemtype_);
+			elem_var_field_->Prepare(env);
+			}
+
+		out_cc->println("if ( %s )", env->RValue(value_var()));
+		out_cc->inc_indent();
+		out_cc->println("{");
+
+		out_cc->println("for ( int i = 0; i < (int) %s->size(); ++i )",
+			env->RValue(value_var()));
+		out_cc->inc_indent();
+		out_cc->println("{");
+		out_cc->println("%s %s = (*%s)[i];", 
+			elemtype_->DataTypeStr().c_str(), 
+			env->LValue(elem_var()), 
+			lvalue());
+		elemtype_->GenCleanUpCode(out_cc, env);
+		out_cc->println("}");
+		out_cc->dec_indent();
+
+		out_cc->println("}");
+		out_cc->dec_indent();
+		}
+	out_cc->println("delete %s;", lvalue());
+	}
+
+string ArrayType::GenArrayInit(Output *out_cc, Env *env, bool known_array_length)
+	{
+	string array_str;
+
+	array_str = lvalue();
+	if ( incremental_parsing() )
+		{
+		out_cc->println("if ( %s < 0 )", 
+			env->LValue(elem_it_var()));
+		out_cc->inc_indent();
+		out_cc->println("{");
+		out_cc->println("// Initialize only once");
+		out_cc->println("%s = 0;", env->LValue(elem_it_var()));
+		}
+
+	out_cc->println("%s = new %s;", 
+		lvalue(), vector_str_.c_str());
+
+	if ( known_array_length )
+		{
+		out_cc->println("%s->reserve(%s);", 
+			lvalue(), env->RValue(arraylength_var()));
+		}
+
+	if ( incremental_parsing() )
+		{
+		out_cc->println("}");
+		out_cc->dec_indent();
+		}
+
+	return array_str;
+	}
+
+void ArrayType::GenElementAssignment(Output *out_cc, Env *env,
+		string const &array_str, bool use_vector)
+	{
+	// Assign the element
+	if ( ! use_vector )
+		{
+		out_cc->println("%s[%s] = %s;", 
+			array_str.c_str(), 
+			env->LValue(elem_it_var()), 
+			env->LValue(elem_var()));
+		}
+	else
+		{
+		out_cc->println("%s->push_back(%s);", 
+			array_str.c_str(), 
+			env->LValue(elem_var()));
+		}
+	}
+
+void ArrayType::DoGenParseCode(Output *out_cc, Env *env, 
+		const DataPtr& data, int flags)
+	{
+	GenArrayLength(out_cc, env, data);
+
+	// Otherwise these variables are declared as member variables
+	if ( ! incremental_parsing() )
+		{
+		// Declare and initialize temporary variables
+		elem_var_field_->GenInitCode(out_cc, env);
+		elem_it_var_field_->GenTempDecls(out_cc, env);
+		out_cc->println("%s = 0;", env->LValue(elem_it_var()));
+		env->SetEvaluated(elem_it_var());
+		}
+
+	/*
+	If the input length can be determined without parsing
+	individual elements, generate the boundary checking before
+	parsing (unless in the case of incremental parsing).
+
+	There are two cases when the input length can be determined:
+	1. The array has a static size;
+	2. The array length can be computed before parsing and
+   	each element is of constant size.
+	*/
+
+	bool compute_size_var = false;
+
+	if ( incremental_input() )
+		{
+		// Do not compute size_var on incremental input
+		compute_size_var = false;
+		
+		if ( ! incremental_parsing() &&
+		     ( StaticSize(env) >= 0 ||
+		       ( env->Evaluated(arraylength_var()) && 
+		         elemtype_->StaticSize(env) >= 0 ) ) )
+			{
+			GenBoundaryCheck(out_cc, env, data);
+			}
+		}
+	else
+		{
+		compute_size_var = AddSizeVar(out_cc, env);
+		}
+
+	bool known_array_length = env->Evaluated(arraylength_var());
+	string array_str = GenArrayInit(out_cc, env, known_array_length);
+
+	bool use_vector = true;
+
+	ASSERT(elem_it_var());
+
+	DataPtr elem_data(env, 0, 0);
+
+	if ( elem_dataptr_var() )
+		{
+		out_cc->println("const_byteptr %s = %s;", 
+			env->LValue(elem_dataptr_var()), data.ptr_expr());
+		env->SetEvaluated(elem_dataptr_var());
+
+		elem_data = DataPtr(env, elem_dataptr_var(), 0);
+		}
+
+	string for_condition = known_array_length ?
+		strfmt("%s < %s", 
+			env->LValue(elem_it_var()), 
+			env->RValue(arraylength_var())) :
+		"/* forever */";
+
+	out_cc->println("for (; %s; ++%s)", 
+		for_condition.c_str(), 
+		env->LValue(elem_it_var()));
+	out_cc->inc_indent();
+	out_cc->println("{");
+
+	if ( attr_generic_until_expr_ )
+		GenUntilCheck(out_cc, env, attr_generic_until_expr_, true);
+
+	if ( elem_dataptr_var() )
+		GenUntilCheck(out_cc, env, elem_dataptr_until_expr_, false);
+	
+	elemtype_->GenPreParsing(out_cc, env);
+	elemtype_->GenParseCode(out_cc, env, elem_data, flags);
+
+	if ( incremental_parsing() ) 
+		{
+		out_cc->println("if ( ! %s )", 
+			elemtype_->parsing_complete(env).c_str());
+		out_cc->inc_indent();
+		out_cc->println("goto %s;", kNeedMoreData);
+		out_cc->dec_indent();
+		}
+
+	GenElementAssignment(out_cc, env, array_str, use_vector);
+
+	if ( elem_dataptr_var() )
+		{
+		out_cc->println("%s += %s;", 
+			env->LValue(elem_dataptr_var()), 
+			elemtype_->DataSize(0, env, elem_data).c_str());
+		out_cc->println("BINPAC_ASSERT(%s <= %s);",
+			env->RValue(elem_dataptr_var()), 
+			env->RValue(end_of_data));
+		}
+
+	if ( attr_until_element_expr_ )
+		GenUntilCheck(out_cc, env, attr_until_element_expr_, false);
+
+	if ( elemtype_->IsPointerType() )
+		out_cc->println("%s = 0;", env->LValue(elem_var()));
+
+	out_cc->println("}");
+	out_cc->dec_indent();
+
+	out_cc->dec_indent();
+	out_cc->println("%s: ;", end_of_array_loop_label_.c_str());
+	out_cc->inc_indent();
+
+	if ( compute_size_var && elem_dataptr_var() && ! env->Evaluated(size_var()) )
+		{
+		// Compute the data size
+		out_cc->println("%s = %s - (%s);", 
+			env->LValue(size_var()),
+			env->RValue(elem_dataptr_var()),
+			data.ptr_expr());
+		env->SetEvaluated(size_var());
+		}
+	}
+
+void ArrayType::GenUntilInputCheck(Output *out_cc, Env *env)
+	{
+	ID *elem_input_var_id = new ID(
+		fmt("%s__elem_input", value_var()->Name()));
+	elem_input_var_field_ = new TempVarField(
+		elem_input_var_id, extern_type_const_bytestring->Clone());
+	elem_input_var_field_->Prepare(env);
+
+	out_cc->println("%s %s(%s, %s);", 
+		extern_type_const_bytestring->DataTypeStr().c_str(),
+		env->LValue(elem_input_var()),
+		env->RValue(begin_of_data),
+		env->RValue(end_of_data));
+	env->SetEvaluated(elem_input_var());
+
+	GenUntilCheck(out_cc, env, attr_until_input_expr_, true);
+	}
+
+void ArrayType::GenUntilCheck(Output *out_cc, Env *env, 
+		Expr *until_expr, bool delete_elem)
+	{
+	ASSERT(until_expr);
+
+	Env check_env(env, this);
+	check_env.AddMacro(element_macro_id, 
+		new Expr(elem_var()->clone()));
+	if ( elem_input_var() )
+		{
+		check_env.AddMacro(input_macro_id, 
+			new Expr(elem_input_var()->clone()));
+		}
+
+	out_cc->println("// Check &until(%s)", until_expr->orig());
+	out_cc->println("if ( %s )", 
+		until_expr->EvalExpr(out_cc, &check_env));
+	out_cc->inc_indent();
+	out_cc->println("{");
+	if ( parsing_complete_var() )
+		{
+		out_cc->println("%s = true;",
+			env->LValue(parsing_complete_var()));
+		}
+
+	if ( elemtype_->IsPointerType() )
+		{
+		if ( delete_elem )
+			elemtype_->GenCleanUpCode(out_cc, env);
+		else
+			out_cc->println("%s = 0;", env->LValue(elem_var()));
+		}
+
+	out_cc->println("goto %s;", end_of_array_loop_label_.c_str());
+	out_cc->println("}");
+	out_cc->dec_indent();
+	}
+
+void ArrayType::GenDynamicSize(Output *out_cc, Env *env,
+		const DataPtr& data)
+	{
+	ASSERT(! incremental_input());
+	DEBUG_MSG("Generating dynamic size for array `%s'\n", 
+		value_var()->Name());
+
+	int elem_w = elemtype_->StaticSize(env);
+	if ( elem_w >= 0 &&
+	     ! attr_until_element_expr_ && 
+	     ! attr_until_input_expr_ &&
+	     ( length_ || attr_restofdata_ ) )
+		{
+		// If the elements have a fixed size,
+		// we only need to compute the number of elements
+		bool compute_size_var = AddSizeVar(out_cc, env);
+		ASSERT(compute_size_var);
+		GenArrayLength(out_cc, env, data);
+		ASSERT(env->Evaluated(arraylength_var()));
+		out_cc->println("%s = %d * %s;",
+			env->LValue(size_var()), elem_w, env->RValue(arraylength_var()));
+		env->SetEvaluated(size_var());
+		}
+	else
+		{
+		// Otherwise we need parse the array dynamically
+		GenParseCode(out_cc, env, data, 0);
+		}
+	}
+
+int ArrayType::StaticSize(Env *env) const
+	{
+	int num = 0;
+
+	if ( ! length_ || ! length_->ConstFold(env, &num) )
+		return -1;
+
+	int elem_w = elemtype_->StaticSize(env);
+	if ( elem_w < 0 )
+		return -1;
+
+	DEBUG_MSG("static size of %s:%s = %d * %d\n", 
+		decl_id()->Name(), lvalue(), elem_w, num);
+
+	return num * elem_w;
+	}
+
+void ArrayType::SetBoundaryChecked()
+	{
+	Type::SetBoundaryChecked();
+	elemtype_->SetBoundaryChecked();
+	}
+
+void ArrayType::DoMarkIncrementalInput()
+	{
+	elemtype_->MarkIncrementalInput(); 
+	}
+
+bool ArrayType::RequiresAnalyzerContext()
+	{ 
+	return Type::RequiresAnalyzerContext() ||
+	       ( length_ && length_->RequiresAnalyzerContext() ) || 
+	       elemtype_->RequiresAnalyzerContext(); 
+	}
+
+bool ArrayType::DoTraverse(DataDepVisitor *visitor)
+	{ 
+	if ( ! Type::DoTraverse(visitor) )
+		return false;
+
+	if ( length_ && ! length_->Traverse(visitor) )
+		return false;
+
+	if ( ! elemtype_->Traverse(visitor) )
+		return false; 
+
+	return true;
+	}

