externals/gridflow/base bitpacking.c, NONE, 1.1 flow_objects.c, NONE, 1.1 flow_objects.rb, NONE, 1.1 flow_objects_for_image.c, NONE, 1.1 flow_objects_for_matrix.c, NONE, 1.1 grid.c, NONE, 1.1 grid.h, NONE, 1.1 main.c, NONE, 1.1 main.rb, NONE, 1.1 number.c, NONE, 1.1 source_filter.rb, NONE, 1.1 test.rb, NONE, 1.1 - Pd-cvs

4 Oct 2005


      Update of /cvsroot/pure-data/externals/gridflow/base
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv21117/base
Added Files:
    bitpacking.c flow_objects.c flow_objects.rb 
    flow_objects_for_image.c flow_objects_for_matrix.c grid.c 
    grid.h main.c main.rb number.c source_filter.rb test.rb 
Log Message:
starting to commit gridflow 0.8.0 ...
if you know how to use "cvs import" please mail me and i'll use it for 0.8.1
--- NEW FILE: source_filter.rb ---
$keywords = %w(class decl def end grdecl)
$stack = []
$classes = []
ClassDecl = Struct.new(:name,:supername,:methods,:grins,:attrs,:info)
MethodDecl = Struct.new(:rettype,:selector,:arglist,:minargs,:maxargs,:where)
Arg = Struct.new(:type,:name,:default)
class MethodDecl
    def ==(o)
    	return false unless rettype==o.rettype &&
    	maxargs==o.maxargs # && minargs==o.minargs 
    	arglist.each_index{|i| arglist[i] == o.arglist[i] or return false }
    	return true
    end
end
class Arg
    def ==(o)
    	type==o.type && name==o.name # && default==o.default
    end
end
In = File.open ARGV[0], "r"
Out = File.open ARGV[1], "w"
def handle_class(line)
    raise "already in class #{where}" if $stack[-1] and ClassDecl===$stack[-1]
    #STDERR.puts "class: #{line}"
    /^(\w+)(?:\s*<\s*(\w+))?$/.match line or raise "syntax error #{where}"
    q=ClassDecl.new($1,$2,{},{},{},false)
    $stack << q
    $classes << q
    Out.puts ""
end
def parse_methoddecl(line,term)
    /^(\w+)\s+(\w+)\s*(([^)]*))\s*#{term}/.match line or
    	raise "syntax error #{where} #{line}"
    rettype,selector,arglist = $1,$2,$3
    arglist,minargs,maxargs = parse_arglist arglist
    MethodDecl.new(rettype,selector,arglist,minargs,maxargs,where)
end
def parse_arglist(arglist)
    arglist = arglist.split(/,/)
    maxargs = arglist.length
    args = arglist.map {|arg|
    	if /^\s*...\s*$/.match arg then maxargs=-1; next end
    	/^\s*([\w\s*<>]+)\s*\b(\w+)\s*(?:=(.*))?/.match arg or
    		raise "syntax error in "#{arg}" #{where}"
    	type,name,default=$1,$2,$3
    	Arg.new(type.sub(/\s+$/,""),name,default)
    }.compact
    minargs = args.length
    minargs-=1 while minargs>0 and args[minargs-1].default
    [args,minargs,maxargs]
end
def unparse_arglist(arglist,with_default=true)
    arglist.map {|arg|
    	x="#{arg.type} #{arg.name} "
    	x<<'='<<arg.default if with_default and arg.default
    	x
    }.join(", ")
end
def where
  "[#{ARGV[0]}:#{$linenumber}]"
end
def handle_attr(line)
    type = line.gsub(%r"//.*$","").gsub(%r"/*.**/","").gsub(%r";?\s*$","")
    name = type.slice!(/\w+$/)
    raise "missing \class #{where}" if
    	not $stack[-1] or not ClassDecl===$stack[-1]
    $stack[-1].attrs[name]=Arg.new(type,name,nil)
    Out.print line
    Out.puts "//FCS"
    handle_decl "void _0_#{name}_m (#{type} #{name});"
    Out.puts "# #{$linenumber}"
end
def handle_decl(line)
    raise "missing \class #{where}" if
    	not $stack[-1] or not ClassDecl===$stack[-1]
    classname = $stack[-1].name
    m = parse_methoddecl(line,";\s*$")
    $stack[-1].methods[m.selector] = m
Out.print "#{m.rettype} #{m.selector}(int argc, Ruby *argv"
    Out.print "," if m.arglist.length>0
    Out.print "#{unparse_arglist m.arglist});"
    Out.puts "static Ruby #{m.selector}_wrap"+
    "(int argc, Ruby *argv, Ruby rself);//FCS"
    Out.puts "# #{$linenumber+1}"
end
def handle_def(line)
    m = parse_methoddecl(line,"{?.*$")
    term = line[/{.*/]
    qlass = $stack[-1]
    raise "missing \class #{where}" if not qlass or not ClassDecl===qlass
    classname = qlass.name
    if qlass.methods[m.selector]
    	n = m; m = qlass.methods[m.selector]
    	if m!=n then
    		STDERR.puts "warning: def does not match decl:"
    		STDERR.puts "#{m.where}: \decl #{m.inspect}"
    		STDERR.puts "#{n.where}: \def #{n.inspect}"
    	end
    else
    	qlass.methods[m.selector] = m
    end
Out.print "Ruby #{classname}::#{m.selector}_wrap"+
    "(int argc, Ruby *argv, Ruby rself) {"+
    "static const char *methodspec = "+
    ""#{qlass.name}::#{m.selector}(#{unparse_arglist m.arglist,false})";"+
    "DGS(#{classname});"
Out.print "if (argc<#{m.minargs}"
    Out.print "||argc>#{m.maxargs}" if m.maxargs!=-1
    Out.print ") RAISE("got %d args instead of %d..%d in %s""+
    	",argc,#{m.minargs},#{m.maxargs},methodspec);"
error = proc {|x,y|
    	"RAISE("got %s instead of #{x} in %s","+
    	"rb_str_ptr(rb_inspect(rb_obj_class(#{y}))),methodspec)"
    }
m.arglist.each_with_index{|arg,i|
    	case arg.type
    	when "Symbol"
    		Out.print "if (argc>#{i} && TYPE(argv[#{i}])!=T_SYMBOL) "+
    		error[arg.type,"argv[#{i}]"]+";"
    	when "Array"
    		Out.print "if (argc>#{i} && TYPE(argv[#{i}])!=T_ARRAY) "+
    		error[arg.type,"argv[#{i}]"]+";"
    	when "String"
    		Out.print "if (argc>#{i} && TYPE(argv[#{i}])==T_SYMBOL) "+
    		"argv[#{i}]=rb_funcall(argv[#{i}],SI(to_s),0);"
    		Out.print "if (argc>#{i} && TYPE(argv[#{i}])!=T_STRING) "+
    		error[arg.type,"argv[#{i}]"]+";"
    	end
    }
#	Out.print "return " if m.rettype!="void"
    Out.print "VALUE foo = " if m.rettype!="void" ###
Out.print " self->#{m.selector}(argc,argv"
    m.arglist.each_with_index{|arg,i|
    	if arg.default then
    		Out.print ",argc<#{i+1}?#{arg.default}:convert(argv[#{i}],(#{arg.type}*)0)"
    	else
    		Out.print ",convert(argv[#{i}],(#{arg.type}*)0)"
    	end
    }
    Out.print ");"
    Out.print "self->check_magic();"
    Out.print "return Qnil;" if m.rettype=="void"
    Out.print "return foo;" if m.rettype!="void" ###
    Out.print "} #{m.rettype} #{classname}::#{m.selector}(int argc, Ruby *argv"
    Out.print "," if m.arglist.length>0
    Out.puts "#{unparse_arglist m.arglist, false})#{term}//FCS"
end
def handle_classinfo(line)
    frame = $stack[-1]
    cl = frame.name
    line="{}" if /^\s*$/ =~ line
    Out.puts "static void #{cl}_startup (Ruby rself);"
    Out.puts "static void *#{cl}_allocator () {return new #{cl};}"
    Out.puts "static MethodDecl #{cl}_methods[] = {"
    Out.puts frame.methods.map {|foo,method|
    	c,s = frame.name,method.selector
    	"{ "#{s}",(RMethod)#{c}::#{s}_wrap }"
    }.join(",")
    Out.puts "}; FClass ci#{cl} = { #{cl}_allocator, #{cl}_startup,"
    Out.puts "#{cl.inspect}, COUNT(#{cl}_methods), #{cl}_methods };"
    Out.puts "void #{frame.name}_startup (Ruby rself) "+line
end
def handle_grin(line)
    fields = line.split(/\s+/)
    i = fields[0].to_i
    c = $stack[-1].name
    Out.print "template <class T> void grin_#{i}(GridInlet *in, int n, Pt<T> data);"
    Out.print "template <class T> static void grinw_#{i} (GridInlet *in, int n, Pt<T> data);"
    Out.print "static GridHandler grid_#{i}_hand;"
    handle_decl "Ruby _#{i}_grid(...);"
    $stack[-1].grins[i] = fields.dup
end
def handle_end(line)
    frame = $stack.pop
    fields = line.split(/\s+/)
    n = fields.length
    if ClassDecl===frame then
    	#handle_classinfo if not frame.info
    	cl = frame.name
    	if fields[0]!="class" or
    	(n>1 and fields[1]!=cl)
    	then raise "end not matching #{where}" end
    	$stack.push frame
    	frame.attrs.each {|name,attr|
    		type,name,default = attr.to_a
    		#STDERR.puts "type=#{type} name=#{name} default=#{default}"
    		handle_def "void _0_#{name}_m (#{type} #{name}) { this->#{name}=#{name}; }"
    	}
    	frame.grins.each {|i,v|
    		k = case v[1]
    		when nil; '4'
    		when 'int32'; '1'
    		when 'int'; '2'
    		when 'float'; 'F'
    		else raise 'BORK BORK BORK' end
    		Out.print "static GridHandler #{cl}_grid_#{i}_hand = GRIN#{k}(#{cl},#{i});"
    		handle_def "Ruby _#{i}_grid(...) {"+
    			"if (in.size()<=#{i}) in.resize(#{i}+1);"+
    			"if (!in[#{i}]) in[#{i}]=new GridInlet((GridObject *)this,&#{cl}_grid_#{i}_hand);"+
    			"return in[#{i}]->begin(argc,argv);}"
}
    	$stack.pop
    	Out.puts "# #{$linenumber}"
    end
    if :ruby==frame then
    	if fields[0]!="ruby" then raise "expected \end ruby" end
    end
    Out.puts ""
end
def handle_startall(line)
    $classes.each {|q|
    	Out.print "rb_funcall(EVAL("GridFlow"),SI(fclass_install),2,PTR2FIX(&ci#{q.name}),"
    	if q.supername then
    		Out.print "EVAL("GridFlow::#{q.supername}"));"
    	else
    		Out.print "Qnil);"
    	end
    }
    Out.puts ""
end
def handle_ruby(line)
    Out.puts ""
    $stack.push :ruby
end
$rubymode=false
$linenumber=1
loop{
    x = In.gets
    break if not x
    if /^\s*\(\w+)\s*(.*)$/.match x then
    	begin
    		send("handle_#{$1}",$2)
    	rescue StandardError => e
    		STDERR.puts e.inspect
    		STDERR.puts "at line #{$linenumber}"
    		STDERR.puts e.backtrace
    		File.unlink ARGV[1]
    		exit 1
    	end
    else
    	if $stack[-1]==:ruby then
    		x.gsub!(/([\"])/) { "\"+$1 }
    		x=""#{x.chomp}\n"\n"
    	end
    	Out.puts x
    end
    $linenumber+=1
}
--- NEW FILE: flow_objects_for_matrix.c ---
/*
    $Id: flow_objects_for_matrix.c,v 1.1 2005/10/04 02:02:13 matju Exp $
GridFlow
    Copyright (c) 2001,2002,2003 by Mathieu Bouchard
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.
See file ../COPYING for further informations on licensing terms.
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.
*/
#include <math.h>
#include "grid.h.fcs"
// produce an upper triangular matrix with ones on the diagonal
// will also affect any additional columns using the same row-operations
void expect_complete_matrix (P<Dim> d) {
    if (d->n!=2) RAISE("bletch");
    if (d->get(0)>d->get(1)) RAISE("argh");
}
\class GridMatrixSolve < GridObject
struct GridMatrixSolve : GridObject {
    Numop *op_sub;
    Numop *op_mul;
    Numop *op_div;
    PtrGrid matrix;
    GridMatrixSolve() {
    	matrix.constrain(expect_complete_matrix);
    }
    \decl void initialize ();
    \grin 0 float
};
GRID_INPUT(GridMatrixSolve,0,matrix) {
    int n = matrix->dim->get(0); // # rows
    int m = matrix->dim->get(1); // # columns
    Pt<T> mat = (Pt<T>)*matrix;
    for (int j=0; j<n; j++) {
    	op_div->map(m,mat+j*m,mat[j*m+j]);
    	for (int i=j+1; i<n; i++) {
    		STACK_ARRAY(T,row,m);
    		COPY(row,mat+j,m);
    		op_mul->map(m,row,mat[i*m+j]);
    		op_sub->zip(m,mat+i*m,row);
    	}
    }
    GridOutlet out(this,0,matrix->dim);
    out.send(n*m,mat);
} GRID_END
\def void initialize () {
    rb_call_super(argc,argv);
    this->op_sub = op_sub;
    this->op_mul = op_mul;
    this->op_div = op_div;
}
\classinfo { IEVAL(rself,"install '#matrix_solve',1,1"); }
\end class
void startup_flow_objects_for_matrix () {
    \startall
}
--- NEW FILE: main.rb ---
=begin
    $Id: main.rb,v 1.1 2005/10/04 02:02:13 matju Exp $
GridFlow
    Copyright (c) 2001,2002 by Mathieu Bouchard
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.