aux/binpac/src/pac_array.h

@@ -0,0 +1,92 @@
+#ifndef pac_array_h
+#define pac_array_h
+
+#include "pac_common.h"
+#include "pac_type.h"
+
+// Fixed-length array and variable length sequence with an ending pattern
+
+class ArrayType : public Type
+{
+public:
+	ArrayType(Type *arg_elemtype, Expr *arg_length = 0);
+	~ArrayType();
+
+	bool DefineValueVar() const;
+	string DataTypeStr() const;
+	string DefaultValue() const	{ return "0"; }
+	Type *ElementDataType() const;
+
+	string EvalElement(const string &array, const string &index) const;
+
+	void ProcessAttr(Attr *a);
+
+	void Prepare(Env *env, int flags);
+
+	void GenPubDecls(Output *out, Env *env);
+	void GenPrivDecls(Output *out, Env *env);
+
+	void GenInitCode(Output *out, Env *env);
+	void GenCleanUpCode(Output *out, Env *env);
+
+	int StaticSize(Env *env) const;
+
+	void SetBoundaryChecked();
+	void GenUntilInputCheck(Output *out_cc, Env *env);
+
+	bool IsPointerType() const	{ return true; }
+
+protected:
+	void init();
+
+	void DoGenParseCode(Output *out, Env *env, const DataPtr& data, int flags);
+	void GenDynamicSize(Output *out, Env *env, const DataPtr& data);
+	void GenArrayLength(Output *out_cc, Env *env, const DataPtr& data);
+	string GenArrayInit(Output *out_cc, Env *env, bool known_array_length);
+	void GenElementAssignment(Output *out_cc, Env *env,
+		string const &array_str, bool use_vector);
+	void GenUntilCheck(Output *out_cc, Env *env, 
+		Expr *until_condition, bool delete_elem);
+
+	bool ByteOrderSensitive() const 
+		{ 
+		return elemtype_->RequiresByteOrder(); 
+		}
+	bool RequiresAnalyzerContext();
+
+	Type *DoClone() const;
+
+	void DoMarkIncrementalInput();
+
+	const ID *arraylength_var() const;
+	const ID *elem_it_var() const;
+	const ID *elem_var() const;
+	const ID *elem_dataptr_var() const;
+	const ID *elem_input_var() const;
+
+protected:
+	bool DoTraverse(DataDepVisitor *visitor);
+
+private:
+	Type *elemtype_;
+	Expr *length_;
+
+	string vector_str_;
+	string datatype_str_;
+	string end_of_array_loop_label_;
+
+	Field *arraylength_var_field_;
+	Field *elem_it_var_field_;
+	Field *elem_var_field_;
+	Field *elem_dataptr_var_field_;
+	Field *elem_input_var_field_;
+
+	// This does not come from &until, but is internally generated
+	Expr *elem_dataptr_until_expr_;
+
+	Expr *attr_generic_until_expr_;
+	Expr *attr_until_element_expr_;
+	Expr *attr_until_input_expr_;
+};
+
+#endif  // pac_array_h

aux/binpac/src/pac_attr.cc

@@ -0,0 +1,57 @@
+#include "pac_attr.h"
+#include "pac_expr.h"
+
+bool Attr::DoTraverse(DataDepVisitor *visitor)
+	{
+	if ( expr_ && ! expr_->Traverse(visitor) )
+		return false;
+	return true;
+	}
+
+bool Attr::RequiresAnalyzerContext() const
+	{
+	return (expr_ && expr_->RequiresAnalyzerContext());
+	}
+
+void Attr::init()
+	{
+	expr_ = 0;
+	seqend_ = 0;
+	}
+
+Attr::Attr(AttrType type)
+	: DataDepElement(DataDepElement::ATTR)
+	{
+	type_ = type;
+	init();
+	}
+
+Attr::Attr(AttrType type, Expr *expr)
+	: DataDepElement(DataDepElement::ATTR)
+	{
+	type_ = type;
+	init();
+	expr_ = expr;
+	}
+
+Attr::Attr(AttrType type, ExprList *exprlist)
+	: DataDepElement(DataDepElement::ATTR)
+	{
+	type_ = type;
+	init();
+	expr_ = new Expr(exprlist);
+	}
+
+Attr::Attr(AttrType type, SeqEnd *seqend)
+	: DataDepElement(DataDepElement::ATTR)
+	{
+	type_ = type;
+	init();
+	seqend_ = seqend;
+	}
+
+LetAttr::LetAttr(FieldList *letfields)
+	: Attr(ATTR_LET)
+	{
+	letfields_ = letfields;
+	}

aux/binpac/src/pac_attr.h

@@ -0,0 +1,61 @@
+#ifndef pac_attr_h
+#define pac_attr_h
+
+#include "pac_common.h"
+#include "pac_datadep.h"
+
+enum AttrType { 
+	ATTR_BYTEORDER, 
+	ATTR_CHECK, 
+	ATTR_CHUNKED,
+	ATTR_EXPORTSOURCEDATA,
+	ATTR_IF,
+	ATTR_LENGTH, 
+	ATTR_LET,
+	ATTR_LINEBREAKER,
+	ATTR_MULTILINE,
+	ATTR_ONELINE,
+	ATTR_REFCOUNT,
+	ATTR_REQUIRES,
+	ATTR_RESTOFDATA, 
+	ATTR_RESTOFFLOW, 
+	ATTR_TRANSIENT,
+	ATTR_UNTIL,
+};
+
+class Attr : public Object, public DataDepElement
+{
+public:
+	Attr(AttrType type);
+	Attr(AttrType type, Expr *expr);
+	Attr(AttrType type, ExprList *exprlist);
+	Attr(AttrType type, SeqEnd *seqend);
+
+	AttrType type() const 		{ return type_; }
+	Expr *expr() const		{ return expr_; }
+	SeqEnd *seqend() const		{ return seqend_; }
+
+	bool RequiresAnalyzerContext() const;
+
+protected:
+	bool DoTraverse(DataDepVisitor *visitor);
+
+protected:
+	void init();
+
+	AttrType type_;
+	Expr *expr_;
+	SeqEnd *seqend_;
+};
+
+class LetAttr : public Attr
+{
+public:
+	LetAttr(FieldList *letfields);
+	FieldList *letfields() const	{ return letfields_; }
+
+private:
+	FieldList *letfields_;
+};
+
+#endif  // pac_attr_h

aux/binpac/src/pac_btype.cc

@@ -0,0 +1,144 @@
+#include "pac_btype.h"
+#include "pac_dataptr.h"
+#include "pac_id.h"
+#include "pac_output.h"
+
+Type *BuiltInType::DoClone() const
+	{
+	return new BuiltInType(bit_type());
+	}
+
+bool BuiltInType::IsNumericType() const
+	{
+	BITType t = bit_type();
+	return (t == INT8 || t == INT16 || t == INT32 ||
+	        t == UINT8 || t == UINT16 || t == UINT32);
+	}
+
+bool BuiltInType::CompatibleBuiltInTypes(BuiltInType *type1, 
+	                                 BuiltInType *type2)
+	{
+	return type1->IsNumericType() && type2->IsNumericType();
+	}
+
+static const char* basic_pactype_name[] = {
+#	define TYPE_DEF(name, pactype, ctype, size)	pactype,
+#	include "pac_type.def"
+#	undef TYPE_DEF
+	0,
+};
+
+void BuiltInType::static_init()
+	{
+	for ( int bit_type = 0; basic_pactype_name[bit_type]; ++bit_type )
+		{
+		Type::AddPredefinedType(
+			basic_pactype_name[bit_type], 
+			new BuiltInType((BITType) bit_type));
+		}
+	}
+
+int BuiltInType::LookUpByName(const char* name)
+	{
+	ASSERT(0);
+	for ( int i = 0; basic_pactype_name[i]; ++i )
+		if ( strcmp(basic_pactype_name[i], name) == 0 )
+			return i;
+	return -1;
+	}
+
+static const char* basic_ctype_name[] = {
+#	define TYPE_DEF(name, pactype, ctype, size)	ctype,
+#	include "pac_type.def"
+#	undef TYPE_DEF
+	0,
+};
+
+bool BuiltInType::DefineValueVar() const
+	{
+	return bit_type_ != EMPTY;
+	}
+
+string BuiltInType::DataTypeStr() const
+	{
+	return basic_ctype_name[bit_type_];
+	}
+
+int BuiltInType::StaticSize(Env* /* env */) const
+	{
+	static const size_t basic_type_size[] = 
+		{
+#		define TYPE_DEF(name, pactype, ctype, size)	size,
+#		include "pac_type.def"
+#		undef TYPE_DEF
+		};
+
+	return basic_type_size[bit_type_];
+	}
+
+void BuiltInType::DoMarkIncrementalInput()
+	{
+	if ( bit_type_ == EMPTY )
+		return;
+	Type::DoMarkIncrementalInput();
+	}
+
+void BuiltInType::GenInitCode(Output* out_cc, Env* env)
+	{
+	if ( bit_type_ != EMPTY )
+		out_cc->println("%s = 0;", env->LValue(value_var()));
+	Type::GenInitCode(out_cc, env);
+	}
+
+void BuiltInType::GenDynamicSize(Output* out_cc, Env* env, const DataPtr& data)
+	{
+	/* should never be called */
+	ASSERT(0);
+	}
+
+void BuiltInType::DoGenParseCode(Output* out_cc, Env* env,
+		const DataPtr& data, int flags)
+	{
+	if ( bit_type_ == EMPTY )
+		return;
+
+	// There is no need to generate the size variable
+	// out_cc->println("%s = sizeof(%s);", size_var(), DataTypeStr().c_str());
+
+	GenBoundaryCheck(out_cc, env, data);
+
+	if ( anonymous_value_var() )
+		return;
+
+	switch ( bit_type_ ) 
+		{
+		case EMPTY:
+			// do nothing
+			break;
+
+		case INT8:
+		case UINT8:
+			out_cc->println("%s = *((%s const *) (%s));",
+				lvalue(), DataTypeStr().c_str(), data.ptr_expr()); 
+			break;
+		case INT16:
+		case UINT16:
+		case INT32:
+		case UINT32:
+#if 0
+			out_cc->println("%s = UnMarshall<%s>(%s, %s);",
+				lvalue(), 
+				DataTypeStr().c_str(), 
+				data.ptr_expr(),
+				EvalByteOrder(out_cc, env).c_str());
+#else
+			out_cc->println("%s = FixByteOrder(%s, *((%s const *) (%s)));",
+				lvalue(), 
+				EvalByteOrder(out_cc, env).c_str(),
+				DataTypeStr().c_str(), 
+				data.ptr_expr());
+#endif
+			break;
+		}
+	}
+

aux/binpac/src/pac_btype.h

@@ -0,0 +1,52 @@
+#ifndef pac_btype_h
+#define pac_btype_h
+
+#include "pac_type.h"
+
+class BuiltInType : public Type
+{
+public:
+	enum BITType {
+#		define TYPE_DEF(name, pactype, ctype, size)	name,
+#		include "pac_type.def"
+#		undef TYPE_DEF
+	};
+
+	static int LookUpByName(const char *name);
+
+	BuiltInType(BITType bit_type)
+		: Type(bit_type == BuiltInType::EMPTY ? Type::EMPTY : BUILTIN),
+		  bit_type_(bit_type) {}
+
+	BITType bit_type() const	{ return bit_type_; }
+
+	bool IsNumericType() const;
+
+	bool DefineValueVar() const;
+	string DataTypeStr() const;
+	string DefaultValue() const	{ return "0"; }
+
+	int StaticSize(Env *env) const;
+
+	bool IsPointerType() const	{ return false; }
+
+	bool ByteOrderSensitive() const { return StaticSize(0) >= 2; }
+
+	void GenInitCode(Output *out_cc, Env *env);
+
+	void DoMarkIncrementalInput();
+
+protected:
+	void DoGenParseCode(Output *out, Env *env, const DataPtr& data, int flags);
+	void GenDynamicSize(Output *out, Env *env, const DataPtr& data);
+	Type *DoClone() const;
+
+	BITType bit_type_;
+
+public:
+	static void static_init();
+	static bool CompatibleBuiltInTypes(BuiltInType *type1, 
+	                                   BuiltInType *type2);
+};
+
+#endif  // pac_btype_h

aux/binpac/src/pac_case.cc

@@ -0,0 +1,391 @@
+#include "pac_exception.h"
+#include "pac_expr.h"
+#include "pac_exttype.h"
+#include "pac_id.h"
+#include "pac_output.h"
+#include "pac_typedecl.h"
+#include "pac_utils.h"
+
+#include "pac_case.h"
+
+CaseType::CaseType(Expr* index_expr, CaseFieldList* cases)
+	: Type(CASE), index_expr_(index_expr), cases_(cases) 
+	{
+	index_var_ = 0;
+	foreach(i, CaseFieldList, cases_)
+		AddField(*i);
+	}
+
+CaseType::~CaseType()
+	{
+	delete index_var_;
+	delete index_expr_;
+	delete cases_;
+	}
+
+void CaseType::AddCaseField(CaseField *f)
+	{
+	// All fields must be added before Prepare()
+	ASSERT(!env());
+
+	AddField(f);
+	cases_->push_back(f);
+	}
+
+bool CaseType::DefineValueVar() const
+	{
+	return false;
+	}
+
+string CaseType::DataTypeStr() const
+	{
+	ASSERT(type_decl());
+	return strfmt("%s *", type_decl()->class_name().c_str());
+	}
+
+Type *CaseType::ValueType() const
+	{
+	foreach (i, CaseFieldList, cases_)
+		{
+		CaseField *c = *i;
+		return c->type();
+		}
+	ASSERT(0);
+	return 0;
+	}
+
+string CaseType::DefaultValue() const
+	{
+	return ValueType()->DefaultValue();
+	}
+
+void CaseType::Prepare(Env* env, int flags)
+	{
+	ASSERT(flags & TO_BE_PARSED);
+
+	index_var_ = new ID(fmt("%s_case_index", value_var()->Name()));
+	env->AddID(index_var_, MEMBER_VAR, extern_type_int);
+
+	// Sort the cases_ to put the default case at the end of the list
+	CaseFieldList::iterator default_case_it = 
+		cases_->end(); // to avoid warning
+	CaseField *default_case = 0;
+
+	foreach (i, CaseFieldList, cases_)
+		{
+		CaseField *c = *i;
+		if ( ! c->index() )
+			{
+			if ( default_case )
+				throw Exception(c, "duplicate default case");
+			default_case_it = i;
+			default_case = c;
+			}
+		}
+	if ( default_case )
+		{
+		cases_->erase(default_case_it);
+		cases_->push_back(default_case);
+		}
+
+	foreach (i, CaseFieldList, cases_)
+		{
+		CaseField *c = *i;
+		c->set_index_var(index_var_);
+		c->set_case_type(this);
+		}
+
+	Type::Prepare(env, flags);
+	}
+
+void CaseType::GenPrivDecls(Output* out_h, Env* env)
+	{
+	out_h->println("int %s;", env->LValue(index_var_));
+	Type::GenPrivDecls(out_h, env);
+	}
+
+void CaseType::GenPubDecls(Output* out_h, Env* env)
+	{
+	out_h->println("int %s const	{ return %s; }",
+		env->RValue(index_var_), env->LValue(index_var_));
+	Type::GenPubDecls(out_h, env);
+	}
+
+void CaseType::GenInitCode(Output* out_cc, Env* env)
+	{
+	out_cc->println("%s = -1;", env->LValue(index_var_));
+	Type::GenInitCode(out_cc, env);
+	}
+
+void CaseType::GenCleanUpCode(Output* out_cc, Env* env)
+	{
+	Type::GenCleanUpCode(out_cc, env);
+
+	env->set_in_branch(true);
+	out_cc->println("switch ( %s )", env->RValue(index_var_));
+	out_cc->inc_indent();
+	out_cc->println("{");
+	foreach (i, CaseFieldList, cases_)
+		{
+		CaseField *c = *i;
+		c->GenCleanUpCode(out_cc, env);
+		}
+	out_cc->println("}");
+	out_cc->dec_indent();
+	env->set_in_branch(false);
+	}
+
+void CaseType::DoGenParseCode(Output* out_cc, Env* env,
+		const DataPtr& data, int flags)
+	{
+	if ( StaticSize(env) >= 0 )
+		GenBoundaryCheck(out_cc, env, data);
+
+	bool compute_size_var = false;
+
+	if ( ! incremental_input() )
+		compute_size_var = AddSizeVar(out_cc, env);
+
+	out_cc->println("%s = %s;", 
+		env->LValue(index_var_), index_expr_->EvalExpr(out_cc, env));
+	env->SetEvaluated(index_var_);
+	
+	env->set_in_branch(true);
+	out_cc->println("switch ( %s )", env->RValue(index_var_));
+	out_cc->inc_indent();
+	out_cc->println("{");
+	bool has_default_case = false;
+	foreach (i, CaseFieldList, cases_)
+		{
+		CaseField *c = *i;
+		c->GenParseCode(out_cc, env, data, 
+			compute_size_var ? size_var() : 0);
+		if ( c->IsDefaultCase() )
+			has_default_case = true;
+		}
+
+	if ( ! has_default_case )
+		{
+		out_cc->println("default:");
+		out_cc->inc_indent();
+		out_cc->println("throw ExceptionInvalidCaseIndex(\"%s\", %s);", 
+			decl_id()->Name(), env->RValue(index_var_));
+		out_cc->println("break;");
+		out_cc->dec_indent();
+		}
+	out_cc->println("}");
+	out_cc->dec_indent();
+	env->set_in_branch(false);
+
+	if ( compute_size_var )
+		env->SetEvaluated(size_var());
+	}
+
+void CaseType::GenDynamicSize(Output* out_cc, Env* env,
+		const DataPtr& data)
+	{
+	GenParseCode(out_cc, env, data, 0);
+	}
+
+int CaseType::StaticSize(Env* env) const
+	{
+	int static_w = -1;
+	foreach (i, CaseFieldList, cases_)
+		{
+		CaseField *c = *i;
+		int w = c->StaticSize(env);
+		if ( w < 0 || ( static_w >= 0 && w != static_w ) )
+			return -1;
+		static_w = w;
+		}
+	return static_w;
+	}
+
+void CaseType::SetBoundaryChecked()
+	{
+	Type::SetBoundaryChecked();
+	foreach (i, CaseFieldList, cases_)
+		{
+		CaseField *c = *i;
+		c->SetBoundaryChecked();
+		}
+	}
+
+void CaseType::DoMarkIncrementalInput()
+	{
+	foreach (i, CaseFieldList, cases_)
+		{
+		CaseField *c = *i;
+		c->type()->MarkIncrementalInput();
+		}
+	}
+
+bool CaseType::ByteOrderSensitive() const
+	{
+	foreach (i, CaseFieldList, cases_)
+		{
+		CaseField *c = *i;
+		if ( c->RequiresByteOrder() )
+			return true;
+		}
+	return false;
+	}
+
+CaseField::CaseField(ExprList* index, ID* id, Type* type)
+	: Field(CASE_FIELD, 
+		TYPE_TO_BE_PARSED | CLASS_MEMBER | PUBLIC_READABLE, 
+		id, type), 
+	  index_(index)
+	{
+	ASSERT(type_);
+	type_->set_value_var(id, MEMBER_VAR);
+	case_type_ = 0;
+	}
+
+CaseField::~CaseField()
+	{
+	delete_list(ExprList, index_);
+	}
+
+void GenCaseStr(ExprList *index_list, Output *out_cc, Env *env)
+	{
+	if ( index_list )
+		{
+		foreach(i, ExprList, index_list)
+			{
+			Expr *index_expr = *i;
+			int index_const;
+			
+			if ( ! index_expr->ConstFold(env, &index_const) )
+				throw ExceptionNonConstExpr(index_expr);
+			out_cc->println("case %d:", index_const);
+			}
+		}
+	else
+		{
+		out_cc->println("default:");
+		}
+	}
+
+void CaseField::Prepare(Env* env)
+	{
+	ASSERT(index_var_);
+	Field::Prepare(env);
+	}
+
+void CaseField::GenPubDecls(Output* out_h, Env* env)
+	{
+	if ( ! ((flags_ & PUBLIC_READABLE) && (flags_ & CLASS_MEMBER)) )
+		return;
+
+	// Skip type "empty"
+	if ( type_->DataTypeStr().empty() )
+		return;
+
+	out_h->println("%s %s const",
+		type_->DataTypeConstRefStr().c_str(), env->RValue(id_));
+
+	out_h->inc_indent();
+	out_h->println("{");
+
+	if ( ! index_ )
+		out_h->println("return %s;", lvalue());
+	else
+		{
+		out_h->println("switch ( %s )", env->RValue(index_var_));
+		out_h->inc_indent();
+		out_h->println("{");
+		GenCaseStr(index_, out_h, env);
+		out_h->inc_indent();
+		out_h->println("break;  // OK");
+		out_h->dec_indent();
+
+		out_h->println("default:");
+		out_h->inc_indent();
+		out_h->println("throw ExceptionInvalidCase(\"%s\", %s, \"%s\");", 
+		               id_->LocName(), 
+		               env->RValue(index_var_), 
+		               OrigExprList(index_).c_str());
+		out_h->println("break;");
+		out_h->dec_indent();
+
+		out_h->println("}");
+		out_h->dec_indent();
+
+		out_h->println("return %s;", lvalue());
+		}
+
+	out_h->println("}");
+	out_h->dec_indent();
+	}
+
+void CaseField::GenInitCode(Output* out_cc, Env* env)
+	{
+	// GenCaseStr(index_, out_cc, env);
+	// out_cc->inc_indent();
+	// out_cc->println("{");
+	// out_cc->println("// Initialize \"%s\"", id_->Name());
+	type_->GenInitCode(out_cc, env);
+	// out_cc->println("}");
+	// out_cc->println("break;");
+	// out_cc->dec_indent();
+	}
+
+void CaseField::GenCleanUpCode(Output* out_cc, Env* env)
+	{
+	GenCaseStr(index_, out_cc, env);
+	out_cc->inc_indent();
+	out_cc->println("// Clean up \"%s\"", id_->Name());
+	out_cc->println("{");
+	if ( ! anonymous_field() )
+		type_->GenCleanUpCode(out_cc, env);
+	out_cc->println("}");
+	out_cc->println("break;");
+	out_cc->dec_indent();
+	}
+
+void CaseField::GenParseCode(Output* out_cc, Env* env, 
+		const DataPtr& data, const ID* size_var)
+	{
+	GenCaseStr(index_, out_cc, env);
+	out_cc->inc_indent();
+	out_cc->println("// Parse \"%s\"", id_->Name());
+	out_cc->println("{");
+	
+	{
+	Env case_env(env, this);
+	Env *env = &case_env;
+
+	type_->GenPreParsing(out_cc, env);
+	type_->GenParseCode(out_cc, env, data, 0);
+	if ( size_var )
+		{
+		out_cc->println("%s = %s;", 
+			env->LValue(size_var),
+			type_->DataSize(out_cc, env, data).c_str());
+		}
+	if ( type_->incremental_input() )
+		{
+		ASSERT(case_type()->parsing_complete_var());
+		out_cc->println("%s = %s;", 
+			env->LValue(case_type()->parsing_complete_var()),
+			env->RValue(type_->parsing_complete_var()));
+		}
+	out_cc->println("}");
+	}
+
+	out_cc->println("break;");
+	out_cc->dec_indent();
+	}
+
+bool CaseField::DoTraverse(DataDepVisitor *visitor)
+	{ 
+	return Field::DoTraverse(visitor) &&
+	       type()->Traverse(visitor); 
+	}
+
+bool CaseField::RequiresAnalyzerContext() const 
+	{ 
+	return Field::RequiresAnalyzerContext() ||
+	       type()->RequiresAnalyzerContext(); 
+	}

aux/binpac/src/pac_case.h

@@ -0,0 +1,99 @@
+#ifndef pac_case_h
+#define pac_case_h
+
+#include "pac_common.h"
+#include "pac_field.h"
+#include "pac_id.h"
+#include "pac_type.h"
+
+class CaseType : public Type
+{
+public:
+	CaseType(Expr *index, CaseFieldList *cases);
+	~CaseType();
+
+	void AddCaseField(CaseField *f);
+
+	bool DefineValueVar() const;
+	string DataTypeStr() const;
+	string DefaultValue() const;
+
+	void Prepare(Env *env, int flags);
+
+	void GenPubDecls(Output *out, Env *env);
+	void GenPrivDecls(Output *out, Env *env);
+
+	void GenInitCode(Output *out, Env *env);
+	void GenCleanUpCode(Output *out, Env *env);
+
+	int StaticSize(Env *env) const;
+
+	void SetBoundaryChecked();
+
+	Type *ValueType() const;
+
+	bool IsPointerType() const	{ return ValueType()->IsPointerType(); }
+
+protected:
+	void DoGenParseCode(Output *out, Env *env, const DataPtr& data, int flags);
+	void GenDynamicSize(Output *out, Env *env, const DataPtr& data);
+	Type *DoClone() const 	{ return 0; }
+	void DoMarkIncrementalInput();
+
+	bool ByteOrderSensitive() const;
+
+	Expr *index_expr_;
+	ID *index_var_;
+	CaseFieldList *cases_;
+
+	typedef map<const ID*, CaseField*, ID_ptr_cmp> member_map_t;
+	member_map_t member_map_;
+};
+
+class CaseField : public Field
+{
+public:
+	CaseField(ExprList *index, ID *id, Type *type);
+	~CaseField();
+
+	CaseType *case_type() const	{ return case_type_; }
+	void set_case_type(CaseType *t)	{ case_type_ = t; }
+
+	ExprList *index() const		{ return index_; }
+
+	const char *lvalue() const	{ return type_->lvalue(); }
+
+	const char *CaseStr(Env *env);
+	void set_index_var(const ID *var) { index_var_ = var; }
+	
+	void Prepare(Env *env);
+
+	void GenPubDecls(Output *out, Env *env);
+
+	void GenInitCode(Output *out, Env *env);
+	void GenCleanUpCode(Output *out, Env *env);
+	void GenParseCode(Output *out, Env *env, 
+		const DataPtr& data, const ID *size_var);
+
+	int StaticSize(Env *env) const 	{ return type_->StaticSize(env); }
+
+	bool IsDefaultCase() const	{ return ! index_; }
+	void SetBoundaryChecked()	{ type_->SetBoundaryChecked(); }
+
+	bool RequiresByteOrder() const	{ return type_->RequiresByteOrder(); }
+	bool RequiresAnalyzerContext() const;
+
+protected:
+	bool DoTraverse(DataDepVisitor *visitor);
+
+protected:
+	CaseType *case_type_;
+	ExprList *index_;
+	const ID *index_var_;
+};
+
+// Generate a list of "case X:" lines from index_list. Each index
+// expression must be constant foldable.
+void GenCaseStr(ExprList *index_list, Output *out_cc, Env *env);
+
+#endif  // pac_case_h