See file ../COPYING for further informations on licensing terms.
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.
=end
# ENV["RUBY_VERBOSE_GC"]="yes"
# this file gets loaded by main.c upon startup
# module GridFlow is supposed to be created by main.c
# this includes GridFlow.post_string(s)
# because Ruby1.6 has no #object_id and Ruby1.8 warns on #id
unless Object.instance_methods(true).include? "object_id"
    class Object; alias object_id id end
end
# in case of bug in Ruby ("Error: Success")
module Errno; class E000 < StandardError; end; end
#$post_log = File.open "/tmp/gridflow.log", "w"
$post_log = nil
class Array
    def split(elem)
    	r=[]
    	j=0
    	for i in 0...length
    		(r<<self[j,i-j]; j=i+1) if self[i]==elem
    	end
    	r<<self[j,length-j]
    end
end
module GridFlow #------------------
def self.post(s,*a)
    post_string(sprintf("%s"+s,post_header,*a))
    ($post_log << sprintf(s,*a); $post_log.flush) if $post_log
end
class<<self
    attr_accessor :data_path
    attr_accessor :post_header
    attr_accessor :verbose
    attr_reader :fobjects
    attr_reader :fclasses
    attr_reader :cpu_hertz
    attr_reader :subprocesses
    attr_reader :bridge_name
    alias gfpost post
end
@subprocesses={}
@verbose=false
@data_path=[]
if GridFlow.respond_to? :config then
  @data_path << GridFlow.config["PUREDATA_PATH"]+"/extra/gridflow/images"
end
def self.hunt_zombies
    #STDERR.puts "GridFlow.hunt_zombies"
    # the $$ value is bogus
    begin
    	died = []
    	subprocesses.each {|x,v|
    		Process.waitpid2(x,Process::WNOHANG) and died<<x
    	}
    rescue Errno::ECHILD
    end
    #STDERR.puts died.inspect
    died.each {|x| subprocesses.delete x }
end
def self.packstring_for_nt(nt)
    case nt
    when :u, :u8,  :uint8; "C*"
    when :s, :i16, :int16; "s*"
    when :i, :i32, :int32; "l*"
    when :f, :f32, :float32; "f*"
    when :d, :f64, :float64; "d*"
    else raise "no decoder for #{nt.inspect}"
    end
end
self.post_header = "[gf] "
def self.gfpost2(fmt,s); post("%s",s) end
if GridFlow.bridge_name then
  post "This is GridFlow #{GridFlow::GF_VERSION} within Ruby version #{RUBY_VERSION}"
  post "base/main.c was compiled on #{GridFlow::GF_COMPILE_TIME}"
  post "Please use at least 1.6.6 if you plan to use sockets" if RUBY_VERSION<"1.6.6"
end
if not GridFlow.bridge_name then
  require "gridflow/bridge/placebo"
end
Brace1 = "{".intern
Brace2 = "}".intern
Paren1 = "(".intern
Paren2 = ")".intern
def self.parse(m)
    m = m.gsub(/({|})/," \1 ").split(/\s+/)
    m.map! {|x| case x
    	when Integer, Symbol; x
    	when /^[+-]?[0-9]+$/; x.to_i
    	when String; x.intern
    	end
    }
    m
end
def self.stringify_list(argv)
    argv.map {|x| stringify x }.join(" ")
end
def self.stringify(arg)
    case arg
    when Integer, Float, Symbol; arg.to_s
    when Array; "{#{stringify_list arg}}"
    end
end
::Object.module_eval do def FloatOrSymbol(x) Float(x) rescue x.intern end end
# adding some functionality to that:
class FObject
    @broken_ok = false
    @do_loadbangs = true
    class<<self
    	# global
    	attr_accessor :broken_ok
    	# per-class
    	attr_reader :ninlets
    	attr_reader :noutlets
    	attr_accessor :do_loadbangs
    	attr_accessor :comment
    	def foreign_name; @foreign_name if defined? @foreign_name end
    end
    def post(*a) GridFlow.post(*a) end
    def self.subclass(*args,&b)
    	qlass = Class.new self
    	qlass.install(*args)
    	qlass.module_eval(&b)
    end
    alias :total_time :total_time_get
    alias :total_time= :total_time_set
    attr_writer :args # String
    attr_accessor :argv # Array
    attr_reader :outlets
    attr_accessor :parent_patcher
    attr_accessor :properties
    attr_accessor :classname
    def initialize2; end
    def args
    	if defined? @args
    		@args
    	else
    		"[#{self.class} ...]"
    	end
    end
    alias info args
    def connect outlet, object, inlet
    	@outlets ||= []
    	@outlets[outlet] ||= []
    	@outlets[outlet].push [object, inlet]
    end
    def self.name_lookup sym
    	qlasses = GridFlow.fclasses
    	qlass = qlasses[sym.to_s]
    	if not qlass
    		return qlasses['broken'] if @broken_ok
    		raise "object class '#{sym}' not found"
    	end
    	qlass
    end
    def self.[](*m)
    	o=nil
    	if m.length==1 and m[0] =~ / /
    		o="[#{m[0]}]"
    		m=GridFlow.parse(m[0]) 
    	else
    		o=m.inspect
    	end
    	GridFlow.handle_braces!(m)
    	ms = m.split ','.intern
    	m = ms.shift
    	qlass = m.shift
    	qlassname = qlass.to_s
    	qlass = name_lookup qlass.to_s unless Class===qlass
    	r = qlass.new(*m)
    	r.classname = qlassname
    	GridFlow.post "%s",r.args if GridFlow.verbose
    	for x in ms do r.send_in(-2, *x) end if FObject.do_loadbangs
    	r
    end
    def inspect
    	if args then "#<#{self.class} #{args}>" else super end
    end
    def initialize(*argv)
    	s = GridFlow.stringify_list argv
    	@argv = argv
    	@args = "["
    	@args << (self.class.foreign_name || self.to_s)
    	@args << " " if s.length>0
    	@args << s << "]"
    	@parent_patcher = nil
    	@properties = {}
    	@init_messages = []
    end
end
class FPatcher < FObject
    class << self
    	attr_reader :fobjects
    	attr_reader :wires
    end
    def initialize(*)
    	super
    	fobjects = self.class.fobjects
    	wires = self.class.wires
    	@fobjects = fobjects.map {|x| if String===x then FObject[x] else x.call end }
    	@inlets = []
    	@ninlets = self.class.ninlets or raise "oops"
    	i=0
    	@fobjects << self
    	while i<wires.length do
    		a,b,c,d = wires[i,4]
    		if a==-1 then
    			a=self
    			@inlets[b]||=[]
    			@inlets[b] << [@fobjects[c],d]
    		else
    			if c==-1 then
    				@fobjects[a].connect b,self,d+@ninlets
    			else
    				@fobjects[a].connect b,@fobjects[c],d
    			end
    		end
    		i+=4
    	end
    end
    def method_missing(sym,*args)
    	sym=sym.to_s
    	if sym =~ /^_(\d)_(.*)/ then
    		inl = Integer $1
    		sym = $2.intern
    		if inl<@ninlets then
    		raise "#{inspect} has not @inlets[#{inl}]" if not @inlets[inl]
    			for x in @inlets[inl] do
    			 x[0].send_in x[1],sym,*args end
    		else
    			send_out(inl-@ninlets,sym,*args)
    		end
    	else super end
    end
end
def GridFlow.estimate_cpu_clock
    u0,t0=GridFlow.rdtsc,Time.new.to_f; sleep 0.01
    u1,t1=GridFlow.rdtsc,Time.new.to_f; (u1-u0)/(t1-t0)
end
begin
    @cpu_hertz = (0...3).map {
    	GridFlow.estimate_cpu_clock
    }.sort[1] # median of three tries
rescue
    GridFlow.post $!
end
def GridFlow.find_file s
    s=s.to_s
    if s==File.basename(s) then
    	dir = GridFlow.data_path.find {|x| File.exist? "#{x}/#{s}" }
    	if dir then "#{dir}/#{s}" else s end
    elsif GridFlow.respond_to? :find_file_2
    	GridFlow.find_file_2 s
    else
    	s
    end
end
def GridFlow.macerr(i)
  begin
    f=File.open("/System/Library/Frameworks/CoreServices.framework/"+
      "Versions/A/Frameworks/CarbonCore.framework/Versions/A/Headers/"+
      "MacErrors.h")
    while f.gets
      m = /^\s*(\w+)\s*=\s*(-\d+),\s*/*\s*(.*)\s**/$/.match $_
      next if not m
      if m[2].to_i == i then return "#{m[2]}: "#{m[3]}"" end
    end
    return "no error message available for this error number"
  rescue FileError
    return "Can't find Apple's precious copyrighted list of error messages on this system."
  ensure
    f.close if f	
  end
end
end # module GridFlow
class IO
  def nonblock= flag
    bit = Fcntl::O_NONBLOCK
    state = fcntl(Fcntl::F_GETFL, 0)
    fcntl(Fcntl::F_SETFL, (state & ~bit) |
      (if flag; bit else 0 end))
  end
end
def protect
  yield
rescue Exception => e
  STDERR.puts "#{e.class}: #{e}"
  STDERR.puts e.backtrace
end
def GridFlow.load_user_config
    require "gridflow/bridge/puredata.rb" if GridFlow.bridge_name == "puredata"
    user_config_file = ENV["HOME"] + "/.gridflow_startup"
    begin
    	load user_config_file if File.exist? user_config_file
    rescue Exception => e
    	GridFlow.post "#{e.class}: #{e}:\n" + e.backtrace.join("\n")
    	GridFlow.post "while loading ~/.gridflow_startup"
    end
end
require "gridflow/base/flow_objects.rb"
require "gridflow/format/main.rb"
%w(
  # #for #finished #type #dim #transpose #perspective #store #outer
  #grade #redim #import #export #export_list #cast
  #scale_by #downscale_by #draw_polygon #draw_image #layer
  #print #pack #export_symbol #rotate
  #in #out
).each {|k|
    GridFlow::FObject.name_lookup(k).add_creator k.gsub(/#/,"@")
}
END {
    GridFlow.fobjects.each {|k,v| k.delete if k.respond_to? :delete }
    GridFlow.fobjects.clear
    GC.start
}
--- NEW FILE: flow_objects_for_image.c ---
/*
    $Id: flow_objects_for_image.c,v 1.1 2005/10/04 02:02:13 matju Exp $
GridFlow
    Copyright (c) 2001,2002,2003 by Mathieu Bouchard
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.
See file ../COPYING for further informations on licensing terms.
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.
*/
#include <math.h>
#include "grid.h.fcs"
static void expect_picture (P<Dim> d) {
    if (d->n!=3) RAISE("(height,width,chans) dimensions please");}
static void expect_rgb_picture (P<Dim> d) {
    expect_picture(d);
    if (d->get(2)!=3) RAISE("(red,green,blue) channels please");}
static void expect_rgba_picture (P<Dim> d) {
    expect_picture(d);
    if (d->get(2)!=4) RAISE("(red,green,blue,alpha) channels please");}
static void expect_max_one_dim (P<Dim> d) {
    if (d->n>1) { RAISE("expecting Dim[] or Dim[n], got %s",d->to_s()); }}
//****************************************************************
//{ Dim[A,B,*Cs]<T>,Dim[D,E]<T> -> Dim[A,B,*Cs]<T> }
static void expect_convolution_matrix (P<Dim> d) {
    if (d->n != 2) RAISE("only exactly two dimensions allowed for now (got %d)",
    	d->n);
}
// entry in a compiled convolution kernel
struct PlanEntry { int y,x; bool neutral; };
\class GridConvolve < GridObject
struct GridConvolve : GridObject {
    \attr Numop *op_para;
    \attr Numop *op_fold;
    \attr PtrGrid seed;
    \attr PtrGrid b;
    PtrGrid a;
    int plann;
    PlanEntry *plan; //Pt?