aux/binpac/src/pac_cclass.h

@@ -0,0 +1,81 @@
+#ifndef pac_cclass_h
+#define pac_cclass_h
+
+class CClass;
+class CClassMember;
+class CClassMethod;
+class CType;
+class CVariable;
+
+typedef vector<CClassMember *> CClassMemberList;
+typedef vector<CClassMethod *> CClassMethodList;
+typedef vector<CVariable *> CVariableList;
+
+#include "pac_common.h"
+
+// Represents a C++ class.
+// 
+// For now we adopt a simple model:
+// 
+// 1. All members have a protected member variable "name_" and a
+// public constant access method "name()".
+// 
+// 2. All methods are public.
+//
+// 3. We do not check repeated names.
+
+class CClass
+{
+public:
+	CClass(const string &class_name);
+
+	void AddMember(CClassMember *member);
+	void AddMethod(CClassMember *method);
+
+	void GenForwardDeclaration(Output *out_h);
+	void GenCode(Output *out_h, Output *out_cc);
+
+protected:
+	string class_name_;
+	CClassMemberList *members_;
+	CClassMethodList *methods_;
+};
+
+class CVariable
+{
+public:
+	CClassMember(const string &name, CType *type);
+
+	string name() const { return name_; }
+	CType *type() const { return type_; }
+
+protected:
+	string name_;
+	CType *type_;
+};
+
+class CClassMember
+{
+public:
+	CClassMember(CVariable *var);
+	void GenCode(Output *out_h, Output *out_cc);
+
+	string decl() const;
+
+protected:
+	CVariable *var_;
+};
+
+class CClassMethod
+{
+public:
+	CClassMethod(CVariable *var, CVariableList *params);
+
+	string decl() const;
+
+protected:
+	CVariable *var_;
+	CVariableList *params_;
+};
+
+#endif  // pac_cclass_h

aux/binpac/src/pac_common.h

@@ -0,0 +1,134 @@
+#ifndef pac_common_h
+#define pac_common_h
+
+#include "pac_utils.h"
+
+#include <vector>
+
+#include <ctype.h>
+#include <string.h>
+#include <stdlib.h>
+
+using namespace std;
+
+extern bool FLAGS_pac_debug;
+extern vector<string> FLAGS_include_directories;
+extern string input_filename;
+extern int line_number;
+
+// Definition of class Object, which is the base class for all objects
+// representing language elements -- identifiers, types, expressions,
+// etc.
+
+class Object
+{
+public:
+	Object()
+		{
+		filename = input_filename;
+		line_num = line_number;
+		location = strfmt("%s:%d", filename.c_str(), line_number);
+		}
+
+	~Object()
+		{
+		}
+
+	const char* Location() const	{ return location.c_str(); }
+
+protected:
+	string filename;
+	int line_num;
+	string location;
+};
+
+class ActionParam;
+class ActionParamType;
+class AnalyzerAction;
+class AnalyzerContextDecl;
+class AnalyzerDecl;
+class AnalyzerElement;
+class ArrayType;
+class Attr;
+class CClass;
+class CType;
+class ConstString;
+class CaseExpr;
+class CaseField;
+class ContextField;
+class DataPtr;
+class Decl;
+class EmbeddedCode;
+class Enum;
+class Env;
+class ExternType;
+class Expr;
+class Field;
+class Function;
+class InputBuffer;
+class LetDef;
+class LetField;
+class ID;
+class Number;
+class Output;
+class PacPrimitive;
+class Param;
+class ParameterizedType;
+class RecordType;
+class RecordField;
+class RecordDataField;
+class RecordPaddingField;
+class RegEx;
+class SeqEnd;
+class StateVar;
+class Type;
+class TypeDecl;
+class WithInputField;
+
+// The ID of the current declaration.
+extern const ID* current_decl_id;
+
+typedef vector<ActionParam*>		ActionParamList;
+typedef vector<AnalyzerAction*> 	AnalyzerActionList;
+typedef vector<AnalyzerElement*>	AnalyzerElementList;
+typedef vector<Attr*> 			AttrList;
+typedef vector<CaseExpr*> 		CaseExprList;
+typedef vector<CaseField*> 		CaseFieldList;
+typedef vector<ContextField*> 		ContextFieldList;
+typedef vector<Decl*> 			DeclList;
+typedef vector<Enum*>			EnumList;
+typedef vector<Expr*> 			ExprList;
+typedef vector<Field*> 			FieldList;
+typedef vector<LetField*> 		LetFieldList;
+typedef vector<Number*>			NumList;
+typedef vector<Param*> 			ParamList;
+typedef vector<RecordField*> 		RecordFieldList;
+typedef vector<StateVar*>		StateVarList;
+
+#define foreach(i, ct, pc) \
+	if ( pc ) \
+		for ( ct::iterator i = (pc)->begin(); i != (pc)->end(); ++i )
+
+#define delete_list(ct, pc) \
+	{ \
+	foreach(delete_list_i, ct, pc) 		\
+		delete *delete_list_i;		\
+	delete pc;				\
+	pc = 0;					\
+	}
+
+// Constants
+const char * const kComputeFrameLength = "compute_frame_length";
+const char * const kFlowBufferClass = "FlowBuffer";
+const char * const kFlowBufferVar = "flow_buffer";
+const char * const kFlowEOF = "FlowEOF";
+const char * const kFlowGap = "NewGap";
+const char * const kInitialBufferLengthFunc = "initial_buffer_length";
+const char * const kNeedMoreData = "need_more_data";
+const char * const kNewData = "NewData";
+const char * const kParseFuncWithBuffer = "ParseBuffer";
+const char * const kParseFuncWithoutBuffer = "Parse";
+const char * const kRefCountClass = "binpac::RefCount";
+const char * const kTypeWithLengthClass = "binpac::TypeWithLength";
+
+#endif  // pac_common_h

aux/binpac/src/pac_conn.cc

@@ -0,0 +1,169 @@
+#include "pac_analyzer.h"
+#include "pac_dataunit.h"
+#include "pac_embedded.h"
+#include "pac_exception.h"
+#include "pac_expr.h"
+#include "pac_flow.h"
+#include "pac_output.h"
+#include "pac_paramtype.h"
+#include "pac_type.h"
+
+#include "pac_conn.h"
+
+ConnDecl::ConnDecl(ID *conn_id, 
+                   ParamList *params,
+                   AnalyzerElementList *elemlist)
+	: AnalyzerDecl(conn_id, CONN, params)
+	{
+	flows_[0] = flows_[1] = 0;
+	AddElements(elemlist);
+	data_type_ = new ParameterizedType(conn_id->clone(), 0);
+	}
+
+ConnDecl::~ConnDecl()
+	{
+	delete flows_[0];
+	delete flows_[1];
+	}
+
+void ConnDecl::AddBaseClass(vector<string> *base_classes) const
+	{
+	base_classes->push_back("binpac::ConnectionAnalyzer"); 
+	} 
+
+void ConnDecl::ProcessFlowElement(AnalyzerFlow *flow_elem)
+	{
+	int flow_index;
+
+	if ( flow_elem->dir() == AnalyzerFlow::UP )
+		flow_index = 0;
+	else
+		flow_index = 1;
+
+	if ( flows_[flow_index] )
+		{
+		throw Exception(flow_elem,
+		                fmt("%sflow already defined", 
+		                    flow_index == 0 ? "up" : "down"));
+		}
+
+	flows_[flow_index] = flow_elem;
+	type_->AddField(flow_elem->flow_field());
+	}
+
+void ConnDecl::ProcessDataUnitElement(AnalyzerDataUnit *dataunit_elem)
+	{
+	throw Exception(
+		dataunit_elem, 
+		"dataunit should be defined in only a flow declaration");
+	}
+
+void ConnDecl::Prepare()
+	{
+	AnalyzerDecl::Prepare();
+
+	flows_[0]->flow_decl()->set_conn_decl(this);
+	flows_[1]->flow_decl()->set_conn_decl(this);
+	}
+
+void ConnDecl::GenPubDecls(Output *out_h, Output *out_cc)
+	{
+	AnalyzerDecl::GenPubDecls(out_h, out_cc);
+	}
+
+void ConnDecl::GenPrivDecls(Output *out_h, Output *out_cc)
+	{
+	AnalyzerDecl::GenPrivDecls(out_h, out_cc);
+	}
+
+void ConnDecl::GenEOFFunc(Output *out_h, Output *out_cc)
+	{
+	string proto = strfmt("%s(bool is_orig)", kFlowEOF);
+
+	out_h->println("void %s;", proto.c_str());
+
+	out_cc->println("void %s::%s", class_name().c_str(), proto.c_str());
+	out_cc->inc_indent();
+	out_cc->println("{");
+
+	out_cc->println("if ( is_orig )");
+	out_cc->inc_indent();
+	out_cc->println("%s->%s();",
+	                env_->LValue(upflow_id),
+	                kFlowEOF);
+	out_cc->dec_indent();
+	out_cc->println("else");
+	out_cc->inc_indent();
+	out_cc->println("%s->%s();",
+	                env_->LValue(downflow_id),
+	                kFlowEOF);
+
+	foreach(i, AnalyzerHelperList, eof_helpers_)
+		{
+		(*i)->GenCode(0, out_cc, this);
+		}
+
+	out_cc->dec_indent();
+	
+	out_cc->println("}");
+	out_cc->dec_indent();
+	out_cc->println("");
+	}
+
+void ConnDecl::GenGapFunc(Output *out_h, Output *out_cc)
+	{
+	string proto = strfmt("%s(bool is_orig, int gap_length)", kFlowGap);
+
+	out_h->println("void %s;", proto.c_str());
+
+	out_cc->println("void %s::%s", class_name().c_str(), proto.c_str());
+	out_cc->inc_indent();
+	out_cc->println("{");
+
+	out_cc->println("if ( is_orig )");
+	out_cc->inc_indent();
+	out_cc->println("%s->%s(gap_length);",
+	                env_->LValue(upflow_id),
+	                kFlowGap);
+	out_cc->dec_indent();
+	out_cc->println("else");
+	out_cc->inc_indent();
+	out_cc->println("%s->%s(gap_length);",
+	                env_->LValue(downflow_id),
+	                kFlowGap);
+	out_cc->dec_indent();
+	
+	out_cc->println("}");
+	out_cc->dec_indent();
+	out_cc->println("");
+	}
+
+void ConnDecl::GenProcessFunc(Output *out_h, Output *out_cc)
+	{
+	string proto = 
+		strfmt("%s(bool is_orig, const_byteptr begin, const_byteptr end)",
+		       kNewData);
+
+	out_h->println("void %s;", proto.c_str());
+
+	out_cc->println("void %s::%s", class_name().c_str(), proto.c_str());
+	out_cc->inc_indent();
+	out_cc->println("{");
+
+	out_cc->println("if ( is_orig )");
+	out_cc->inc_indent();
+	out_cc->println("%s->%s(begin, end);",
+	                env_->LValue(upflow_id),
+	                kNewData);
+	out_cc->dec_indent();
+	out_cc->println("else");
+	out_cc->inc_indent();
+	out_cc->println("%s->%s(begin, end);",
+	                env_->LValue(downflow_id),
+	                kNewData);
+	out_cc->dec_indent();
+	
+	out_cc->println("}");
+	out_cc->dec_indent();
+	out_cc->println("");
+	}

aux/binpac/src/pac_conn.h

@@ -0,0 +1,34 @@
+#ifndef pac_conn_h
+#define pac_conn_h
+
+#include "pac_decl.h"
+#include "pac_analyzer.h"
+
+class ConnDecl : public AnalyzerDecl
+{
+public:
+	ConnDecl(ID *conn_id, ParamList *params, AnalyzerElementList *elemlist);
+	~ConnDecl();
+
+	void Prepare();
+
+	Type* DataType() const	{ return data_type_; }
+
+protected:
+	void AddBaseClass(vector<string> *base_classes) const;
+
+	void GenProcessFunc(Output *out_h, Output *out_cc);
+	void GenGapFunc(Output *out_h, Output *out_cc);
+	void GenEOFFunc(Output *out_h, Output *out_cc);
+
+	void GenPubDecls(Output *out_h, Output *out_cc);
+	void GenPrivDecls(Output *out_h, Output *out_cc);
+
+	void ProcessFlowElement(AnalyzerFlow *flow_elem);
+	void ProcessDataUnitElement(AnalyzerDataUnit *dataunit_elem);
+
+	AnalyzerFlow *flows_[2];
+	Type *data_type_;
+};
+
+#endif  // pac_conn_h

aux/binpac/src/pac_context.cc

@@ -0,0 +1,120 @@
+#include "pac_analyzer.h"
+#include "pac_exception.h"
+#include "pac_exttype.h"
+#include "pac_flow.h"
+#include "pac_id.h"
+#include "pac_output.h"
+#include "pac_param.h"
+#include "pac_paramtype.h"
+#include "pac_type.h"
+#include "pac_utils.h"
+
+#include "pac_context.h"
+
+ContextField::ContextField(ID *id, Type *type)
+	: Field(CONTEXT_FIELD, 
+		TYPE_NOT_TO_BE_PARSED | CLASS_MEMBER | PUBLIC_READABLE, 
+		id, type)
+	{
+	}
+
+AnalyzerContextDecl *AnalyzerContextDecl::current_analyzer_context_ = 0;
+
+namespace {
+	ParamList *ContextFieldsToParams(ContextFieldList *context_fields)
+		{
+		// Convert context fields to parameters
+		ParamList *params = new ParamList();
+		foreach(i, ContextFieldList, context_fields)
+			{
+			ContextField *f = *i;
+			params->push_back(
+				new Param(f->id()->clone(), 
+				f->type()));
+			}
+		return params;
+		}
+} // namespace private
+
+AnalyzerContextDecl::AnalyzerContextDecl(
+		ID *id, 
+		ContextFieldList *context_fields)
+	: TypeDecl(new ID(fmt("Context%s", id->Name())), 
+		ContextFieldsToParams(context_fields), 
+		new DummyType())
+	{
+	context_name_id_ = id;
+	if ( current_analyzer_context_ != 0 )
+		{
+		throw Exception(this, 
+		                fmt("multiple declaration of analyzer context; "
+		                    "the previous one is `%s'",
+		                    current_analyzer_context_->id()->Name()));
+		}
+	else
+		current_analyzer_context_ = this;
+
+	context_fields_ = context_fields;
+
+	param_type_ = new ParameterizedType(id_->clone(), 0);
+
+	flow_buffer_added_ = false;
+
+	DEBUG_MSG("Context type: %s\n", param_type()->class_name().c_str());
+	}
+
+AnalyzerContextDecl::~AnalyzerContextDecl()
+	{
+	delete context_name_id_;
+	delete_list(ContextFieldList, context_fields_);
+	}
+
+void AnalyzerContextDecl::GenForwardDeclaration(Output *out_h)
+	{
+	GenNamespaceBegin(out_h);
+	TypeDecl::GenForwardDeclaration(out_h);
+	}
+
+void AnalyzerContextDecl::GenCode(Output *out_h, Output *out_cc)
+	{
+	GenNamespaceBegin(out_h);
+	GenNamespaceBegin(out_cc);
+	TypeDecl::GenCode(out_h, out_cc);
+	}
+
+void AnalyzerContextDecl::GenNamespaceBegin(Output *out) const
+	{
+	out->println("namespace %s {", context_name_id()->Name());
+	}
+
+void AnalyzerContextDecl::GenNamespaceEnd(Output *out) const
+	{
+	out->println("} // namespace %s", context_name_id()->Name());
+	}
+
+void AnalyzerContextDecl::AddFlowBuffer()
+	{
+	if ( flow_buffer_added_ )
+		return;
+
+	AddParam(new Param(
+		new ID(kFlowBufferVar), 
+		FlowDecl::flow_buffer_type()->Clone()));
+
+	flow_buffer_added_ = true;
+	}
+
+string AnalyzerContextDecl::mb_buffer(Env *env)
+	{
+	// A hack. The orthodox way would be to build an Expr of
+	// context.flow_buffer_var, and then EvalExpr.
+	return fmt("%s->%s()", 
+		env->RValue(analyzer_context_id), 
+		kFlowBufferVar);
+	}
+
+Type *DummyType::DoClone() const
+	{
+	// Fields will be copied in Type::Clone().
+	return new DummyType();
+	}

aux/binpac/src/pac_context.h

@@ -0,0 +1,97 @@
+#ifndef pac_context_h
+#define pac_context_h
+
+#include "pac_common.h"
+#include "pac_field.h"
+#include "pac_type.h"
+#include "pac_typedecl.h"
+
+// AnalyzerContext represents a cookie that an analyzer gives to
+// parse functions of various message types. The cookie is parsed
+// to every parse function (if necessary) as parameter 'binpac_context'.
+// 
+// The members of the cookie is declared through 'analyzer' declarations,
+// such as in:
+//
+// analyzer SunRPC withcontext {
+//        connection:     RPC_Conn;
+//        flow:           RPC_Flow;
+// };      
+//
+// The cookie usually contains the connection and flow in which 
+// the message appears, and the context information can be 
+// accessed as members of the cookie, such as 
+// ``binpac_context.connection''.
+
+class ContextField : public Field
+{
+public:
+	ContextField(ID *id, Type *type);
+};
+
+class AnalyzerContextDecl : public TypeDecl
+{
+public:
+	AnalyzerContextDecl(ID *id, ContextFieldList *context_fields);
+	~AnalyzerContextDecl();
+
+	void AddFlowBuffer();
+
+	const ID *context_name_id() const { return context_name_id_; }
+
+	// The type of analyzer context as a parameter
+	ParameterizedType *param_type() const { return param_type_; }
+
+	void GenForwardDeclaration(Output *out_h);
+	void GenCode(Output *out_h, Output *out_cc);
+
+	void GenNamespaceBegin(Output *out) const;
+	void GenNamespaceEnd(Output *out) const;
+
+private:
+	ID *context_name_id_;
+	ContextFieldList *context_fields_;
+	ParameterizedType *param_type_;
+	bool flow_buffer_added_;
+
+// static members
+public:
+	static AnalyzerContextDecl *current_analyzer_context()
+		{
+		return current_analyzer_context_;
+		}
+
+	static string mb_buffer(Env *env);
+
+private:
+	static AnalyzerContextDecl *current_analyzer_context_;
+};
+
+class DummyType : public Type 
+{
+public:
+	DummyType() : Type(DUMMY) {}
+
+	bool DefineValueVar() const 	{ return false; }
+	string DataTypeStr() const 	{ ASSERT(0); return ""; }
+
+	int StaticSize(Env* env) const 	{ ASSERT(0); return -1; }
+
+	bool ByteOrderSensitive() const { return false; }
+
+	bool IsPointerType() const 	{ ASSERT(0); return false; }
+
+	void DoGenParseCode(Output* out, Env* env,
+			const DataPtr& data, int flags)
+		{ ASSERT(0); }
+
+	// Generate code for computing the dynamic size of the type
+	void GenDynamicSize(Output* out, Env* env, const DataPtr& data)
+		{ ASSERT(0); }
+
+protected:
+	Type *DoClone()	const;
+	void DoMarkIncrementalInput()		{ ASSERT(0); }
+};
+
+#endif  // pac_context_h

aux/binpac/src/pac_cstr.cc

@@ -0,0 +1,127 @@
+#include "pac_cstr.h"
+#include "pac_dbg.h"
+#include "pac_exception.h"
+
+namespace {
+
+class EscapeException
+{
+public:
+	explicit EscapeException(const string &s)
+		{
+		msg_ = s;
+		}
+
+	const string &msg() const	{ return msg_; }
+
+private:
+	string msg_;
+};
+
+// Copied from util.cc of Bro
+int expand_escape(const char*& s)
+	{
+	switch ( *(s++) ) {
+	case 'b': return '\b';
+	case 'f': return '\f';
+	case 'n': return '\n';
+	case 'r': return '\r';
+	case 't': return '\t';
+	case 'a': return '\a';
+	case 'v': return '\v';
+
+	case '0': case '1': case '2': case '3': case '4':
+	case '5': case '6': case '7':
+		{ // \<octal>{1,3}
+		--s;	// put back the first octal digit
+		const char* start = s;
+
+		// Don't increment inside loop control
+		// because if isdigit() is a macro it might
+		// expand into multiple increments ...
+
+		// Here we define a maximum length for escape sequence
+		// to allow easy handling of string like: "^H0" as
+		// "\0100".
+
+		for ( int len = 0; len < 3 && isascii(*s) && isdigit(*s); ++s, ++len)
+			;
+
+		int result;
+		if ( sscanf(start, "%3o", &result) != 1 )
+			{
+			throw EscapeException(fmt("bad octal escape: \"%s", start));
+			result = 0;
+			}
+
+		return result;
+		}
+
+	case 'x':
+		{ /* \x<hex> */
+		const char* start = s;
+
+		// Look at most 2 characters, so that "\x0ddir" -> "^Mdir".
+		for ( int len = 0; len < 2 && isascii(*s) && isxdigit(*s);
+		      ++s, ++len)
+			;
+
+		int result;
+		if ( sscanf(start, "%2x", &result) != 1 )
+			{
+			throw EscapeException(fmt("bad hexadecimal escape: \"%s", start));
+			result = 0;
+			}
+
+		return result;
+		}
+
+	default:
+		return s[-1];
+	}
+	}
+
+}  // private namespace
+
+ConstString::ConstString(const string &s)
+	: str_(s)
+	{
+	// Copied from scan.l of Bro
+	try 
+		{
+		const char* text = str_.c_str();
+		int len = strlen(text) + 1;
+		int i = 0;
+
+		char* s = new char[len];
+
+		// Skip leading quote.
+		for ( ++text; *text; ++text )
+			{
+			if ( *text == '\\' )
+				{
+				++text;	// skip '\'
+				s[i++] = expand_escape(text);
+				--text;	// point to end of sequence
+				}
+			else
+				{
+				s[i++] = *text;
+				}
+			}
+		ASSERT(i < len);
+
+		// Get rid of trailing quote.
+		ASSERT(s[i-1] == '"');
+		s[i-1] = '\0';
+
+		unescaped_ = s;
+		delete [] s;
+		}
+	catch(EscapeException const &e)
+		{
+		// Throw again with the object
+		throw Exception(this, e.msg().c_str());
+		}
+	}
+

aux/binpac/src/pac_cstr.h

@@ -0,0 +1,23 @@
+#ifndef pac_cstr_h
+#define pac_cstr_h
+
+#include "pac_common.h"
+
+class ConstString : public Object
+{
+public:
+	ConstString(const string &s);
+
+	// The string in its escaped form, with surrounding '"'s
+	const string &str() const	{ return str_; }
+	const char *c_str() const	{ return str_.c_str(); }
+
+	// The unescaped string, without surrounding '"'s
+	const string &unescaped() const { return unescaped_; }
+
+private:
+	string str_;
+	string unescaped_;
+};
+
+#endif  // pac_cstr_h

aux/binpac/src/pac_ctype.cc

@@ -0,0 +1,21 @@
+#include "pac_ctype.h"
+
+string CType::DeclareInstance(const string &var) const
+	{
+	return strfmt("%s %s", name().c_str(), var.c_str());
+	}
+
+string CType::DeclareConstReference(const string &var) const
+	{
+	return strfmt("%s const &%s", name().c_str(), var.c_str());
+	}
+
+string CType::DeclareConstPointer(const string &var) const
+	{
+	return strfmt("%s const *%s", name().c_str(), var.c_str());
+	}
+
+string CType::DeclarePointer(const string &var) const
+	{
+	return strfmt("%s *%s", name().c_str(), var.c_str());
+	}

aux/binpac/src/pac_ctype.h

@@ -0,0 +1,23 @@
+#ifndef pac_ctype_h
+#define pac_ctype_h
+
+#include "pac_common.h"
+
+// Represents a C++ type
+class CType
+{
+public:
+	CType(const string &name);
+
+	string name() const 	{ return name_; }
+
+	string DeclareInstance(const string &var) const;
+	string DeclareConstReference(const string &var) const;
+	string DeclareConstPointer(const string &var) const;
+	string DeclarePointer(const string &var) const;
+
+protected:
+	string name_;
+};
+
+#endif  // pac_ctype_h