    int margx,margy; // margins
    GridConvolve () : plan(0) { b.constrain(expect_convolution_matrix); plan=0; }
    \decl void initialize (Grid *r=0);
    \decl void _0_op   (Numop *op);
    \decl void _0_fold (Numop *op);
    \decl void _0_seed (Grid *seed);
    \grin 0
    \grin 1
    template <class T> void copy_row (Pt<T> buf, int sx, int y, int x);
    template <class T> void make_plan (T bogus);
    ~GridConvolve () {if (plan) delete[] plan;}
};
template <class T> void GridConvolve::copy_row (Pt<T> buf, int sx, int y, int x) {
    int day = a->dim->get(0), dax = a->dim->get(1), dac = a->dim->prod(2);
    y=mod(y,day); x=mod(x,dax);
    Pt<T> ap = (Pt<T>)*a + y*dax*dac;
    while (sx) {
    	int sx1 = min(sx,dax-x);
    	COPY(buf,ap+x*dac,sx1*dac);
    	x=0;
    	buf += sx1*dac;
    	sx -= sx1;
    }
}
static Numop *OP(Ruby x) {return FIX2PTR(Numop,rb_hash_aref(op_dict,x));}
template <class T> void GridConvolve::make_plan (T bogus) {
    P<Dim> da = a->dim, db = b->dim;
    int dby = db->get(0);
    int dbx = db->get(1);
    if (plan) delete[] plan;
    plan = new PlanEntry[dbx*dby];
    int i=0;
    for (int y=0; y<dby; y++) {
    	for (int x=0; x<dbx; x++) {
    		T rh = ((Pt<T>)*b)[y*dbx+x];
    		bool neutral = op_para->on(rh)->is_neutral(rh,at_right);
    		bool absorbent = op_para->on(rh)->is_absorbent(rh,at_right);
    		STACK_ARRAY(T,foo,1);
    		if (absorbent) {
    			foo[0] = 0;
    			op_para->map(1,foo,rh);
    			absorbent = op_fold->on(rh)->is_neutral(foo[0],at_right);
    		}
    		if (absorbent) continue;
    		plan[i].y = y;
    		plan[i].x = x;
    		plan[i].neutral = neutral;
    		i++;
    	}
    }
    plann = i;
}
GRID_INLET(GridConvolve,0) {
    SAME_TYPE(in,b);
    SAME_TYPE(in,seed);
    P<Dim> da = in->dim, db = b->dim;
    if (!db) RAISE("right inlet has no grid");
    if (!seed) RAISE("seed missing");
    if (db->n != 2) RAISE("right grid must have two dimensions");
    if (da->n < 2) RAISE("left grid has less than two dimensions");
    if (seed->dim->n != 0) RAISE("seed must be scalar");
    if (da->get(0) < db->get(0)) RAISE("grid too small (y): %d < %d", da->get(0), db->get(0));
    if (da->get(1) < db->get(1)) RAISE("grid too small (x): %d < %d", da->get(1), db->get(1));
    margy = (db->get(0)-1)/2;
    margx = (db->get(1)-1)/2;
    a=new Grid(in->dim,in->nt);
    out=new GridOutlet(this,0,da,in->nt);
} GRID_FLOW {
    COPY((Pt<T>)*a+in->dex, data, n);
} GRID_FINISH {
    Numop *op_put = OP(SYM(put));
    make_plan((T)0);
    int dbx = b->dim->get(1);
    int day = a->dim->get(0);
    int n = a->dim->prod(1);
    int sx = a->dim->get(1)+dbx-1;
    int n2 = sx*a->dim->prod(2);
    STACK_ARRAY(T,buf,n);
    STACK_ARRAY(T,buf2,n2);
    T orh=0;
    for (int iy=0; iy<day; iy++) {
    	op_put->map(n,buf,*(T *)*seed);
    	for (int i=0; i<plann; i++) {
    		int jy = plan[i].y;
    		int jx = plan[i].x;
    		T rh = ((Pt<T>)*b)[jy*dbx+jx];
    		if (i==0 || plan[i].y!=plan[i-1].y || orh!=rh) {
    			copy_row(buf2,sx,iy+jy-margy,-margx);
    			if (!plan[i].neutral) op_para->map(n2,buf2,rh);
    		}
    		op_fold->zip(n,buf,buf2+jx*a->dim->prod(2));
    		orh=rh;
    	}
    	out->send(n,buf);
    }
    a=0;
} GRID_END
GRID_INPUT(GridConvolve,1,b) {} GRID_END
\def void _0_op   (Numop *op ) { this->op_para=op; }
\def void _0_fold (Numop *op ) { this->op_fold=op; }
\def void _0_seed (Grid *seed) { this->seed=seed; }
\def void initialize (Grid *r) {
    rb_call_super(argc,argv);
    this->op_para = op_mul;
    this->op_fold = op_add;
    this->seed = new Grid(new Dim(),int32_e,true);
    this->b= r ? r : new Grid(new Dim(1,1),int32_e,true);
}
\classinfo { IEVAL(rself,"install '#convolve',2,1"); }
\end class GridConvolve
/* ---------------------------------------------------------------- */
/* "#scale_by" does quick scaling of pictures by integer factors */
/*{ Dim[A,B,3]<T> -> Dim[C,D,3]<T> }*/
\class GridScaleBy < GridObject
struct GridScaleBy : GridObject {
    \attr PtrGrid scale; // integer scale factor
    int scaley;
    int scalex;
    \decl void initialize (Grid *factor=0);
    \grin 0
    \grin 1
    void prepare_scale_factor () {
    	scaley = ((Pt<int32>)*scale)[0];
    	scalex = ((Pt<int32>)*scale)[scale->dim->prod()==1 ? 0 : 1];
    	if (scaley<1) scaley=2;
    	if (scalex<1) scalex=2;
    }
};
GRID_INLET(GridScaleBy,0) {
    P<Dim> a = in->dim;
    expect_picture(a);
    out=new GridOutlet(this,0,new Dim(a->get(0)*scaley,a->get(1)*scalex,a->get(2)),in->nt);
    in->set_factor(a->get(1)*a->get(2));
} GRID_FLOW {
    int rowsize = in->dim->prod(1);
    STACK_ARRAY(T,buf,rowsize*scalex);
    int chans = in->dim->get(2);
    #define Z(z) buf[p+z]=data[i+z]
    for (; n>0; data+=rowsize, n-=rowsize) {
    	int p=0;
    	#define LOOP(z) \
    		for (int i=0; i<rowsize; i+=z) \
    		for (int k=0; k<scalex; k++, p+=3)
    	switch (chans) {
    	case 3: LOOP(3) {Z(0);Z(1);Z(2);} break;
    	case 4: LOOP(4) {Z(0);Z(1);Z(2);Z(3);} break;
    	default: LOOP(chans) {for (int c=0; c<chans; c++) Z(c);}
    	}
    	#undef LOOP
    	for (int j=0; j<scaley; j++) out->send(rowsize*scalex,buf);
    }
    #undef Z
} GRID_END
static void expect_scale_factor (P<Dim> dim) {
    if (dim->prod()!=1 && dim->prod()!=2)
    	RAISE("expecting only one or two numbers");
}
GRID_INPUT(GridScaleBy,1,scale) { prepare_scale_factor(); } GRID_END
\def void initialize (Grid *factor) {
    scale.constrain(expect_scale_factor);
    rb_call_super(argc,argv);
    scale=new Grid(INT2NUM(2));
    if (factor) scale=factor;
    prepare_scale_factor();
}
\classinfo { IEVAL(rself,"install '#scale_by',2,1"); }
\end class GridScaleBy
// ----------------------------------------------------------------
//{ Dim[A,B,3]<T> -> Dim[C,D,3]<T> }
\class GridDownscaleBy < GridObject
struct GridDownscaleBy : GridObject {
    \attr PtrGrid scale;
    \attr bool smoothly;
    int scaley;
    int scalex;
    PtrGrid temp;
    \decl void initialize (Grid *factor=0, Symbol option=Qnil);
    \grin 0
    \grin 1
    void prepare_scale_factor () {
    	scaley = ((Pt<int32>)*scale)[0];
    	scalex = ((Pt<int32>)*scale)[scale->dim->prod()==1 ? 0 : 1];
    	if (scaley<1) scaley=2;
    	if (scalex<1) scalex=2;
    }
};
GRID_INLET(GridDownscaleBy,0) {
P<Dim> a = in->dim;
    if (a->n!=3) RAISE("(height,width,chans) please");
    out=new GridOutlet(this,0,new Dim(a->get(0)/scaley,a->get(1)/scalex,a->get(2)),in->nt);
    in->set_factor(a->get(1)*a->get(2));
    // i don't remember why two rows instead of just one.
    temp=new Grid(new Dim(2,in->dim->get(1)/scalex,in->dim->get(2)),in->nt);
} GRID_FLOW {
    int rowsize = in->dim->prod(1);
    int rowsize2 = temp->dim->prod(1);
    Pt<T> buf = (Pt<T>)*temp; //!@#$ maybe should be something else than T ?
    int xinc = in->dim->get(2)*scalex;
    int y = in->dex / rowsize;
    int chans=in->dim->get(2);
    #define Z(z) buf[p+z]+=data[i+z]
    if (smoothly) {
    	while (n>0) {
    		if (y%scaley==0) CLEAR(buf,rowsize2);
    		#define LOOP(z) \
    			for (int i=0,p=0; p<rowsize2; p+=z) \
    			for (int j=0; j<scalex; j++,i+=z)
    		switch (chans) {
    		case 1: LOOP(1) {Z(0);} break;
    		case 2: LOOP(2) {Z(0);Z(1);} break;
    		case 3: LOOP(3) {Z(0);Z(1);Z(2);} break;
    		case 4: LOOP(4) {Z(0);Z(1);Z(2);Z(3);} break;
    		default:LOOP(chans) {for (int k=0; k<chans; k++) Z(k);} break;
    		}
    		#undef LOOP
    		y++;
    		if (y%scaley==0 && out->dim) {
    			op_div->map(rowsize2,buf,(T)(scalex*scaley));
    			out->send(rowsize2,buf);
    			CLEAR(buf,rowsize2);
    		}
    		data+=rowsize;
    		n-=rowsize;
    	}
    #undef Z
    } else {
    #define Z(z) buf[p+z]=data[i+z]
    	for (; n>0 && out->dim; data+=rowsize, n-=rowsize,y++) {
    		if (y%scaley!=0) continue;
    		#define LOOP(z) for (int i=0,p=0; p<rowsize2; i+=xinc, p+=z)
    		switch(in->dim->get(2)) {
    		case 1: LOOP(1) {Z(0);} break;
    		case 2: LOOP(2) {Z(0);Z(1);} break;
    		case 3: LOOP(3) {Z(0);Z(1);Z(2);} break;
    		case 4: LOOP(4) {Z(0);Z(1);Z(2);Z(3);} break;
    		default:LOOP(chans) {for (int k=0; k<chans; k++) Z(k);}break;
    		}
    		#undef LOOP
    		out->send(rowsize2,buf);
    	}
    }
    #undef Z
} GRID_END
GRID_INPUT(GridDownscaleBy,1,scale) { prepare_scale_factor(); } GRID_END
\def void initialize (Grid *factor, Symbol option) {
    scale.constrain(expect_scale_factor);
    rb_call_super(argc,argv);
    scale=new Grid(INT2NUM(2));
    if (factor) scale=factor;
    prepare_scale_factor();
    smoothly = option==SYM(smoothly);
}
\classinfo { IEVAL(rself,"install '#downscale_by',2,1"); }
\end class GridDownscaleBy
//****************************************************************
\class GridLayer < GridObject
struct GridLayer : GridObject {
    PtrGrid r;
    GridLayer() { r.constrain(expect_rgb_picture); }
    \grin 0 int
    \grin 1 int
};
GRID_INLET(GridLayer,0) {
    NOTEMPTY(r);
    SAME_TYPE(in,r);
    P<Dim> a = in->dim;
    expect_rgba_picture(a);
    if (a->get(1)!=r->dim->get(1)) RAISE("same width please");
    if (a->get(0)!=r->dim->get(0)) RAISE("same height please");
    in->set_factor(a->prod(2));
    out=new GridOutlet(this,0,r->dim);
} GRID_FLOW {
    Pt<T> rr = ((Pt<T>)*r) + in->dex*3/4;
    STACK_ARRAY(T,foo,n*3/4);
#define COMPUTE_ALPHA(c,a) \
    foo[j+c] = (data[i+c]*data[i+a] + rr[j+c]*(256-data[i+a])) >> 8
    for (int i=0,j=0; i<n; i+=4,j+=3) {
    	COMPUTE_ALPHA(0,3);
    	COMPUTE_ALPHA(1,3);
    	COMPUTE_ALPHA(2,3);
    }
#undef COMPUTE_ALPHA
    out->send(n*3/4,foo);
} GRID_END
GRID_INPUT(GridLayer,1,r) {} GRID_END
\classinfo { IEVAL(rself,"install '#layer',2,1"); }
\end class GridLayer
// ****************************************************************
// pad1,pad2 only are there for 32-byte alignment
struct Line { int32 y1,x1,y2,x2,x,m,pad1,pad2; };
static void expect_polygon (P<Dim> d) {
    if (d->n!=2 || d->get(1)!=2) RAISE("expecting Dim[n,2] polygon");
}
\class DrawPolygon < GridObject
struct DrawPolygon : GridObject {
    \attr Numop *op;
    \attr PtrGrid color;
    \attr PtrGrid polygon;
    PtrGrid color2;
    PtrGrid lines;
    int lines_start;
    int lines_stop;
    DrawPolygon() {
    	color.constrain(expect_max_one_dim);
    	polygon.constrain(expect_polygon);
    }
    \decl void initialize (Numop *op, Grid *color=0, Grid *polygon=0);
    \grin 0
    \grin 1
    \grin 2 int32
    void init_lines();
};
void DrawPolygon::init_lines () {
    int nl = polygon->dim->get(0);
    lines=new Grid(new Dim(nl,8), int32_e);
    Pt<Line> ld = Pt<Line>((Line *)(int32 *)*lines,nl);
    Pt<int32> pd = *polygon;
    for (int i=0,j=0; i<nl; i++) {
    	ld[i].y1 = pd[j+0];
    	ld[i].x1 = pd[j+1];
    	j=(j+2)%(2*nl);
    	ld[i].y2 = pd[j+0];
    	ld[i].x2 = pd[j+1];
    	if (ld[i].y1>ld[i].y2) memswap(Pt<int32>(ld+i)+0,Pt<int32>(ld+i)+2,2);
    }
}
static int order_by_starting_scanline (const void *a, const void *b) {
    return ((Line *)a)->y1 - ((Line *)b)->y1;
}
static int order_by_column (const void *a, const void *b) {
    return ((Line *)a)->x - ((Line *)b)->x;
}
GRID_INLET(DrawPolygon,0) {
    NOTEMPTY(color);
    NOTEMPTY(polygon);
    NOTEMPTY(lines);
    SAME_TYPE(in,color);
    if (in->dim->n!=3) RAISE("expecting 3 dimensions");
    if (in->dim->get(2)!=color->dim->get(0))
    	RAISE("image does not have same number of channels as stored color");
    out=new GridOutlet(this,0,in->dim,in->nt);
    lines_start = lines_stop = 0;
    in->set_factor(in->dim->get(1)*in->dim->get(2));
    int nl = polygon->dim->get(0);
    qsort((int32 *)*lines,nl,sizeof(Line),order_by_starting_scanline);
    int cn = color->dim->prod();
    color2=new Grid(new Dim(cn*16), color->nt);
    for (int i=0; i<16; i++) COPY((Pt<T>)*color2+cn*i,(Pt<T>)*color,cn);
} GRID_FLOW {
    int nl = polygon->dim->get(0);
    Pt<Line> ld = Pt<Line>((Line *)(int32 *)*lines,nl);
    int f = in->factor();
    int y = in->dex/f;
    int cn = color->dim->prod();
    Pt<T> cd = (Pt<T>)*color2;
    
    while (n) {
    	while (lines_stop != nl && ld[lines_stop].y1<=y) lines_stop++;
    	for (int i=lines_start; i<lines_stop; i++) {
    		if (ld[i].y2<=y) {
    			memswap(ld+i,ld+lines_start,1);
    			lines_start++;
    		}
    	}
    	if (lines_start == lines_stop) {
    		out->send(f,data);
    	} else {
    		int32 xl = in->dim->get(1);
    		Pt<T> data2 = ARRAY_NEW(T,f);
    		COPY(data2,data,f);
    		for (int i=lines_start; i<lines_stop; i++) {
    			Line &l = ld[i];
    			l.x = l.x1 + (y-l.y1)*(l.x2-l.x1+1)/(l.y2-l.y1+1);
    		}
    		qsort(ld+lines_start,lines_stop-lines_start,
    			sizeof(Line),order_by_column);
    		for (int i=lines_start; i<lines_stop-1; i+=2) {
    			int xs = max(ld[i].x,(int32)0), xe = min(ld[i+1].x,xl);
    			if (xs>=xe) continue; /* !@#$ WHAT? */
    			while (xe-xs>=16) { op->zip(16*cn,data2+cn*xs,cd); xs+=16; }
    			op->zip((xe-xs)*cn,data2+cn*xs,cd);
    		}
    		out->give(f,data2);
    	}
    	n-=f;
    	data+=f;
    	y++;
    }
} GRID_END
GRID_INPUT(DrawPolygon,1,color) {} GRID_END
GRID_INPUT(DrawPolygon,2,polygon) {init_lines();} GRID_END
\def void initialize (Numop *op, Grid *color, Grid *polygon) {
    rb_call_super(argc,argv);
    this->op = op;
    if (color) this->color=color;
    if (polygon) { this->polygon=polygon; init_lines(); }
}
\classinfo { IEVAL(rself,"install '#draw_polygon',3,1"); }
\end class DrawPolygon
//****************************************************************
static void expect_position(P<Dim> d) {
    if (d->n!=1) RAISE("position should have 1 dimension, not %d", d->n);
    if (d->v[0]!=2) RAISE("position dim 0 should have 2 elements, not %d", d->v[0]);
}
\class DrawImage < GridObject
struct DrawImage : GridObject {
    \attr Numop *op;
    \attr PtrGrid image;
    \attr PtrGrid position;
    \attr bool alpha;
    \attr bool tile;
    
    DrawImage() : alpha(false), tile(false) {
    	position.constrain(expect_position);
    	image.constrain(expect_picture);
    }
\decl void initialize (Numop *op, Grid *image=0, Grid *position=0);
    \decl void _0_alpha (bool v=true);
    \decl void _0_tile (bool v=true);
    \grin 0
    \grin 1
    \grin 2 int32
    // draw row # ry of right image in row buffer buf, starting at xs
    // overflow on both sides has to be handled automatically by this method
    template <class T> void draw_segment(Pt<T> obuf, Pt<T> ibuf, int ry, int x0);
};
#define COMPUTE_ALPHA(c,a) obuf[j+(c)] = ibuf[j+(c)] + (rbuf[a])*(obuf[j+(c)]-ibuf[j+(c)])/256;
#define COMPUTE_ALPHA4(b) \
    COMPUTE_ALPHA(b+0,b+3); \
    COMPUTE_ALPHA(b+1,b+3); \
    COMPUTE_ALPHA(b+2,b+3); \
    obuf[b+3] = rbuf[b+3] + (255-rbuf[b+3])*(ibuf[j+b+3])/256;
template <class T> void DrawImage::draw_segment(Pt<T> obuf, Pt<T> ibuf, int ry, int x0) {
    if (ry<0 || ry>=image->dim->get(0)) return; // outside of image
    int sx = in[0]->dim->get(1), rsx = image->dim->get(1);
    int sc = in[0]->dim->get(2), rsc = image->dim->get(2);
    Pt<T> rbuf = (Pt<T>)*image + ry*rsx*rsc;
    if (x0>sx || x0<=-rsx) return; // outside of buffer
    int n=rsx;
    if (x0+n>sx) n=sx-x0;
    if (x0<0) { rbuf-=rsc*x0; n+=x0; x0=0; }
    if (alpha && rsc==4 && sc==3) { // RGB by RGBA //!@#$ optimise
    	int j=sc*x0;
    	for (; n; n--, rbuf+=4, j+=3) {
    		op->zip(sc,obuf+j,rbuf); COMPUTE_ALPHA(0,3); COMPUTE_ALPHA(1,3); COMPUTE_ALPHA(2,3);
    	}
    } else if (alpha && rsc==4 && sc==4) { // RGBA by RGBA
    	op->zip(n*rsc,obuf+x0*rsc,rbuf);
    	int j=sc*x0;
    	for (; n>=4; n-=4, rbuf+=16, j+=16) {
    		COMPUTE_ALPHA4(0);COMPUTE_ALPHA4(4);
    		COMPUTE_ALPHA4(8);COMPUTE_ALPHA4(12);
    	}
    	for (; n; n--, rbuf+=4, j+=4) {
    		COMPUTE_ALPHA4(0);
    	}
    } else { // RGB by RGB, etc
    	op->zip(n*rsc,obuf+sc*x0,rbuf);
    }
}
GRID_INLET(DrawImage,0) {
    NOTEMPTY(image);
    NOTEMPTY(position);
    SAME_TYPE(in,image);
    if (in->dim->n!=3) RAISE("expecting 3 dimensions");
    int lchan = in->dim->get(2);
    int rchan = image->dim->get(2);
    if (alpha && rchan!=4) {
    	RAISE("alpha mode works only with 4 channels in right_hand");
    }
    if (lchan != rchan-(alpha?1:0) && lchan != rchan) {
    	RAISE("right_hand has %d channels, alpha=%d, left_hand has %d, expecting %d or %d",
    		rchan, alpha?1:0, lchan, rchan-(alpha?1:0), rchan);
    }
    out=new GridOutlet(this,0,in->dim,in->nt);
    in->set_factor(in->dim->get(1)*in->dim->get(2));
} GRID_FLOW {
    int f = in->factor();
    int y = in->dex/f;
    if (position->nt != int32_e) RAISE("position has to be int32");
    int py = ((int32*)*position)[0], rsy = image->dim->v[0], sy=in->dim->get(0);
    int px = ((int32*)*position)[1], rsx = image->dim->v[1], sx=in->dim->get(1);
    for (; n; y++, n-=f, data+=f) {
    	int ty = div2(y-py,rsy);
    	if (tile || ty==0) {
    		Pt<T> data2 = ARRAY_NEW(T,f);
    		COPY(data2,data,f);
    		if (tile) {
    			for (int x=px-div2(px+rsx-1,rsx)*rsx; x<sx; x+=rsx) {
    				draw_segment(data2,data,mod(y-py,rsy),x);
    			}
    		} else {
    			draw_segment(data2,data,y-py,px);
    		}
    		out->give(f,data2);
    	} else {
    		out->send(f,data);
    	}
    }
} GRID_END
GRID_INPUT(DrawImage,1,image) {} GRID_END
GRID_INPUT(DrawImage,2,position) {} GRID_END
\def void _0_alpha (bool v=true) { alpha = v; gfpost("ALPHA=%d",v); }
\def void _0_tile (bool v=true) {   tile = v; }
\def void initialize (Numop *op, Grid *image, Grid *position) {
    rb_call_super(argc,argv);
    this->op = op;
    if (image) this->image=image;
    if (position) this->position=position;
    else this->position=new Grid(new Dim(2),int32_e,true);
}
\classinfo { IEVAL(rself,"install '#draw_image',3,1"); }
\end class DrawImage
void startup_flow_objects_for_image () {
    \startall
}
--- NEW FILE: number.c ---
/*
    $Id: number.c,v 1.1 2005/10/04 02:02:13 matju Exp $
GridFlow
    Copyright (c) 2001,2002,2003,2004 by Mathieu Bouchard
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.
See file ../COPYING for further informations on licensing terms.
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.
*/
#include "grid.h.fcs"
#include <math.h>
#include <stdlib.h>
#include <stdio.h>
#include <limits.h>
#ifdef PASS1
NumberType number_type_table[] = {
#define FOO(_sym_,_size_,_flags_,args...) NumberType( #_sym_, _size_, _flags_, args ),
NUMBER_TYPES(FOO)
#undef FOO
};
const long number_type_table_n = COUNT(number_type_table);
#endif
// those are bogus class-templates in the sense that you don't create
// objects from those, you just call static functions. The same kind
// of pattern is present in STL to overcome some limitations of C++.
template <class T> class Op {
public:
    // I call abort() on those because I can't say they're purevirtual.
    static T f(T a, T b) {abort();}
    static bool is_neutral(T x, LeftRight side)   {assert(!"Op::is_neutral called?");}
    static bool is_absorbent(T x, LeftRight side) {assert(!"Op::is_absorbent called?");}
};
template <class O> class OpLoops {
public:
    template <class T> static void op_map (int n, T *as, T b) {
    	if (!n) return;
#define FOO(I) as[I]=O::f(as[I],b);
    UNROLL_8(FOO,n,as)
#undef FOO
    }
    template <class T> static void op_zip (int n, T *as, T *bs) {
    	if (!n) return;
    	int ba=bs-as; // really!