aux/binpac/src/pac_datadep.cc

@@ -0,0 +1,76 @@
+#include "pac_datadep.h"
+#include "pac_expr.h"
+#include "pac_id.h"
+#include "pac_type.h"
+
+DataDepElement::DataDepElement(DDE_Type type) 
+	: dde_type_(type), in_traversal(false)
+	{ 
+	}
+
+bool DataDepElement::Traverse(DataDepVisitor *visitor)
+	{
+	// Avoid infinite loop
+	if ( in_traversal )
+		return true;
+	if ( ! visitor->PreProcess(this) )
+		return false;
+
+	in_traversal = true;
+	bool cont = DoTraverse(visitor);
+	in_traversal = false;
+
+	if ( ! cont )
+		return false;
+	if ( ! visitor->PostProcess(this) )
+		return false;
+	return true;
+	}
+
+Expr *DataDepElement::expr()
+	{ 
+	return static_cast<Expr *>(this); 
+	}
+
+Type *DataDepElement::type()
+	{ 
+	return static_cast<Type *>(this); 
+	}
+
+bool RequiresAnalyzerContext::PreProcess(DataDepElement *element)
+	{
+	switch ( element->dde_type() ) 
+		{
+		case DataDepElement::EXPR:
+			ProcessExpr(element->expr());
+			break;
+		default:
+			break;
+		}
+
+	// Continue traversal until we know the answer is 'yes'
+	return ! requires_analyzer_context_;
+	}
+
+bool RequiresAnalyzerContext::PostProcess(DataDepElement *element)
+	{
+	return ! requires_analyzer_context_;
+	}
+
+void RequiresAnalyzerContext::ProcessExpr(Expr *expr)
+	{
+	if ( expr->expr_type() == Expr::EXPR_ID )
+		{
+		requires_analyzer_context_ = 
+			(requires_analyzer_context_ || 
+		       	 *expr->id() == *analyzer_context_id ||
+		       	 *expr->id() == *context_macro_id);
+		}
+	}
+
+bool RequiresAnalyzerContext::compute(DataDepElement *element)
+	{
+	RequiresAnalyzerContext visitor;
+	element->Traverse(&visitor);
+	return visitor.requires_analyzer_context_;
+	}

aux/binpac/src/pac_datadep.h

@@ -0,0 +1,71 @@
+#ifndef pac_datadep_h
+#define pac_datadep_h
+
+// To provide a way to traverse through the data dependency graph.
+// That is, to evaluate X, what must be evaluated.
+
+#include "pac_common.h"
+#include "pac_dbg.h"
+
+class DataDepVisitor;
+
+class DataDepElement {
+public:
+	enum DDE_Type {
+		ATTR,
+		CASEEXPR,
+		EXPR,
+		FIELD,
+		INPUT_BUFFER,
+		PARAM,
+		TYPE,
+	};
+
+	DataDepElement(DDE_Type type);
+	virtual ~DataDepElement() {}
+
+	// Returns whether to continue traversal
+	bool Traverse(DataDepVisitor *visitor);
+
+	// Returns whether to continue traversal
+	virtual bool DoTraverse(DataDepVisitor *visitor) = 0;
+
+	DDE_Type dde_type() const	{ return dde_type_; }
+	Expr *expr();
+	Type *type();
+
+protected:
+	DDE_Type dde_type_;
+	bool in_traversal;
+};
+
+class DataDepVisitor {
+public:
+	virtual ~DataDepVisitor() {}
+	// Returns whether to continue traversal
+	virtual bool PreProcess(DataDepElement *element) = 0;
+	virtual bool PostProcess(DataDepElement *element) = 0;
+};
+
+class RequiresAnalyzerContext : public DataDepVisitor {
+public:
+	RequiresAnalyzerContext() : requires_analyzer_context_(false) {}
+
+	// Returns whether to continue traversal
+	bool PreProcess(DataDepElement *element);
+	bool PostProcess(DataDepElement *element);
+
+	bool requires_analyzer_context() const
+		{
+		return requires_analyzer_context_;
+		}
+
+	static bool compute(DataDepElement *element);
+
+protected:
+	void ProcessExpr(Expr *expr);
+
+	bool requires_analyzer_context_;
+};
+
+#endif // pac_datadep_h

aux/binpac/src/pac_dataptr.cc

@@ -0,0 +1,66 @@
+#include "pac_exception.h"
+#include "pac_id.h"
+#include "pac_output.h"
+#include "pac_utils.h"
+
+#include "pac_dataptr.h"
+
+DataPtr::DataPtr(Env* env, const ID* id, const int offset)
+	: id_(id), offset_(offset)
+	{
+	if ( id_ )
+		{
+		if ( ! env->Evaluated(id_) )
+			throw ExceptionIDNotEvaluated(id_);
+
+		if ( offset_ == 0 )
+			ptr_expr_ = strfmt("%s", env->RValue(id_));
+		else
+			ptr_expr_ = strfmt("(%s + %d)", env->RValue(id_), offset_);
+		}
+	else
+		ptr_expr_ = "(null id)";
+	}
+
+int DataPtr::AbsOffset(const ID* base_ptr) const
+	{
+	return ( id() == base_ptr ) ? offset() : -1;
+	}
+
+char* DataPtr::AbsOffsetExpr(Env* env, const ID* base_ptr) const
+	{
+	if ( AbsOffset(base_ptr) >= 0 )
+		return nfmt("%d", offset());
+	else
+		return nfmt("(%s - %s)", ptr_expr(), env->RValue(base_ptr));
+	}
+
+void DataPtr::GenBoundaryCheck(Output* out_cc, Env* env,
+		const char* data_size, const char* data_name) const
+	{
+	ASSERT(id_);
+
+	out_cc->println("// Checking out-of-bound for \"%s\"", data_name);
+	out_cc->println("if ( %s + (%s) > %s )",
+		ptr_expr(),
+		data_size,
+		env->RValue(end_of_data));
+
+	out_cc->inc_indent(); 
+	out_cc->println("{");
+
+	char* data_offset = AbsOffsetExpr(env, begin_of_data); 
+
+	out_cc->println("// Handle out-of-bound condition");
+	out_cc->println("throw ExceptionOutOfBound(\"%s\",", data_name);
+	out_cc->println("	(%s) + (%s), ", 
+		data_offset, data_size);
+	out_cc->println("	(%s) - (%s));", 
+		env->RValue(end_of_data), env->RValue(begin_of_data));
+
+	delete [] data_offset;
+
+	out_cc->println("}");
+	out_cc->dec_indent(); 
+	}
+

aux/binpac/src/pac_dataptr.h

@@ -0,0 +1,47 @@
+#ifndef pac_dataptr_h
+#define pac_dataptr_h
+
+#include <string>
+#include "pac_common.h"
+
+// A data pointer is represented by an data pointer variable
+// plus a constant offset.
+
+class DataPtr
+{
+public:
+	DataPtr(Env* env, const ID* arg_id, const int arg_off);
+
+	DataPtr const &operator=(DataPtr const &x)
+		{
+		id_ = x.id();
+		offset_ = x.offset();
+		ptr_expr_ = x.ptr_expr();
+
+		return *this;
+		}
+
+	const ID* id() const 	{ return id_; }
+	int offset() const 	{ return offset_; }
+
+	const char* ptr_expr() const
+		{
+		ASSERT(id_); 
+		return ptr_expr_.c_str(); 
+		}
+
+	int AbsOffset(const ID* base_ptr) const;
+	char* AbsOffsetExpr(Env* env, const ID* base_ptr) const;
+
+	void GenBoundaryCheck(Output* out, 
+                              Env* env, 
+                              const char* data_size, 
+                              const char* data_name) const;
+
+protected:
+	const ID* id_;
+	int offset_;
+	string ptr_expr_;
+};
+
+#endif  // pac_dataptr_h

aux/binpac/src/pac_dataunit.cc

@@ -0,0 +1,60 @@
+#include "pac_context.h"
+#include "pac_dataunit.h"
+#include "pac_output.h"
+#include "pac_paramtype.h"
+#include "pac_varfield.h"
+
+AnalyzerDataUnit::AnalyzerDataUnit(
+		DataUnitType type, 
+		ID *id, 
+		ExprList *type_params,
+		ExprList *context_params)
+	: AnalyzerElement(DATAUNIT),
+	  type_(type),
+	  id_(id),
+	  type_params_(type_params),
+	  context_params_(context_params) 
+	{
+	data_type_ = new ParameterizedType(id_, type_params_);
+	context_type_ = new ParameterizedType(
+		AnalyzerContextDecl::current_analyzer_context()->id()->clone(),
+		context_params_);
+
+	dataunit_var_field_ = new ParseVarField(
+		Field::CLASS_MEMBER,
+		dataunit_id->clone(),
+		data_type());
+	context_var_field_ = new PrivVarField(
+		analyzer_context_id->clone(),
+		context_type());
+	}
+
+AnalyzerDataUnit::~AnalyzerDataUnit()
+	{
+	delete dataunit_var_field_;
+	delete context_var_field_;
+	}
+
+void AnalyzerDataUnit::Prepare(Env *env)
+	{
+	dataunit_var_field_->Prepare(env);
+	context_var_field_->Prepare(env);
+	}
+
+void AnalyzerDataUnit::GenNewDataUnit(Output *out_cc, Env *env)
+	{
+	out_cc->println("%s = new %s(%s);",
+		env->LValue(dataunit_id),
+		data_type()->class_name().c_str(),
+		data_type()->EvalParameters(out_cc, env).c_str());
+	}
+
+void AnalyzerDataUnit::GenNewContext(Output *out_cc, Env *env)
+	{
+	out_cc->println("%s = new %s(%s);", 
+		env->LValue(analyzer_context_id),
+		context_type()->class_name().c_str(),
+		context_type()->EvalParameters(out_cc, env).c_str());
+	env->SetEvaluated(analyzer_context_id);
+	}
+

aux/binpac/src/pac_dataunit.h

@@ -0,0 +1,49 @@
+#ifndef pac_dataunit_h
+#define pac_dataunit_h
+
+#include "pac_analyzer.h"
+
+// The type and parameters of input data unit of a flow. For instance, the
+// data unit of a DCE/RPC flow is DCE_RPC_PDU.
+
+class AnalyzerDataUnit : public AnalyzerElement
+{
+public:
+	enum DataUnitType { DATAGRAM, FLOWUNIT };
+	AnalyzerDataUnit(
+		DataUnitType type, 
+		ID *id, 
+		ExprList *type_params, 
+		ExprList *context_params);
+	~AnalyzerDataUnit();
+
+	void Prepare(Env *env);
+
+	// Initializes dataunit_id
+	void GenNewDataUnit(Output *out_cc, Env *env);
+	// Initializes analyzer_context_id
+	void GenNewContext(Output *out_cc, Env *env);
+
+	DataUnitType type() const		{ return type_; }
+	const ID *id() const			{ return id_; }
+	ExprList *type_params() const		{ return type_params_; }
+	ExprList *context_params() const	{ return context_params_; }
+
+	ParameterizedType *data_type() const	{ return data_type_; }
+	ParameterizedType *context_type() const	{ return context_type_; }
+
+	Field *dataunit_var_field() const	{ return dataunit_var_field_; }
+	Field *context_var_field() const	{ return context_var_field_; }
+
+private:
+	DataUnitType type_;
+	ID *id_;
+	ExprList *type_params_;
+	ExprList *context_params_;
+	ParameterizedType *data_type_;
+	ParameterizedType *context_type_;
+	Field *dataunit_var_field_;
+	Field *context_var_field_;
+};
+
+#endif // pac_dataunit_h

aux/binpac/src/pac_dbg.h

@@ -0,0 +1,14 @@
+/* $Id: pac_dbg.h 3265 2006-06-09 21:16:12Z rpang $ */
+
+#ifndef pac_dbg_h
+#define pac_dbg_h
+
+#include <assert.h>
+#include <stdio.h>
+
+extern bool FLAGS_pac_debug;
+
+#define ASSERT(x)	assert(x)
+#define DEBUG_MSG(x...)	if ( FLAGS_pac_debug ) fprintf(stderr, x)
+
+#endif /* pac_dbg_h */

aux/binpac/src/pac_decl-inl.h

@@ -0,0 +1,6 @@
+#ifndef pac_decl_inl_h
+#define pac_decl_inl_h
+
+#include "pac_id.h"
+
+#endif  // pac_decl_inl_h