#define FOO(I) as[I]=O::f(as[I],as[ba+I]);
    UNROLL_8(FOO,n,as)
#undef FOO
    }
    // disabled
    template <class T> static void op_zip2 (int n, T *as, T *bs, T *cs) {
    	if (!n) return;
    	int ba=bs-as, ca=cs-as;
#define FOO(I) as[ca+I]=O::f(as[I],as[ba+I]);
    UNROLL_8(FOO,n,as)
#undef FOO
    }
#define W(i) as[i]=O::f(as[i],bs[i]);
#define Z(i,j) as[i]=O::f(O::f(O::f(O::f(as[i],bs[i]),bs[i+j]),bs[i+j+j]),bs[i+j+j+j]);
    template <class T> static void op_fold (int an, int n, T *as, T *bs) {
    	switch (an) {
    	case 1: for (; (n&3)!=0; bs++, n--) W(0);
    		for (; n; bs+=4, n-=4) { Z(0,1); } break;
    	case 2: for (; (n&3)!=0; bs+=2, n--) { W(0); W(1); }
    		for (; n; bs+=8, n-=4) { Z(0,2); Z(1,2); } break;
    	case 3: for (; (n&3)!=0; bs+=3, n--) { W(0); W(1); W(2); }
    		for (; n; bs+=12, n-=4) { Z(0,3); Z(1,3); Z(2,3); } break;
    	case 4: for (; (n&3)!=0; bs+=4, n--) { W(0); W(1); W(2); W(3); }
    		for (; n; bs+=16, n-=4) { Z(0,4); Z(1,4); Z(2,4); Z(3,4); } break;
    	default:for (; n--; ) {
    			int i=0;
    			for (; i<(an&-4); i+=4, bs+=4) {
    				as[i+0]=O::f(as[i+0],bs[0]);
    				as[i+1]=O::f(as[i+1],bs[1]);
    				as[i+2]=O::f(as[i+2],bs[2]);
    				as[i+3]=O::f(as[i+3],bs[3]);
    			}
    			for (; i<an; i++, bs++) as[i] = O::f(as[i],*bs);
    		}
    	}
    }
    template <class T> static void op_scan (int an, int n, T *as, T *bs) {
    	for (; n--; as=bs-an) {
    		for (int i=0; i<an; i++, as++, bs++) *bs=O::f(*as,*bs);
    	}
    }
};
template <class T>
static void quick_mod_map (int n, T *as, T b) {
    if (!b) return;
#define FOO(I) as[I]=mod(as[I],b);
    UNROLL_8(FOO,n,as)
#undef FOO
}
template <class T> static void quick_ign_map (int n, T *as, T b) {}
template <class T> static void quick_ign_zip (int n, T *as, T *bs) {}
template <class T> static void quick_put_map (int n, T *as, T b) {
#define FOO(I) as[I]=b;
    UNROLL_8(FOO,n,as)
#undef FOO
}
#ifdef PASS1
void quick_put_map (int n, int16 *as, int16 b) {
    if (n&1!=0 && (long)as&4!=0) { *as++=b; n--; }
    quick_put_map (n>>1, (int32 *)as, (int32)(b<<16)+b);
    if (n&1!=0) *as++=b;
}
void quick_put_map (int n, uint8 *as, uint8 b) {
    while (n&3!=0 && (long)as&4!=0) { *as++=b; n--; }
    int32 c=(b<<8)+b; c+=c<<16;
    quick_put_map (n>>2, (int32 *)as, c);
    while (n&3!=0) *as++=b;
}
#endif
template <class T> static void quick_put_zip (int n, T *as, T *bs) {
    gfmemcopy((uint8 *)as, (uint8 *)bs, n*sizeof(T));
}
// classic two-input operator
#define DEF_OP(op,expr,neutral,absorbent) \
    template <class T> class Y##op : Op<T> { public: \
    	inline static T f(T a, T b) { return expr; } \
    	inline static bool is_neutral  (T x, LeftRight side) { return neutral; } \
    	inline static bool is_absorbent(T x, LeftRight side) { return absorbent; } };
#define DEF_OPFT(op,expr,neutral,absorbent,T) \
    template <> class Y##op<T> : Op<T> { public: \
    	inline static T f(T a, T b) { return expr; } \
    	inline static bool is_neutral  (T x, LeftRight side) { return neutral; } \
    	inline static bool is_absorbent(T x, LeftRight side) { return absorbent; } }; \
// this macro is for operators that have different code for the float version
#define DEF_OPF(op,expr,expr2,neutral,absorbent) \
    DEF_OP( op,expr,      neutral,absorbent) \
    DEF_OPFT(op,expr2,neutral,absorbent,float32) \
    DEF_OPFT(op,expr2,neutral,absorbent,float64)
#define DECL_OPON(base,op,T) NumopOn<T>( \
    &base<Y##op<T> >::op_map, &base<Y##op<T> >::op_zip, \
    &base<Y##op<T> >::op_fold, &base<Y##op<T> >::op_scan, \
    &Y##op<T>::is_neutral, &Y##op<T>::is_absorbent)
#define DECL_OPON_NOFOLD(base,op,T) NumopOn<T>( \
    &base<Y##op<T> >::op_map, &base<Y##op<T> >::op_zip, 0,0, \
    &Y##op<T>::is_neutral, &Y##op<T>::is_absorbent)
#define DECL_OP(op,sym,flags) Numop(0, sym, \
    DECL_OPON(OpLoops,op,uint8), DECL_OPON(OpLoops,op,int16), \
    DECL_OPON(OpLoops,op,int32) NONLITE(, DECL_OPON(OpLoops,op,int64), \
    DECL_OPON(OpLoops,op,float32), DECL_OPON(OpLoops,op,float64)), flags)
#define DECL_OP_NOFLOAT(op,sym,flags) Numop(0, sym, \
    DECL_OPON(OpLoops,op,uint8), DECL_OPON(OpLoops,op,int16), \
    DECL_OPON(OpLoops,op,int32) NONLITE(, DECL_OPON(OpLoops,op,int64), \
    NumopOn<float32>(0,0,0,0,0,0), NumopOn<float64>(0,0,0,0,0,0)), flags)
#define DECL_OP_NOFOLD(op,sym,flags) Numop(0, sym, \
    DECL_OPON_NOFOLD(OpLoops,op,uint8), DECL_OPON_NOFOLD(OpLoops,op,int16), \
    DECL_OPON_NOFOLD(OpLoops,op,int32) NONLITE(, DECL_OPON_NOFOLD(OpLoops,op,int64), \
    DECL_OPON_NOFOLD(OpLoops,op,float32), DECL_OPON_NOFOLD(OpLoops,op,float64)), flags)
template <class T> static inline T gf_floor (T a) {
    return (T) floor((double)a); }
template <class T> static inline T gf_trunc (T a) {
    return (T) floor(abs((double)a)) * (a<0?-1:1); }
/*
uint8 clipadd(uint8 a, uint8 b) { int32 c=a+b; return c<0?0:c>255?255:c; }
int16 clipadd(int16 a, int16 b) { int32 c=a+b; return c<-0x8000?-0x8000:c>0x7fff?0x7fff:c; }
int32 clipadd(int32 a, int32 b) { int64 c=a+b; return c<-0x80000000?-0x80000000:c>0x7fffffff?0x7fffffff:c; }
int64 clipadd(int64 a, int64 b) { int64 c=(a>>1)+(b>>1)+(a&b&1);
    return c<(nt_smallest(0LL)/2?nt_smallest(0LL):c>nt_greatest(0LL)/2?nt_greatest(0LL):a+b; }
uint8 clipsub(uint8 a, uint8 b) { int32 c=a-b; return c<0?0:c>255?255:c; }
int16 clipsub(int16 a, int16 b) { int32 c=a-b; return c<-0x8000?-0x8000:c>0x7fff?0x7fff:c; }
int32 clipsub(int32 a, int32 b) { int64 c=a-b; return c<-0x80000000?-0x80000000:c>0x7fffffff?0x7fffffff:c; }
int64 clipsub(int64 a, int64 b) { int64 c=(a>>1)-(b>>1); //???
    return c<(nt_smallest(0LL)/2?nt_smallest(0LL):c>nt_greatest(0LL)/2?nt_greatest(0LL):a-b; }
*/
#ifdef PASS1
DEF_OP(ignore, a, side==at_right, side==at_left)
DEF_OP(put, b, side==at_left, side==at_right)
DEF_OP(add, a+b, x==0, false)
DEF_OP(sub, a-b, side==at_right && x==0, false)
DEF_OP(bus, b-a, side==at_left  && x==0, false)
DEF_OP(mul, a*b, x==1, x==0)
DEF_OP(mulshr8, ((int32)a*(int32)b)>>8, (int64)x==256, x==0) //!@#$ bug with int64
DEF_OP(div, b==0 ? 0 : a/b, side==at_right && x==1, false)
DEF_OP(div2, b==0 ? 0 : div2(a,b), side==at_right && x==1, false)
DEF_OP(vid, a==0 ? 0 : b/a, side==at_left && x==1, false)
DEF_OP(vid2, a==0 ? 0 : div2(b,a), side==at_left && x==1, false)
DEF_OPF(mod, b==0 ? 0 : mod(a,b), b==0 ? 0 : a-b*gf_floor(a/b),
    false, side==at_left && x==0 || side==at_right && x==1)
DEF_OPF(dom, a==0 ? 0 : mod(b,a), a==0 ? 0 : b-a*gf_floor(b/a),
    false, side==at_left && x==0 || side==at_right && x==1)
//DEF_OPF(rem, b==0 ? 0 : a%b, b==0 ? 0 : a-b*gf_trunc(a/b))
//DEF_OPF(mer, a==0 ? 0 : b%a, a==0 ? 0 : b-a*gf_trunc(b/a))
DEF_OP(rem, b==0?0:a%b, false, side==at_left&&x==0 || side==at_right&&x==1)
DEF_OP(mer, a==0?0:b%a, false, side==at_left&&x==0 || side==at_right&&x==1)
#endif
#ifdef PASS2
DEF_OP(gcd, gcd(a,b), x==0, x==1)
DEF_OP(gcd2, gcd2(a,b), x==0, x==1) // should test those and pick one of the two
DEF_OP(lcm, a==0 || b==0 ? 0 : lcm(a,b), x==1, x==0)
DEF_OPF(or , a|b, (float32)((int32)a | (int32)b), x==0, x==nt_all_ones(&x))
DEF_OPF(xor, a^b, (float32)((int32)a ^ (int32)b), x==0, false)
DEF_OPF(and, a&b, (float32)((int32)a & (int32)b), x==nt_all_ones(&x), x==0)
DEF_OPF(shl, a<<b, a*pow(2.0,+b), side==at_right && x==0, false)
DEF_OPF(shr, a>>b, a*pow(2.0,-b), side==at_right && x==0, false)
DEF_OP(sc_and, a ? b : a, side==at_left && x!=0, side==at_left && x==0)
DEF_OP(sc_or,  a ? a : b, side==at_left && x==0, side==at_left && x!=0)
DEF_OP(min, min(a,b), x==nt_greatest(&x), x==nt_smallest(&x))
DEF_OP(max, max(a,b), x==nt_smallest(&x), x==nt_greatest(&x))
#endif
#ifdef PASS3
DEF_OP(cmp, cmp(a,b), false, false)
DEF_OP(eq,  a == b, false, false)
DEF_OP(ne,  a != b, false, false)
DEF_OP(gt,  a >  b, false, side==at_left&&x==nt_smallest(&x)||side==at_right&&x==nt_greatest(&x))
DEF_OP(le,  a <= b, false, side==at_left&&x==nt_smallest(&x)||side==at_right&&x==nt_greatest(&x))
DEF_OP(lt,  a <  b, false, side==at_left&&x==nt_greatest(&x)||side==at_right&&x==nt_smallest(&x))
DEF_OP(ge,  a >= b, false, side==at_left&&x==nt_greatest(&x)||side==at_right&&x==nt_smallest(&x))
DEF_OP(sin, (T)(b * sin(a * (M_PI / 18000))), false, false) // "LN=9000+36000n RA=0 LA=..."
DEF_OP(cos, (T)(b * cos(a * (M_PI / 18000))), false, false) // "LN=36000n RA=0 LA=..."
DEF_OP(atan, (T)(atan2(a,b) * (18000 / M_PI)), false, false) // "LA=0"
DEF_OP(tanh, (T)(b * tanh(a * (M_PI / 18000))), false, x==0)
DEF_OP(gamma, b<=0 ? 0 : (T)(0+floor(pow(a/256.0,256.0/b)*256.0)), false, false) // "RN=256"
DEF_OP(pow, ipow(a,b), false, false) // "RN=1"
DEF_OP(log, (T)(a==0 ? 0 : b * log(gf_abs(a))), false, false) // "RA=0"
// 0.7.8
//DEF_OPF(clipadd, clipadd(a,b), a+b, x==0, false)
//DEF_OPF(clipsub, clipsub(a,b), a-b, side==at_right && x==0, false)
DEF_OP(abssub,  gf_abs(a-b), false, false)
DEF_OP(sqsub, (a-b)*(a-b), false, false)
DEF_OP(avg, (a+b)/2, false, false)
DEF_OP(hypot, (T)(0+floor(sqrt(a*a+b*b))), false, false)
DEF_OP(sqrt, (T)(0+floor(sqrt(a))), false, false)
DEF_OP(rand, a==0 ? 0 : random()%(int32)a, false, false)
//DEF_OP(erf,"erf*", 0)
#endif
extern Numop      op_table1[], op_table2[], op_table3[];
extern const long op_table1_n, op_table2_n, op_table3_n;
#ifdef PASS1
Numop op_table1[] = {
    DECL_OP(ignore, "ignore", OP_ASSOC),
    DECL_OP(put, "put", OP_ASSOC),
    DECL_OP(add, "+", OP_ASSOC|OP_COMM), // "LINV=sub"
    DECL_OP(sub, "-", 0),
    DECL_OP(bus, "inv+", 0),
    DECL_OP(mul, "*", OP_ASSOC|OP_COMM),
    DECL_OP_NOFLOAT(mulshr8, "*>>8", OP_ASSOC|OP_COMM),
    DECL_OP(div, "/", 0),
    DECL_OP_NOFLOAT(div2, "div", 0),
    DECL_OP(vid, "inv*", 0),
    DECL_OP_NOFLOAT(vid2, "swapdiv", 0),
    DECL_OP_NOFLOAT(mod, "%", 0),
    DECL_OP_NOFLOAT(dom, "swap%", 0),
    DECL_OP_NOFLOAT(rem, "rem", 0),
    DECL_OP_NOFLOAT(mer, "swaprem", 0),
};
const long op_table1_n = COUNT(op_table1);
#endif
#ifdef PASS2
Numop op_table2[] = {
    DECL_OP_NOFLOAT(gcd, "gcd", OP_ASSOC|OP_COMM),
    DECL_OP_NOFLOAT(gcd2, "gcd2", OP_ASSOC|OP_COMM),
    DECL_OP_NOFLOAT(lcm, "lcm", OP_ASSOC|OP_COMM),
    DECL_OP(or , "|", OP_ASSOC|OP_COMM),
    DECL_OP(xor, "^", OP_ASSOC|OP_COMM),
    DECL_OP(and, "&", OP_ASSOC|OP_COMM),
    DECL_OP_NOFOLD(shl, "<<", 0),
    DECL_OP_NOFOLD(shr, ">>", 0),
    DECL_OP_NOFOLD(sc_and,"&&", 0),
    DECL_OP_NOFOLD(sc_or, "||", 0),
    DECL_OP(min, "min", OP_ASSOC|OP_COMM),
    DECL_OP(max, "max", OP_ASSOC|OP_COMM),
};
const long op_table2_n = COUNT(op_table2);
#endif
#ifdef PASS3
Numop op_table3[] = {
    DECL_OP_NOFOLD(eq,  "==", OP_COMM),
    DECL_OP_NOFOLD(ne,  "!=", OP_COMM),
    DECL_OP_NOFOLD(gt,  ">", 0),
    DECL_OP_NOFOLD(le,  "<=", 0),
    DECL_OP_NOFOLD(lt,  "<", 0),
    DECL_OP_NOFOLD(ge,  ">=", 0),
    DECL_OP_NOFOLD(cmp, "cmp", 0),
    DECL_OP_NOFOLD(sin, "sin*", 0),
    DECL_OP_NOFOLD(cos, "cos*", 0),
    DECL_OP_NOFOLD(atan, "atan", 0),
    DECL_OP_NOFOLD(tanh, "tanh*", 0),
    DECL_OP_NOFOLD(gamma, "gamma", 0),
    DECL_OP_NOFOLD(pow, "**", 0),
    DECL_OP_NOFOLD(log, "log*", 0),
// 0.7.8
//	DECL_OP(clipadd,"clip+", OP_ASSOC|OP_COMM),
//	DECL_OP(clipsub,"clip-", 0),
    DECL_OP_NOFOLD(abssub,"abs-", OP_COMM),
    DECL_OP_NOFOLD(sqsub,"sq-", OP_COMM),
    DECL_OP_NOFOLD(avg,"avg", OP_COMM),
    DECL_OP_NOFOLD(hypot,"hypot", OP_COMM),
    DECL_OP_NOFOLD(sqrt,"sqrt", 0),
    DECL_OP_NOFOLD(rand,"rand", 0),
    //DECL_OP_NOFOLD(erf,"erf*", 0),
};
const long op_table3_n = COUNT(op_table3);
#endif
// D=dictionary, A=table, A##_n=table count.