aux/binpac/src/pac_decl.cc

@@ -0,0 +1,191 @@
+#include "pac_attr.h"
+#include "pac_context.h"
+#include "pac_dataptr.h"
+#include "pac_embedded.h"
+#include "pac_exception.h"
+#include "pac_expr.h"
+#include "pac_exttype.h"
+#include "pac_id.h"
+#include "pac_output.h"
+#include "pac_param.h"
+#include "pac_record.h"
+#include "pac_type.h"
+#include "pac_utils.h"
+
+#include "pac_decl.h"
+
+DeclList *Decl::decl_list_ = 0;
+Decl::DeclMap Decl::decl_map_;
+
+Decl::Decl(ID* id, DeclType decl_type)
+	: id_(id), decl_type_(decl_type), attrlist_(0)
+	{
+	decl_map_[id_] = this;
+	if ( ! decl_list_ )
+		decl_list_ = new DeclList();
+	decl_list_->push_back(this);
+
+	DEBUG_MSG("Finished Decl %s\n", id_->Name());
+
+	analyzer_context_ = 0;
+	}
+
+Decl::~Decl()
+	{
+	delete id_;
+	delete_list(AttrList, attrlist_);
+	}
+
+void Decl::AddAttrs(AttrList* attrs)
+	{
+	if ( ! attrs )
+		return;
+	if ( ! attrlist_ )
+		attrlist_ = new AttrList();
+	foreach ( i, AttrList, attrs )
+		{
+		attrlist_->push_back(*i);
+		ProcessAttr(*i);
+		}
+	}
+
+void Decl::ProcessAttr(Attr *attr)
+	{
+	throw Exception(attr, "unhandled attribute");
+	}
+
+void Decl::SetAnalyzerContext()
+	{
+	analyzer_context_ = 
+		AnalyzerContextDecl::current_analyzer_context();
+	if ( ! analyzer_context_ )
+		{
+		throw Exception(this, 
+		                "analyzer context not defined");
+		}
+	}
+
+void Decl::ProcessDecls(Output *out_h, Output *out_cc)
+	{
+	if ( ! decl_list_ )
+		return;
+
+	foreach(i, DeclList, decl_list_)
+		{
+		Decl *decl = *i;
+		current_decl_id = decl->id();
+		decl->Prepare();
+		}
+
+	foreach(i, DeclList, decl_list_)
+		{
+		Decl *decl = *i;
+		current_decl_id = decl->id();
+		decl->GenExternDeclaration(out_h);
+		}
+
+	out_h->println("namespace binpac {\n");
+	out_cc->println("namespace binpac {\n");
+
+	AnalyzerContextDecl *analyzer_context =
+		AnalyzerContextDecl::current_analyzer_context();
+
+	foreach(i, DeclList, decl_list_)
+		{
+		Decl *decl = *i;
+		current_decl_id = decl->id();
+		decl->GenForwardDeclaration(out_h);
+		}
+
+	if ( analyzer_context )
+		analyzer_context->GenNamespaceEnd(out_h);
+
+	out_h->println("");
+
+	foreach(i, DeclList, decl_list_)
+		{
+		Decl *decl = *i;
+		current_decl_id = decl->id();
+		decl->GenCode(out_h, out_cc);
+		}
+
+	if ( analyzer_context )
+		{
+		analyzer_context->GenNamespaceEnd(out_h);
+		analyzer_context->GenNamespaceEnd(out_cc);
+		}
+
+	out_h->println("}  // namespace binpac");
+	out_cc->println("}  // namespace binpac");
+	}
+
+Decl* Decl::LookUpDecl(const ID* id)
+	{
+	DeclMap::iterator it = decl_map_.find(id);
+	if ( it == decl_map_.end() )
+		return 0;
+	return it->second;
+	}
+
+int HelperDecl::helper_id_seq = 0;
+
+HelperDecl::HelperDecl(HelperType helper_type, 
+                       ID* context_id, 
+                       EmbeddedCode* code)
+ 	: Decl(new ID(fmt("helper_%d", ++helper_id_seq)), HELPER), 
+	  helper_type_(helper_type),
+	  context_id_(context_id),
+	  code_(code)
+	{
+	}
+
+HelperDecl::~HelperDecl()
+	{
+	delete context_id_;
+	delete code_;
+	}
+
+void HelperDecl::Prepare()
+	{
+	// Do nothing
+	}
+
+void HelperDecl::GenExternDeclaration(Output *out_h)
+	{
+	if ( helper_type_ == EXTERN )
+		code_->GenCode(out_h, global_env());
+	}
+
+void HelperDecl::GenCode(Output *out_h, Output *out_cc)
+	{
+	Env *env = global_env();
+
+#if 0
+	if ( context_id_ )
+		{
+		Decl *decl = Decl::LookUpDecl(context_id_);
+		if ( ! decl )
+			{
+			throw Exception(context_id_, 
+			                fmt("cannot find declaration for %s", 
+			                    context_id_->Name()));
+			}
+		env = decl->env();
+		if ( ! env )
+			{
+			throw Exception(context_id_,
+			                fmt("not a type or analyzer: %s",
+			                    context_id_->Name()));
+			}
+		}
+#endif
+
+	if ( helper_type_ == HEADER )
+		code_->GenCode(out_h, env);
+	else if ( helper_type_ == CODE )
+		code_->GenCode(out_cc, env);
+	else if ( helper_type_ == EXTERN )
+		; // do nothing
+	else
+		ASSERT(0);
+	}

aux/binpac/src/pac_decl.h

@@ -0,0 +1,81 @@
+#ifndef pac_decl_h
+#define pac_decl_h
+
+#include "pac_common.h"
+#include "pac_id.h"
+
+class Decl : public Object
+{
+public:
+	// Note: ANALYZER is not for AnalyzerDecl (which is an
+	// abstract class) , but for AnalyzerContextDecl.
+	enum DeclType { ENUM, LET, TYPE, FUNC, CONN, FLOW, ANALYZER, HELPER, REGEX };
+
+	Decl(ID *id, DeclType decl_type);
+	virtual ~Decl();
+
+	const ID *id() const		{ return id_; }
+	DeclType decl_type() const	{ return decl_type_; }
+	AnalyzerContextDecl *analyzer_context() const 
+		{ return analyzer_context_; }
+
+	// NULL except for TypeDecl or AnalyzerDecl
+	virtual Env *env() const	{ return 0; }
+
+	virtual void Prepare() = 0;
+
+	// Generate declarations out of the "binpac" namespace
+	virtual void GenExternDeclaration(Output *out_h) { /* do nothing */ }
+
+	// Generate declarations before definition of classes
+	virtual void GenForwardDeclaration(Output *out_h) = 0;
+
+	virtual void GenCode(Output *out_h, Output *out_cc) = 0;
+
+	void TakeExprList();
+	void AddAttrs(AttrList *attrlist);
+	void SetAnalyzerContext();
+
+protected:
+	virtual void ProcessAttr(Attr *a);
+
+	ID *id_;
+	DeclType decl_type_;
+	AttrList *attrlist_;
+	ExprList *expr_list_;
+	AnalyzerContextDecl *analyzer_context_;
+
+public:
+	static void ProcessDecls(Output *out_h, Output *out_cc);
+	static Decl *LookUpDecl(const ID *id);
+
+private:
+	static DeclList *decl_list_;
+	typedef map<const ID *, Decl*, ID_ptr_cmp> DeclMap;
+	static DeclMap decl_map_;
+};
+
+class HelperDecl : public Decl
+{
+public:
+	enum HelperType {
+		HEADER, CODE, EXTERN,
+	};
+	HelperDecl(HelperType type, ID *context_id, EmbeddedCode *code);
+	~HelperDecl();
+
+	void Prepare();
+	void GenExternDeclaration(Output *out_h);
+	void GenForwardDeclaration(Output *out_h) { /* do nothing */ }
+	void GenCode(Output *out_h, Output *out_cc);
+
+private:
+	HelperType helper_type_;
+	ID *context_id_;
+	ID *helper_id_;
+	EmbeddedCode *code_;
+
+	static int helper_id_seq;
+};
+
+#endif  // pac_decl_h

aux/binpac/src/pac_embedded.cc

@@ -0,0 +1,82 @@
+#include "pac_id.h"
+#include "pac_primitive.h"
+#include "pac_output.h"
+
+#include "pac_embedded.h"
+
+EmbeddedCodeSegment::EmbeddedCodeSegment(const string &s)
+	: s_(s), primitive_(0)
+	{
+	}
+
+EmbeddedCodeSegment::EmbeddedCodeSegment(PacPrimitive *primitive)
+	: s_(""), primitive_(primitive)
+	{
+	}
+
+EmbeddedCodeSegment::~EmbeddedCodeSegment()
+	{
+	delete primitive_;
+	}
+
+string EmbeddedCodeSegment::ToCode(Env *env)
+	{
+	if ( primitive_ &&  s_.empty() )
+		s_ = primitive_->ToCode(env);
+	return s_;
+	}
+
+EmbeddedCode::EmbeddedCode()
+	{
+	segments_ = new EmbeddedCodeSegmentList();
+	}
+
+EmbeddedCode::~EmbeddedCode()
+	{
+	delete_list(EmbeddedCodeSegmentList, segments_);
+	}
+
+void EmbeddedCode::Append(int atom)
+	{
+	current_segment_ += static_cast<char>(atom);
+	}
+
+void EmbeddedCode::Append(const char *str)
+	{
+	current_segment_ += str;
+	}
+
+void EmbeddedCode::Append(PacPrimitive *primitive)
+	{
+	if ( ! current_segment_.empty() )
+		{
+		segments_->push_back(new EmbeddedCodeSegment(current_segment_));
+		current_segment_ = "";
+		}
+	segments_->push_back(new EmbeddedCodeSegment(primitive));
+	}
+
+void EmbeddedCode::GenCode(Output *out, Env *env)
+	{
+	if ( ! current_segment_.empty() )
+		{
+		segments_->push_back(new EmbeddedCodeSegment(current_segment_));
+		current_segment_ = "";
+		}
+
+	// TODO: return to the generated file after embedded code
+	// out->print("#line %d \"%s\"\n", line_num, filename.c_str());
+
+	// Allow use of RValue for undefined ID, in which case the
+	// ID's name is used as its RValue
+	env->set_allow_undefined_id(true);
+
+	foreach(i, EmbeddedCodeSegmentList, segments_)
+		{
+		EmbeddedCodeSegment *segment = *i;
+		out->print("%s", segment->ToCode(env).c_str());
+		}
+
+	env->set_allow_undefined_id(false);
+	out->print("\n");
+	}

aux/binpac/src/pac_embedded.h

@@ -0,0 +1,42 @@
+#ifndef pac_embedded_h
+#define pac_embedded_h
+
+#include "pac_common.h"
+
+class EmbeddedCodeSegment
+{
+public:
+	explicit EmbeddedCodeSegment(const string &s);
+	explicit EmbeddedCodeSegment(PacPrimitive *primitive);
+	~EmbeddedCodeSegment();
+
+	string ToCode(Env *env);
+
+private:
+	string s_;
+	PacPrimitive *primitive_;
+};
+
+typedef vector<EmbeddedCodeSegment *> EmbeddedCodeSegmentList;
+
+class EmbeddedCode : public Object
+{
+public:
+	EmbeddedCode();
+	~EmbeddedCode();
+
+	// Append a character
+	void Append(int atom);
+	void Append(const char *str);
+
+	// Append a PAC primitive
+	void Append(PacPrimitive *primitive);
+
+	void GenCode(Output *out, Env *env);
+
+private:
+	string current_segment_;
+	EmbeddedCodeSegmentList *segments_;
+};
+
+#endif  // pac_embedded_h

aux/binpac/src/pac_enum.cc

@@ -0,0 +1,70 @@
+#include "pac_exception.h"
+#include "pac_enum.h"
+#include "pac_expr.h"
+#include "pac_exttype.h"
+#include "pac_output.h"
+#include "pac_typedecl.h"
+
+Enum::Enum(ID* id, Expr* expr)
+	: id_(id), expr_(expr)
+	{
+	}
+
+Enum::~Enum()
+	{
+	delete id_;
+	delete expr_;
+	}
+
+void Enum::GenHeader(Output* out_h, int *pval)
+	{
+	ASSERT(pval);
+	if ( expr_ )
+		{
+		if ( ! expr_->ConstFold(global_env(), pval) )
+			throw ExceptionNonConstExpr(expr_);
+		out_h->println("%s = %d,", id_->Name(), *pval);
+		}
+	else
+		out_h->println("%s,", id_->Name());
+	global_env()->AddConstID(id_, *pval);
+	}
+
+EnumDecl::EnumDecl(ID *id, EnumList *enumlist)
+	: Decl(id, ENUM), enumlist_(enumlist) 
+	{
+	ID *type_id = id->clone();
+	datatype_ = new ExternType(type_id, ExternType::NUMBER);
+	extern_typedecl_ = new TypeDecl(type_id, 0, datatype_);
+	}
+
+EnumDecl::~EnumDecl()
+	{ 
+	delete_list(EnumList, enumlist_); 
+	delete extern_typedecl_;
+	}
+
+void EnumDecl::Prepare() 
+	{ 
+	// Do nothing 
+	}
+
+void EnumDecl::GenForwardDeclaration(Output *out_h)
+	{ 
+	out_h->println("enum %s {", id_->Name());
+	out_h->inc_indent();
+	int c = 0;
+	foreach(i, EnumList, enumlist_)
+		{
+		(*i)->GenHeader(out_h, &c);
+		++c;
+		}
+	out_h->dec_indent();
+	out_h->println("};");
+	}
+
+void EnumDecl::GenCode(Output* out_h, Output* /* out_cc */)
+	{
+	// Do nothing 
+	}
+