#define INIT_TABLE(D,A) { D=IEVAL(mGridFlow,"@"#D" ||= {}"); \
    for(int i=0; i<A##_n; i++) { \
    	A[i].sym = ID2SYM(rb_intern(A[i].name)); \
    	rb_hash_aset(D,A[i].sym,PTR2FIX((A+i)));}}
#ifdef PASS1
Ruby op_dict = Qnil;
Ruby number_type_dict = Qnil;
void startup_number () {
    INIT_TABLE(op_dict,op_table1)
    INIT_TABLE(op_dict,op_table2)
    INIT_TABLE(op_dict,op_table3)
    INIT_TABLE(number_type_dict,number_type_table)
for (int i=0; i<COUNT(number_type_table); i++) {
    	number_type_table[i].index = (NumberTypeE) i;
    	char a[64];
    	strcpy(a,number_type_table[i].aliases);
    	char *b = strchr(a,',');
    	if (b) {
    		*b=0;
    		rb_hash_aset(number_type_dict, ID2SYM(rb_intern(b+1)),
    			PTR2FIX(&number_type_table[i]));
    	}
    	rb_hash_aset(number_type_dict, ID2SYM(rb_intern(a)),
    			PTR2FIX(&number_type_table[i]));
    }
// S:name; M:mode; F:replacement function;
#define OVERRIDE_INT(S,M,F) { \
    Numop *foo = FIX2PTR(Numop,rb_hash_aref(op_dict,SYM(S))); \
    foo->on_uint8.op_##M=F; \
    foo->on_int16.op_##M=F; \
    foo->on_int32.op_##M=F; }
    OVERRIDE_INT(ignore,map,quick_ign_map);
    OVERRIDE_INT(ignore,zip,quick_ign_zip);
    //OVERRIDE_INT(put,map,quick_put_map);
    //OVERRIDE_INT(put,zip,quick_put_zip);
    //OVERRIDE_INT(%,map,quick_mod_map); // !@#$ does that make an improvement at all?
}
#endif
--- NEW FILE: flow_objects.rb ---
=begin
    $Id: flow_objects.rb,v 1.1 2005/10/04 02:02:13 matju Exp $
GridFlow
    Copyright (c) 2001,2002,2003,2004 by Mathieu Bouchard
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.
See file ../COPYING for further informations on licensing terms.
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
[...1437 lines suppressed...]
  def _0_float i; @i=i; send_out 0, *@a[@i]; end
  def _1_list(*l) @a[@i]=l; end
  def _0_save(filename,format=nil)
    f=File.open(filename.to_s,"w")
    if format then
      @a.each {|x| f.puts(format.to_s%x) }
    else
      @a.each {|x| f.puts(x.join(",")) }
    end
    f.close
  end
  def _0_load(filename)
    f=File.open(filename.to_s,"r")
    @a.clear
    f.each {|x| @a.push x.split(",").map {|y| Float(y) rescue y.intern }}
    f.close
  end
}
end # module GridFlow
--- NEW FILE: test.rb ---
# $Id: test.rb,v 1.1 2005/10/04 02:02:13 matju Exp $
$:.delete_if {|x| x=='.' }
require "gridflow"
include GridFlow
GridFlow.verbose=true
$imdir = "./images"
$animdir = "/opt/mex"
srand Time.new.to_i
$port = 4200+rand(100)
def pressakey; puts "press return to continue."; readline; end
class Expect < FObject
    def praise(*a)
    	#raise(*a)
    	puts a
[...1047 lines suppressed...]
#test_anim "open file #{$imdir}/g001.jpg"#,"loop 0"
#test_anim "open ppm file #{$animdir}/b.ppm.cat"
#test_anim "open jpeg file #{$imdir}/rgb.jpeg.cat"
#test_anim "open quicktime file BLAH"
#test_anim "open quicktime file #{$imdir}/rgb_uncompressed.mov"
#test_anim "open quicktime file #{$imdir}/test_mjpega.mov"
#test_anim "open ppm gzfile motion_tracking.ppm.cat.gz"
#test_anim "open videodev /dev/video","channel 1","size 480 640"
#test_anim "open videodev /dev/video1 noinit","transfer read"
#test_anim "open videodev /dev/video","channel 1","size 120 160"
#test_anim "open mpeg file /home/matju/net/Animations/washington_zoom_in.mpeg"
#test_anim "open quicktime file /home/matju/Shauna/part_1.mov"
#test_anim "open quicktime file #{$imdir}/gt.mov"
#test_anim "open quicktime file /home/matju/pics/domopers_hi.mov"
#test_anim "open quicktime file /home/matju/net/c.mov"
#test_formats
#test_tcp
#test_sound
#test_metro
#$mainloop.loop
--- NEW FILE: flow_objects.c ---
/*
    $Id: flow_objects.c,v 1.1 2005/10/04 02:02:13 matju Exp $
GridFlow
    Copyright (c) 2001,2002,2003,2004 by Mathieu Bouchard
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.
See file ../COPYING for further informations on licensing terms.
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
[...1150 lines suppressed...]
    }	
} GRID_END
\def void initialize (int32 z) {rb_call_super(argc,argv); this->z=z; }
\classinfo { IEVAL(rself,"install '#perspective',1,1"); }
\end class GridPerspective
static Numop *OP(Ruby x) { return FIX2PTR(Numop,rb_hash_aref(op_dict,x)); }
void startup_flow_objects () {
    op_add = OP(SYM(+));
    op_sub = OP(SYM(-));
    op_mul = OP(SYM(*));
    op_shl = OP(SYM(<<));
    op_mod = OP(SYM(%));
    op_and = OP(SYM(&));
    op_div = OP(SYM(/));
    op_put = OP(SYM(put));
    \startall
}
--- NEW FILE: bitpacking.c ---
/*
    $Id: bitpacking.c,v 1.1 2005/10/04 02:02:13 matju Exp $
GridFlow
    Copyright (c) 2001,2002,2003,2004 by Mathieu Bouchard
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.
See file ../COPYING for further informations on licensing terms.
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.
*/
#include "grid.h.fcs"
#include <math.h>
#include <stdlib.h>
#include <stdio.h>
#define CONVERT1 t = \
    (((in[0] << hb[0]) >> 7) & mask[0]) | \
    (((in[1] << hb[1]) >> 7) & mask[1]) | \
    (((in[2] << hb[2]) >> 7) & mask[2])
#define CONVERT2 \
    for (t=0,i=0; i<self->size; i++) t |= (((in[i] << hb[i]) >> 7) & mask[i]);
#define CONVERT3 \
    for (t=0,i=0; i<self->size; i++) { \
    	t |= ((in[i]>>(7-hb[i]))|(in[i]<<(hb[i]-7))) & mask[i]; \
    }
#define WRITE_LE \
    for (int bytes = self->bytes; bytes; bytes--, t>>=8) *out++ = t;
#define WRITE_BE { int bytes; \
    bytes = self->bytes; \
    while (bytes--) { out[bytes] = t; t>>=8; }\
    out += self->bytes; }
/* this macro would be faster if the _increment_
   was done only once every loop. or maybe gcc does it, i dunno */
#define NTIMES(_x_) \
    for (; n>=4; n-=4) { _x_ _x_ _x_ _x_ } \
    for (; n; n--) { _x_ }
/* this could be faster (use asm) */
void swap32 (int n, Pt<uint32> data) {
    NTIMES({
    	uint32 x = *data;
    	x = (x<<16) | (x>>16);
    	x = ((x&0xff00ff)<<8) | ((x>>8)&0xff00ff);
    	*data++ = x;
    })
}
/* this could be faster (use asm or do it in int32 chunks) */
void swap16 (int n, Pt<uint16> data) {
    NTIMES({ uint16 x = *data; *data++ = (x<<8) | (x>>8); })
}
/* **************************************************************** */
template <class T>
static void default_pack(BitPacking *self, int n, Pt<T> in, Pt<uint8> out) {
    uint32 t;
    int i;
    int hb[4];
    uint32 mask[4];
    int sameorder = self->endian==2 || self->endian==::is_le();
    int size = self->size;
for (i=0; i<self->size; i++) hb[i] = highest_bit(self->mask[i]);
    memcpy(mask,self->mask,size*sizeof(uint32));
if (sameorder && size==3) {
    	switch(self->bytes) {
    	case 2:	NTIMES(t=CONVERT1; *((int16 *)out)=t; out+=2; in+=3;) return;
    	case 4:	NTIMES(t=CONVERT1; *((int32 *)out)=t; out+=4; in+=3;) return;
    	}
    }
    if (self->is_le()) {
    	switch (size) {
    	case 3: for (; n--; in+=3) {CONVERT1; WRITE_LE;} break;
    	case 4:	for (; n--; in+=4) {CONVERT3; WRITE_LE;} break;
    	default:for (; n--; in+=size) {CONVERT2; WRITE_LE;}}
    } else {
    	switch (size) {
    	case 3: for (; n--; in+=3) {CONVERT1; WRITE_BE;} break;
    	case 4: for (; n--; in+=4) {CONVERT3; WRITE_BE;} break;
    	default:for (; n--; in+=size) {CONVERT2; WRITE_BE;}}
    }
}
#define LOOP_UNPACK(_reader_) \
    for (; n; n--) { \
    	int bytes=0; uint32 temp=0; _reader_; \
    	for (int i=0; i<self->size; i++, out++) { \
    		uint32 t=temp&self->mask[i]; \
    		*out = (t<<(7-hb[i]))|(t>>(hb[i]-7)); \
    	} \
    }
//			*out++ = ((temp & self->mask[i]) << 7) >> hb[i];
template <class T>
static void default_unpack(BitPacking *self, int n, Pt<uint8> in, Pt<T> out) {
    int hb[4];
    for (int i=0; i<self->size; i++) hb[i] = highest_bit(self->mask[i]);
    if (is_le()) { // smallest byte first
    	LOOP_UNPACK(
    		for(; self->bytes>bytes; bytes++, in++) temp |= *in<<(8*bytes);
    	)
    } else { // biggest byte first
    	LOOP_UNPACK(
    		bytes=self->bytes; for (; bytes; bytes--, in++) temp=(temp<<8)|*in;
    	)
    }
}
/* **************************************************************** */
template <class T>
static void pack2_565(BitPacking *self, int n, Pt<T> in, Pt<uint8> out) {
    const int hb[3] = {15,10,4};
    const uint32 mask[3] = {0x0000f800,0x000007e0,0x0000001f};
    uint32 t;
    NTIMES( t=CONVERT1; *((short *)out)=t; out+=2; in+=3; )
}
template <class T>
static void pack3_888(BitPacking *self, int n, Pt<T> in, Pt<uint8> out) {
    Pt<int32> o32 = (Pt<int32>)out;
    while (n>=4) {
    	o32[0] = (in[5]<<24) | (in[ 0]<<16) | (in[ 1]<<8) | in[2];
    	o32[1] = (in[7]<<24) | (in[ 8]<<16) | (in[ 3]<<8) | in[4];
    	o32[2] = (in[9]<<24) | (in[10]<<16) | (in[11]<<8) | in[6];
    	o32+=3; in+=12;
    	n-=4;
    }
    out = (Pt<uint8>)o32;
    NTIMES( out[2]=in[0]; out[1]=in[1]; out[0]=in[2]; out+=3; in+=3; )
}
/*
template <>
static void pack3_888(BitPacking *self, int n, Pt<uint8> in, Pt<uint8> out) {
    Pt<uint32> o32 = Pt<uint32>((uint32 *)out.p,n*3/4);
    Pt<uint32> i32 = Pt<uint32>((uint32 *)in.p,n*3/4);
    while (n>=4) {
#define Z(w,i) ((word##w>>(i*8))&255)
    	uint32 word0 = i32[0];
    	uint32 word1 = i32[1];
    	uint32 word2 = i32[2];
    	o32[0] = (Z(1,1)<<24) | (Z(0,0)<<16) | (Z(0,1)<<8) | Z(0,2);
    	o32[1] = (Z(1,3)<<24) | (Z(2,0)<<16) | (Z(0,3)<<8) | Z(1,0);
    	o32[2] = (Z(2,1)<<24) | (Z(2,2)<<16) | (Z(2,3)<<8) | Z(1,2);
    	o32+=3; i32+=3;
    	n-=4;
    }
#undef Z
    out = (Pt<uint8>)o32;
    in  = (Pt<uint8>)i32;
    NTIMES( out[2]=in[0]; out[1]=in[1]; out[0]=in[2]; out+=3; in+=3; )
}
*/
template <class T>
static void pack3_888b(BitPacking *self, int n, Pt<T> in, Pt<uint8> out) {
    Pt<int32> o32 = (Pt<int32>)out;
    while (n>=4) {
    	o32[0] = (in[0]<<16) | (in[1]<<8) | in[2];
    	o32[1] = (in[3]<<16) | (in[4]<<8) | in[5];
    	o32[2] = (in[6]<<16) | (in[7]<<8) | in[8];
    	o32[3] = (in[9]<<16) | (in[10]<<8) | in[11];
    	o32+=4; in+=12;
    	n-=4;
    }
    NTIMES( o32[0] = (in[0]<<16) | (in[1]<<8) | in[2]; o32++; in+=3; )
}
/* (R,G,B,?) -> B:8,G:8,R:8,0:8 */
template <class T>
static void pack3_bgrn8888(BitPacking *self, int n, Pt<T> in, Pt<uint8> out) {
/* NTIMES( out[2]=in[0]; out[1]=in[1]; out[0]=in[2]; out+=4; in+=4; ) */
    Pt<int32> i32 = (Pt<int32>)in;
    Pt<int32> o32 = (Pt<int32>)out;
    while (n>=4) {