aux/binpac/src/pac_enum.h

@@ -0,0 +1,37 @@
+#ifndef pac_enum_h
+#define pac_enum_h
+
+#include "pac_decl.h"
+
+class Enum
+{
+public:
+	Enum(ID *id, Expr *expr = 0);
+	~Enum();
+
+	void GenHeader(Output *out_h, int *pval);
+
+private:
+	ID *id_;
+	Expr *expr_;
+};
+
+class EnumDecl : public Decl
+{
+public:
+	EnumDecl(ID *id, EnumList *enumlist);
+	~EnumDecl();
+
+	Type *DataType() const	{ return datatype_; }
+
+	void Prepare();
+	void GenForwardDeclaration(Output *out_h);
+	void GenCode(Output *out_h, Output *out_cc);
+
+private:
+	EnumList *enumlist_;
+	Type *datatype_;
+	TypeDecl *extern_typedecl_;
+};
+
+#endif // pac_enum_h

aux/binpac/src/pac_exception.cc

@@ -0,0 +1,70 @@
+#include "pac_exception.h"
+#include "pac_expr.h"
+#include "pac_id.h"
+#include "pac_utils.h"
+
+Exception::Exception(const Object* o, const char* msg)
+	{
+	if ( o )
+		{
+		msg_ = o->Location();
+		msg_ += ": error : ";
+		}
+	if ( msg )
+		msg_ += msg;
+	if ( FLAGS_pac_debug ) 
+		{
+		DEBUG_MSG("Exception: %s\n", msg_.c_str());
+		abort();
+		}
+	}
+
+ExceptionIDNotFound::ExceptionIDNotFound(const ID* id)
+	: Exception(id), id_(id) 
+	{
+	append(fmt("`%s' undeclared", id_->Name()));
+	}
+
+ExceptionIDRedefinition::ExceptionIDRedefinition(const ID* id)
+	: Exception(id), id_(id) 
+	{
+	append(fmt("`%s' redefined", id_->Name()));
+	}
+
+ExceptionIDNotEvaluated::ExceptionIDNotEvaluated(const ID* id)
+	: Exception(id), id_(id) 
+	{
+	append(fmt("ID `%s' not evaluated before used", id->Name()));
+	}
+
+ExceptionIDNotField::ExceptionIDNotField(const ID* id)
+	: Exception(id), id_(id) 
+	{
+	append(fmt("ID `%s' is not a field", id_->Name()));
+	}
+
+ExceptionMemberNotFound::ExceptionMemberNotFound(const ID* type_id, 
+                                                 const ID *member_id)
+	: Exception(member_id), type_id_(type_id), member_id_(member_id)
+	{
+	append(fmt("type %s does not have member `%s'",
+	           type_id_->Name(), member_id_->Name()));
+	}
+
+ExceptionCyclicDependence::ExceptionCyclicDependence(const ID* id)
+	: Exception(id), id_(id) 
+	{
+	append(fmt("cyclic dependence through `%s'", id_->Name()));
+	}
+
+ExceptionPaddingError::ExceptionPaddingError(const Object* o, const char* msg)
+	: Exception(o)
+	{
+	append(msg);
+	}
+
+ExceptionNonConstExpr::ExceptionNonConstExpr(const Expr* expr)
+	: Exception(expr)
+	{
+	append(fmt("Expression `%s' is not constant", expr->orig()));
+	}

aux/binpac/src/pac_exception.h

@@ -0,0 +1,97 @@
+// $Id: pac_exception.h 3225 2006-06-08 00:00:01Z vern $
+
+#ifndef pac_exception_h
+#define pac_exception_h
+
+#include <string>
+using namespace std;
+
+#include "pac_common.h"
+
+class Exception
+{
+public:
+	Exception(const Object* o, const char* msg = 0);
+
+	const char* msg() const 	{ return msg_.c_str(); }
+	void append(const char* s) 	{ msg_ += s; }
+
+private:
+	string msg_;
+};
+
+class ExceptionIDNotFound : public Exception
+{
+public:
+	ExceptionIDNotFound(const ID* id);
+	const ID* id() const { return id_; }
+
+private:
+	const ID* id_;
+};
+
+class ExceptionIDRedefinition : public Exception
+{
+public:
+	ExceptionIDRedefinition(const ID* id);
+	const ID* id() const { return id_; }
+
+private:
+	const ID* id_;
+};
+
+class ExceptionIDNotEvaluated : public Exception
+{
+public:
+	ExceptionIDNotEvaluated(const ID* id);
+	const ID* id() const { return id_; }
+
+private:
+	const ID* id_;
+};
+
+class ExceptionCyclicDependence : public Exception
+{
+public:
+	ExceptionCyclicDependence(const ID* id);
+	const ID* id() const { return id_; }
+
+private:
+	const ID* id_;
+};
+
+class ExceptionPaddingError : public Exception
+{
+public:
+	ExceptionPaddingError(const Object* o, const char* msg);
+};
+
+class ExceptionIDNotField : public Exception
+{
+public:
+	ExceptionIDNotField(const ID* id);
+	const ID* id() const { return id_; }
+
+private:
+	const ID* id_;
+};
+
+class ExceptionMemberNotFound : public Exception
+{
+public:
+	ExceptionMemberNotFound(const ID* type_id, const ID *member_id);
+
+private:
+	const ID *type_id_, *member_id_;
+};
+
+class ExceptionNonConstExpr : public Exception
+{
+public:
+	ExceptionNonConstExpr(const Expr* expr);
+
+private:
+	const Expr *expr;
+};
+
+#endif /* pac_exception_h */

aux/binpac/src/pac_expr.def

@@ -0,0 +1,34 @@
+EXPR_DEF(EXPR_ID, 		0, "%s")
+EXPR_DEF(EXPR_NUM,		0, "%s")
+EXPR_DEF(EXPR_CSTR,		0, "%s")
+EXPR_DEF(EXPR_REGEX,		0, "REGEX(%s)")
+EXPR_DEF(EXPR_SUBSCRIPT,	2, "@element@(%s[%s])") 
+EXPR_DEF(EXPR_MEMBER,		2, "@%s->%s@") 
+EXPR_DEF(EXPR_PAREN,		1, " ( %s ) ") 
+EXPR_DEF(EXPR_CALL,		1, "%s(%s)")
+EXPR_DEF(EXPR_CALLARGS,		-1, "@custom@")
+EXPR_DEF(EXPR_SIZEOF,		1, "@sizeof(%s)@")
+EXPR_DEF(EXPR_OFFSETOF,		1, "@offsetof(%s)@")
+EXPR_DEF(EXPR_NEG,		1, "-%s") 
+EXPR_DEF(EXPR_PLUS,		2, "%s + %s")
+EXPR_DEF(EXPR_MINUS,		2, "%s - %s")
+EXPR_DEF(EXPR_TIMES,		2, "%s * %s")
+EXPR_DEF(EXPR_DIV,		2, "%s / %s")
+EXPR_DEF(EXPR_MOD,		2, "%s %% %s")
+EXPR_DEF(EXPR_BITNOT,		1, "~%s")
+EXPR_DEF(EXPR_BITAND,		2, "%s & %s")
+EXPR_DEF(EXPR_BITOR,		2, "%s | %s")
+EXPR_DEF(EXPR_BITXOR,		2, "%s ^ %s")
+EXPR_DEF(EXPR_LSHIFT,		2, "%s << %s")
+EXPR_DEF(EXPR_RSHIFT,		2, "%s >> %s")
+EXPR_DEF(EXPR_EQUAL,		2, "%s == %s")
+EXPR_DEF(EXPR_NEQ,		2, "%s != %s")
+EXPR_DEF(EXPR_GE,		2, "%s >= %s")
+EXPR_DEF(EXPR_LE,		2, "%s <= %s")
+EXPR_DEF(EXPR_GT,		2, "%s > %s")
+EXPR_DEF(EXPR_LT,		2, "%s < %s")
+EXPR_DEF(EXPR_NOT,		1, "! %s")
+EXPR_DEF(EXPR_AND,		2, "%s && %s")
+EXPR_DEF(EXPR_OR,		2, "%s || %s")
+EXPR_DEF(EXPR_COND,		3, "%s ? %s : %s")
+EXPR_DEF(EXPR_CASE,		-1, "@custom@")

aux/binpac/src/pac_expr.h

@@ -0,0 +1,139 @@
+#ifndef pac_expr_h
+#define pac_expr_h
+
+#include "pac_common.h"
+#include "pac_datadep.h"
+
+class CaseExpr;
+
+class Expr : public Object, public DataDepElement
+{
+public:
+	enum ExprType {
+#		define EXPR_DEF(type, x, y) type,
+#		include "pac_expr.def"
+#		undef EXPR_DEF
+	};
+
+	void init();
+
+	Expr(ID *id);
+	Expr(Number *num);
+	Expr(ConstString *s);
+	Expr(RegEx *regex);
+	Expr(ExprList *args);	// for EXPR_CALLARGS
+	Expr(Expr *index, CaseExprList *cases);
+
+	Expr(ExprType type, Expr *op1);
+	Expr(ExprType type, Expr *op1, Expr *op2);
+	Expr(ExprType type, Expr *op1, Expr *op2, Expr *op3);
+
+	virtual ~Expr();
+
+	const char *orig() const	{ return orig_.c_str(); }
+	const ID *id() const 		{ return id_; }
+	const char *str() const		{ return str_.c_str(); }
+	ExprType expr_type() const	{ return expr_type_; }
+
+	void AddCaseExpr(CaseExpr *case_expr);
+
+	// Returns the data "type" of the expression. Here we only
+	// do a serious job for the EXPR_MEMBER and EXPR_SUBSCRIPT
+	// operators. For arithmetic operations, we fall back
+	// to "int".
+	Type *DataType(Env *env) const;
+	string DataTypeStr(Env *env) const;
+
+	// Note: EvalExpr() may generate C++ statements in order to evaluate 
+	// variables in the expression, so the following is wrong:
+	//
+	// out->print("int x = ");
+	// out->println("%s", expr->EvalExpr(out, env));
+	//
+	// While putting them together is right:
+	//
+	// out->println("int x = %s", expr->EvalExpr(out, env));
+	//
+	const char *EvalExpr(Output *out, Env *env);
+
+	// force evaulation of IDs contained in this expression;
+	// necessary with case expr and conditional let fields (&if)
+	// for correct parsing of fields
+	void ForceIDEval(Output *out_cc, Env *env);
+
+	// Returns the set_* function of the expression. 
+	// The expression must be of form ID or x.ID.
+	string SetFunc(Output *out, Env *env);
+
+	// Returns true if the expression folds to an integer
+	// constant with env, and puts the constant in *pn.
+	//
+	bool ConstFold(Env *env, int *pn) const;
+
+	// Whether id is referenced in the expression
+	bool HasReference(const ID *id) const;
+
+	// Suppose the data for type might be incomplete, what is
+	// the minimal number of bytes from data head required to
+	// compute the expression? For example, how many bytes of frame
+	// header do we need to determine the length of the frame?
+	//
+	// The parameter <env> points to the Env of a type.
+	// 
+	// Returns -1 if the number is not a constant.
+	//
+	int MinimalHeaderSize(Env *env);
+
+	// Whether evaluation of the expression requires the analyzer context
+	bool RequiresAnalyzerContext() const;
+
+protected:
+	bool DoTraverse(DataDepVisitor *visitor);
+
+private:
+	ExprType expr_type_;
+
+	int num_operands_;
+	Expr *operand_[3];
+
+	ID *id_;		// EXPR_ID
+	Number *num_;		// EXPR_NUM
+	ConstString *cstr_;	// EXPR_CSTR
+	RegEx *regex_;		// EXPR_REGEX
+	ExprList *args_;	// EXPR_CALLARGS
+	CaseExprList *cases_;	// EXPR_CASE
+
+	string str_;		// value string
+	string orig_;		// original string for debugging info
+
+	void GenStrFromFormat(Env *env);
+	void GenEval(Output *out, Env *env);
+	void GenCaseEval(Output *out_cc, Env *env);
+};
+
+string OrigExprList(ExprList *exprlist);
+string EvalExprList(ExprList *exprlist, Output *out, Env *env);
+
+// An entry of the case expression, consisting of one or more constant
+// expressions for the case index and a value expression.
+class CaseExpr : public Object, public DataDepElement
+{
+public:
+	CaseExpr(ExprList *index, Expr *value);
+	virtual ~CaseExpr();
+
+	ExprList *index() const 	{ return index_; }
+	Expr *value() const 		{ return value_; }
+
+	bool HasReference(const ID *id) const;
+	bool RequiresAnalyzerContext() const;
+
+protected:
+	bool DoTraverse(DataDepVisitor *visitor);
+
+private:
+	ExprList *index_;
+	Expr *value_;
+};
+
+#endif  // pac_expr_h