    	o32[0] = ((i32[0]&0xff)<<16) | (i32[0]&0xff00) | ((i32[0]>>16)&0xff);
    	o32[1] = ((i32[1]&0xff)<<16) | (i32[1]&0xff00) | ((i32[1]>>16)&0xff);
    	o32[2] = ((i32[2]&0xff)<<16) | (i32[2]&0xff00) | ((i32[2]>>16)&0xff);
    	o32[3] = ((i32[3]&0xff)<<16) | (i32[3]&0xff00) | ((i32[3]>>16)&0xff);
    	o32+=4; i32+=4; n-=4;
    }
    NTIMES( o32[0] = ((i32[0]&0xff)<<16) | (i32[0]&0xff00) | ((i32[0]>>16)&0xff); o32++; i32++; )
}
static uint32 bp_masks[][4] = {
    {0x0000f800,0x000007e0,0x0000001f,0},
    {0x00ff0000,0x0000ff00,0x000000ff,0},
};
static Packer bp_packers[] = {
    {default_pack, default_pack, default_pack},
    {pack2_565, pack2_565, pack2_565},
    {pack3_888, pack3_888, pack3_888},
    {pack3_888b, default_pack, default_pack},
    {pack3_bgrn8888, default_pack, default_pack},
};
static Unpacker bp_unpackers[] = {
    {default_unpack, default_unpack, default_unpack},
};
static BitPacking builtin_bitpackers[] = {
    BitPacking(2, 2, 3, bp_masks[0], &bp_packers[1], &bp_unpackers[0]),
    BitPacking(1, 3, 3, bp_masks[1], &bp_packers[2], &bp_unpackers[0]),
    BitPacking(1, 4, 3, bp_masks[1], &bp_packers[3], &bp_unpackers[0]),
    BitPacking(1, 4, 4, bp_masks[1], &bp_packers[4], &bp_unpackers[0]),
};
/* **************************************************************** */
bool BitPacking::eq(BitPacking *o) {
    if (!(bytes == o->bytes)) return false;
    if (!(size == o->size)) return false;
    for (int i=0; i<size; i++) {
    	if (!(mask[i] == o->mask[i])) return false;
    }
    if (endian==o->endian) return true;
    /* same==little on a little-endian; same==big on a big-endian */
    return (endian ^ o->endian ^ ::is_le()) == 2;
}
BitPacking::BitPacking(int endian, int bytes, int size, uint32 *mask,
Packer *packer, Unpacker *unpacker) {
    this->endian = endian;
    this->bytes = bytes;
    this->size = size;
    for (int i=0; i<size; i++) this->mask[i] = mask[i];
    if (packer) {
    	this->packer = packer;
    	this->unpacker = unpacker;
    	return;
    }
    int packeri=-1;
    this->packer = &bp_packers[0];
    this->unpacker = &bp_unpackers[0];
for (int i=0; i<(int)(sizeof(builtin_bitpackers)/sizeof(BitPacking)); i++) {
    	BitPacking *bp = &builtin_bitpackers[i];
    	if (this->eq(bp)) {
    		this->packer = bp->packer;
    		this->unpacker = bp->unpacker;
    		packeri=i;
    		goto end;
    	}
    }
end:;
/*
    ::gfpost("Bitpacking: endian=%d bytes=%d size=%d packeri=%d",
    	endian, bytes, size, packeri);
    ::gfpost("  packer=0x%08x unpacker=0x%08x",this->packer,this->unpacker);
    ::gfpost("  mask=[0x%08x,0x%08x,0x%08x,0x%08x]",mask[0],mask[1],mask[2],mask[3]);
*/
}
bool BitPacking::is_le() {
    return endian==1 || (endian ^ ::is_le())==3;
}
template <class T>
void BitPacking::pack(int n, Pt<T> in, Pt<uint8> out) {
    switch (NumberTypeE_type_of(*in)) {
    case uint8_e: packer->as_uint8(this,n,(Pt<uint8>)in,out); break;
    case int16_e: packer->as_int16(this,n,(Pt<int16>)in,out); break;
    case int32_e: packer->as_int32(this,n,(Pt<int32>)in,out); break;
    default: RAISE("argh");
    }
}
template <class T>
void BitPacking::unpack(int n, Pt<uint8> in, Pt<T> out) {
    switch (NumberTypeE_type_of(*out)) {
    case uint8_e: unpacker->as_uint8(this,n,in,(Pt<uint8>)out); break;
    case int16_e: unpacker->as_int16(this,n,in,(Pt<int16>)out); break;
    case int32_e: unpacker->as_int32(this,n,in,(Pt<int32>)out); break;
    default: RAISE("argh");
    }
}
// i'm sorry... see the end of grid.c for an explanation...
//static
void make_hocus_pocus () {
//	exit(1);
#define FOO(S) \
    ((BitPacking*)0)->pack(0,Pt<S>(),Pt<uint8>()); \
    ((BitPacking*)0)->unpack(0,Pt<uint8>(),Pt<S>());
EACH_NUMBER_TYPE(FOO)
#undef FOO
}
--- NEW FILE: grid.h ---
/*
    $Id: grid.h,v 1.1 2005/10/04 02:02:13 matju Exp $
GridFlow
    Copyright (c) 2001,2002,2003,2004 by Mathieu Bouchard
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.
See file ../COPYING for further informations on licensing terms.
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
[...1070 lines suppressed...]
    GFStackFrame s[GF_STACK_MAX];
    int n;
    GFStack() { n = 0; }
    void push (FObject *o) __attribute__((noinline));
    void pop () __attribute__((noinline));
};
extern GFStack gf_stack;
struct GFStackMarker {
    int n;
    bool flag;
    GFStackMarker(FObject *o) { n = gf_stack.n; gf_stack.push(o); flag=true; }
    ~GFStackMarker() { while (gf_stack.n != n) gf_stack.pop(); }
    bool once () {
    	if (flag) { flag=false; return true; } else return false;
    }
};
typedef GridObject Format;
#endif // __GF_GRID_H
--- NEW FILE: main.c ---
/*
    $Id: main.c,v 1.1 2005/10/04 02:02:13 matju Exp $
GridFlow
    Copyright (c) 2001,2002,2003,2004 by Mathieu Bouchard
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.
See file ../COPYING for further informations on licensing terms.
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.
*/
#include <stdlib.h>
#include <sys/stat.h>
#include <sys/time.h>
#include <time.h>
#include <stdarg.h>
#include <string.h>
#include <signal.h>
#include <stdio.h>
#include <fcntl.h>
#include <unistd.h>
#include <stdarg.h>
#include "grid.h.fcs"
#include "../config.h"
#include <assert.h>
#include <limits.h>
BuiltinSymbols bsym;
GFStack gf_stack;
Ruby mGridFlow;
Ruby cFObject;
extern "C"{
void rb_raise0(
const char *file, int line, const char *func, VALUE exc, const char *fmt, ...) {
    va_list args;
    char buf[BUFSIZ];
    va_start(args,fmt);
    vsnprintf(buf, BUFSIZ, fmt, args);
    buf[BUFSIZ-1]=0;
    va_end(args);
    VALUE e = rb_exc_new2(exc, buf);
    char buf2[BUFSIZ];
    snprintf(buf2, BUFSIZ, "%s:%d:in `%s'", file, line, func);
    buf2[BUFSIZ-1]=0;
    VALUE ary = rb_funcall(e,SI(caller),0);
    if (gf_stack.n) {
    	rb_funcall(ary,SI(unshift),2,rb_str_new2(buf2),
    		rb_str_new2(INFO(gf_stack.s[gf_stack.n-1].o)));
    } else {
    	rb_funcall(ary,SI(unshift),1,rb_str_new2(buf2));
    }
    rb_funcall(e,SI(set_backtrace),1,ary);
    rb_exc_raise(e);
}};
Ruby rb_ary_fetch(Ruby rself, int i) {
    Ruby argv[] = { INT2NUM(i) };
    return rb_ary_aref(COUNT(argv),argv,rself);
}
//----------------------------------------------------------------
// CObject
static void CObject_mark (void *z) {}
void CObject_free (void *foo) {
    CObject *self = (CObject *)foo;
    self->check_magic();
    if (!self->rself) {
    	fprintf(stderr,"attempt to free object that has no rself\n");
    	abort();
    }
    self->rself = 0; /* paranoia */
    delete self;
}
//----------------------------------------------------------------
// Dim
void Dim::check() {
    if (n>MAX_DIMENSIONS) RAISE("too many dimensions");
    for (int i=0; i<n; i++) if (v[i]<0) RAISE("Dim: negative dimension");
}
// !@#$ big leak machine?
// returns a string like "Dim[240,320,3]"
char *Dim::to_s() {
    // if you blow 256 chars it's your own fault
    char buf[256];
    char *s = buf;
    s += sprintf(s,"Dim[");
    for(int i=0; i<n; i++) s += sprintf(s,"%s%d", ","+!i, v[i]);
    s += sprintf(s,"]");
    return strdup(buf);
}
//----------------------------------------------------------------
\class FObject < CObject
static void FObject_prepare_message(int &argc, Ruby *&argv, Ruby &sym, FObject *foo=0) {
    if (argc<1) {
    	sym = bsym._bang;
    } else if (argc>1 && !SYMBOL_P(*argv)) {
    	sym = bsym._list;
    } else if (INTEGER_P(*argv)||FLOAT_P(*argv)) {
    	sym = bsym._float;
    } else if (SYMBOL_P(*argv)) {
    	sym = *argv;
    	argc--, argv++;
    } else if (argc==1 && TYPE(*argv)==T_ARRAY) {
    	sym = bsym._list;
    	argc = rb_ary_len(*argv);
    	argv = rb_ary_ptr(*argv);
    } else {
    	RAISE("%s received bad message: argc=%d; argv[0]=%s",foo?INFO(foo):"", argc,
    		argc ? rb_str_ptr(rb_inspect(argv[0])) : "");
    }
}
struct Helper {
    int argc;
    Ruby *argv;
    FObject *self;
    Ruby rself;
    int n; // stack level
};
static Ruby GridFlow_handle_braces(Ruby rself, Ruby argv);
// inlet #-1 is reserved for SystemInlet messages
// inlet #-2 is for inlet #0 messages that happen at start time
static void send_in_2 (Helper *h) { PROF(h->self) {
    int argc = h->argc;
    Ruby *argv = h->argv;
    if (h->argc<1) RAISE("not enough args");
    int inlet = INT(argv[0]);
    argc--, argv++;
    Ruby foo;
    if (argc==1 && TYPE(argv[0])==T_STRING /* && argv[0] =~ / / */) {
    	foo = rb_funcall(mGridFlow,SI(parse),1,argv[0]);
    	argc = rb_ary_len(foo);
    	argv = rb_ary_ptr(foo);
    }
    if (argc>1) {
    	foo = rb_ary_new4(argc,argv);
    	GridFlow_handle_braces(0,foo);
    	argc = rb_ary_len(foo);
    	argv = rb_ary_ptr(foo);
    }
    if (inlet==-2) {
    	Array init_messages = rb_ivar_get(h->rself,SI(@init_messages));
    	rb_ary_push(init_messages, rb_ary_new4(argc,argv));
    	inlet=0;
    }
    if (inlet<0 || inlet>9 /*|| inlet>real_inlet_max*/)
    	if (inlet!=-3 && inlet!=-1) RAISE("invalid inlet number: %d", inlet);
    Ruby sym;
    FObject_prepare_message(argc,argv,sym,h->self);
//	if (rb_const_get(mGridFlow,SI(@verbose))==Qtrue) gfpost m.inspect
    char buf[256];
    if (inlet==-1) sprintf(buf,"_sys_%s",rb_sym_name(sym));
    else           sprintf(buf,"_%d_%s",inlet,rb_sym_name(sym));
    rb_funcall2(h->rself,rb_intern(buf),argc,argv);
} /* PROF */ }
static void send_in_3 (Helper *h) {
    while (gf_stack.n > h->n) gf_stack.pop();
}
\def void send_in (...) {
    Helper h = {argc,argv,this,rself,gf_stack.n};
    rb_ensure(
    	(RMethod)send_in_2,(Ruby)&h,
    	(RMethod)send_in_3,(Ruby)&h);
}
\def void send_out (...) {
    int n=0;
    if (argc<1) RAISE("not enough args");
    int outlet = INT(*argv);
    if (outlet<0 || outlet>9 /*|| outlet>real_outlet_max*/)
    	RAISE("invalid outlet number: %d",outlet);
    argc--, argv++;
    Ruby sym;
    FObject_prepare_message(argc,argv,sym,this);
    Ruby noutlets2 = rb_ivar_get(rb_obj_class(rself),SYM2ID(SYM(@noutlets)));
    if (TYPE(noutlets2)!=T_FIXNUM) {
    	IEVAL(rself,"STDERR.puts inspect");
    	RAISE("don't know how many outlets this has");
    }
    int noutlets = INT(noutlets2);
    //if (outlet<0 || outlet>=noutlets) RAISE("outlet %d does not exist",outlet);
    // was PROF(0) a hack because of exception-handling problems?
    PROF(0) {
    Ruby argv2[argc+2];
    for (int i=0; i<argc; i++) argv2[2+i] = argv[i];
    argv2[0] = INT2NUM(outlet);
    argv2[1] = sym;
    rb_funcall2(rself,SI(send_out2), argc+2, argv2);
Ruby ary = rb_ivar_defined(rself,SYM2ID(bsym.iv_outlets)) ?
    	rb_ivar_get(rself,SYM2ID(bsym.iv_outlets)) : Qnil;
    if (ary==Qnil) goto end;
    if (TYPE(ary)!=T_ARRAY) RAISE("send_out: expected array");
    ary = rb_ary_fetch(ary,outlet);
    if (ary==Qnil) goto end;
    if (TYPE(ary)!=T_ARRAY) RAISE("send_out: expected array");
    n = rb_ary_len(ary);
for (int i=0; i<n; i++) {
    	Ruby conn = rb_ary_fetch(ary,i);
    	Ruby rec = rb_ary_fetch(conn,0);
    	int inl = INT(rb_ary_fetch(conn,1));
    	argv2[0] = INT2NUM(inl);
    	rb_funcall2(rec,SI(send_in),argc+2,argv2);
    }
    } /* PROF */
end:;
}
Ruby FObject_s_new(Ruby argc, Ruby *argv, Ruby qlass) {
    Ruby allocator = rb_ivar_defined(qlass,SI(@allocator)) ?
    	rb_ivar_get(qlass,SI(@allocator)) : Qnil;
    FObject *self;
    if (allocator==Qnil) {
    	// this is a pure-ruby FObject/GridObject
    	// !@#$ GridObject is in FObject constructor (ugly)
    	self = new GridObject;
    } else {
    	// this is a C++ FObject/GridObject
    	void*(*alloc)() = (void*(*)())FIX2PTR(void,allocator);
    	self = (FObject *)alloc();
    }
    self->check_magic();
    Ruby keep = rb_ivar_get(mGridFlow, SI(@fobjects));
    self->bself = 0;
    Ruby rself = Data_Wrap_Struct(qlass, CObject_mark, CObject_free, self);
    self->rself = rself;
    rb_hash_aset(keep,rself,Qtrue); // prevent sweeping
    rb_funcall2(rself,SI(initialize),argc,argv);
    return rself;
}
Ruby FObject_s_install(Ruby rself, Ruby name, Ruby inlets2, Ruby outlets2) {
    int inlets, outlets;
    Ruby name2;
    if (SYMBOL_P(name)) {
    	name2 = rb_funcall(name,SI(to_str),0);
    } else if (TYPE(name) == T_STRING) {
    	name2 = rb_funcall(name,SI(dup),0);
    } else {
    	RAISE("expect symbol or string");
    }
    inlets  =  INT(inlets2); if ( inlets<0 ||  inlets>9) RAISE("...");
    outlets = INT(outlets2); if (outlets<0 || outlets>9) RAISE("...");
    rb_ivar_set(rself,SI(@ninlets),INT2NUM(inlets));
    rb_ivar_set(rself,SI(@noutlets),INT2NUM(outlets));
    rb_ivar_set(rself,SI(@foreign_name),name2);
    rb_hash_aset(rb_ivar_get(mGridFlow,SI(@fclasses)), name2, rself);
    rb_funcall(rself, SI(install2), 1, name2);
    return Qnil;
}
\def Ruby total_time_get () {return gf_ull2num(total_time);}
\def Ruby total_time_set (Ruby x) {
    if (argc<1) RAISE("muh");
    total_time = TO(uint64,x);
    return argv[0];
}
\def void delete_m () {
    Ruby keep = rb_ivar_get(mGridFlow, SI(@fobjects));
    rb_funcall(keep,SI(delete),1,rself);
}
\classinfo
\end class FObject
/* ---------------------------------------------------------------- */
/* C++<->Ruby bridge for classes/functions in base/number.c */
static Ruby String_swap32_f (Ruby rself) {
    int n = rb_str_len(rself)/4;
    swap32(n,Pt<uint32>((uint32 *)rb_str_ptr(rself),n));
    return rself;
}
static Ruby String_swap16_f (Ruby rself) {
    int n = rb_str_len(rself)/2;
    swap16(n,Pt<uint16>((uint16 *)rb_str_ptr(rself),n));
    return rself;
}
NumberTypeE NumberTypeE_find (Ruby sym) {
    if (TYPE(sym)!=T_SYMBOL) RAISE("expected symbol (not %s)",
    	rb_str_ptr(rb_inspect(rb_obj_class(sym))));
    Ruby nt_dict = rb_ivar_get(mGridFlow,SI(@number_type_dict));
    Ruby v = rb_hash_aref(nt_dict,sym);
    if (v!=Qnil) return FIX2PTR(NumberType,v)->index;
    RAISE("unknown number type "%s"", rb_sym_name(sym));
}
/* **************************************************************** */
\class BitPacking < CObject
\def void initialize(Ruby foo1, Ruby foo2, Ruby foo3) {}
// !@#$ doesn't support number types
\def String pack2 (String ins, String outs=Qnil) {
    int n = rb_str_len(ins) / sizeof(int32) / size;
    Pt<int32> in = Pt<int32>((int32 *)rb_str_ptr(ins),rb_str_len(ins));
    int bytes2 = n*bytes;
    Ruby out = outs!=Qnil ? rb_str_resize(outs,bytes2) : rb_str_new("",bytes2);
    rb_str_modify(out);
    pack(n,Pt<int32>(in,n),Pt<uint8>((uint8 *)rb_str_ptr(out),bytes2));
    return out;
}
// !@#$ doesn't support number types
\def String unpack2 (String ins, String outs=Qnil) {
    int n = rb_str_len(argv[0]) / bytes;
    Pt<uint8> in = Pt<uint8>((uint8 *)rb_str_ptr(ins),rb_str_len(ins));
    int bytes2 = n*size*sizeof(int32);
    Ruby out = outs!=Qnil ? rb_str_resize(outs,bytes2) : rb_str_new("",bytes2);
    rb_str_modify(out);
    unpack(n,Pt<uint8>((uint8 *)in,bytes2),Pt<int32>((int32 *)rb_str_ptr(out),n));
    return out;
}
static Ruby BitPacking_s_new(Ruby argc, Ruby *argv, Ruby qlass) {
    Ruby keep = rb_ivar_get(mGridFlow, rb_intern("@fobjects"));
    if (argc!=3) RAISE("bad args");
    if (TYPE(argv[2])!=T_ARRAY) RAISE("bad mask");
    int endian = INT(argv[0]);
    int bytes = INT(argv[1]);
    Ruby *masks = rb_ary_ptr(argv[2]);
    uint32 masks2[4];
    int size = rb_ary_len(argv[2]);
    if (size<1) RAISE("not enough masks");
    if (size>4) RAISE("too many masks (%d)",size);
    for (int i=0; i<size; i++) masks2[i] = NUM2UINT(masks[i]);
    BitPacking *self = new BitPacking(endian,bytes,size,masks2);
    Ruby rself = Data_Wrap_Struct(qlass, 0, CObject_free, self);
    self->rself = rself;
    rb_hash_aset(keep,rself,Qtrue); // prevent sweeping (leak) (!@#$ WHAT???)
    rb_funcall2(rself,SI(initialize),argc,argv);
    return rself;
}
\classinfo
\end class BitPacking
void gfpost(const char *fmt, ...) {
    va_list args;
    int length;
    va_start(args,fmt);
    const int n=256;
    char post_s[n];
    length = vsnprintf(post_s,n,fmt,args);
    if (length<0 || length>=n) sprintf(post_s+n-6,"[...]"); /* safety */
    va_end(args);
    rb_funcall(mGridFlow,SI(gfpost2),2,rb_str_new2(fmt),rb_str_new2(post_s));
}
void define_many_methods(Ruby rself, int n, MethodDecl *methods) {
    for (int i=0; i<n; i++) {
    	MethodDecl *md = &methods[i];
    	char *buf = strdup(md->selector);
    	if (strlen(buf)>2 && strcmp(buf+strlen(buf)-2,"_m")==0)
    		buf[strlen(buf)-2]=0;
    	rb_define_method(rself,buf,(RMethod)md->method,-1);
    	rb_enable_super(rself,buf);
    	free(buf);
    }
}
static Ruby GridFlow_fclass_install(Ruby rself_, Ruby fc_, Ruby super) {
    FClass *fc = FIX2PTR(FClass,fc_);
    Ruby rself = super!=Qnil ?
    	rb_define_class_under(mGridFlow, fc->name, super) :
    	rb_funcall(mGridFlow,SI(const_get),1,rb_str_new2(fc->name));
    define_many_methods(rself,fc->methodsn,fc->methods);
    rb_ivar_set(rself,SI(@allocator),PTR2FIX((void*)(fc->allocator))); //#!@$??
    if (fc->startup) fc->startup(rself);
    return Qnil;
}
//----------------------------------------------------------------
// GridFlow.class
//\class GridFlow_s < patate
typedef void (*Callback)(void*);
static Ruby GridFlow_exec (Ruby rself, Ruby data, Ruby func) {
    void *data2 = FIX2PTR(void,data);
    Callback func2 = (Callback) FIX2PTR(void,func);
    func2(data2);
    return Qnil;
}
static Ruby GridFlow_get_id (Ruby rself, Ruby arg) {
    fprintf(stderr,"%ld\n",arg);
    return INT2NUM((int)arg);
}
Ruby GridFlow_rdtsc (Ruby rself) { return gf_ull2num(rdtsc()); }
/* This code handles nested lists because PureData (0.38) doesn't do it */
static Ruby GridFlow_handle_braces(Ruby rself, Ruby argv) {
    int stack[16];
    int stackn=0;
    Ruby *av = rb_ary_ptr(argv);
    int ac = rb_ary_len(argv);
    int j=0;
    for (int i=0; i<ac; ) {
    	int close=0;
    	if (SYMBOL_P(av[i])) {
    		const char *s = rb_sym_name(av[i]);
    		while (*s=='(' || *s=='{') {
    			if (stackn==16) RAISE("too many nested lists (>16)");
    			stack[stackn++]=j;
    			s++;
    		}
    		const char *se = s+strlen(s);
    		while (se[-1]==')' || se[-1]=='}') { se--; close++; }
    		if (s!=se) {
    			Ruby u = rb_str_new(s,se-s);
    			av[j++] = rb_funcall(rself,SI(FloatOrSymbol),1,u);
    		}
    	} else {
    		av[j++]=av[i];
    	}
    	i++;
    	while (close--) {
    		if (!stackn) RAISE("unbalanced '}' or ')'",av[i]);
    		Ruby a2 = rb_ary_new();
    		int j2 = stack[--stackn];
    		for (int k=j2; k<j; k++) rb_ary_push(a2,av[k]);
    		j=j2;
    		av[j++] = a2;
    	}
    }
    if (stackn) RAISE("unbalanced '{' or '(' (stackn=%d)",stackn);
    RARRAY(argv)->len = j;
    return rself;
}
/* ---------------------------------------------------------------- */
static uint32 memcpy_calls = 0;
static uint64 memcpy_bytes = 0;
static uint64 memcpy_time  = 0;
static uint32 malloc_calls = 0; /* only new not delete */
static uint64 malloc_bytes = 0; /* only new not delete */
static uint64 malloc_time  = 0; /* in cpu ticks */
// don't touch.
static void gfmemcopy32(int32 *as, int32 *bs, int n) {
    int32 ba = bs-as;
#define FOO(I) as[I] = (as+ba)[I];
    	UNROLL_8(FOO,n,as)
#undef FOO
}
void gfmemcopy(uint8 *out, const uint8 *in, int n) {
    uint64 t = rdtsc();
    memcpy_calls++;
    memcpy_bytes+=n;
    for (; n>16; in+=16, out+=16, n-=16) {
    	((int32*)out)[0] = ((int32*)in)[0];
    	((int32*)out)[1] = ((int32*)in)[1];
    	((int32*)out)[2] = ((int32*)in)[2];
    	((int32*)out)[3] = ((int32*)in)[3];
    }
    for (; n>4; in+=4, out+=4, n-=4) { *(int32*)out = *(int32*)in; }
    for (; n; in++, out++, n--) { *out = *in; }
    t=rdtsc()-t;
    memcpy_time+=t;
}
extern "C" {
void *gfmalloc(size_t n) {
    uint64 t = rdtsc();
    void *p = malloc(n);
    long align = (long)p & 7;
    if (align) fprintf(stderr,"malloc alignment = %ld mod 8\n",align);
    t=rdtsc()-t;
    malloc_time+=t;
    malloc_calls++;
    malloc_bytes+=n;
    return p;
}
void gffree(void *p) {
    uint64 t = rdtsc();
    free(p);
    t=rdtsc()-t;
    malloc_time+=t;
}};
Ruby GridFlow_memcpy_calls (Ruby rself) { return   LONG2NUM(memcpy_calls); }
Ruby GridFlow_memcpy_bytes (Ruby rself) { return gf_ull2num(memcpy_bytes); }
Ruby GridFlow_memcpy_time  (Ruby rself) { return gf_ull2num(memcpy_time); }
Ruby GridFlow_malloc_calls (Ruby rself) { return   LONG2NUM(malloc_calls); }
Ruby GridFlow_malloc_bytes (Ruby rself) { return gf_ull2num(malloc_bytes); }
Ruby GridFlow_malloc_time  (Ruby rself) { return gf_ull2num(malloc_time); }
Ruby GridFlow_profiler_reset2 (Ruby rself) {
    memcpy_calls = memcpy_bytes = memcpy_time = 0;
    malloc_calls = malloc_bytes = malloc_time = 0;
    return Qnil;
}
/* ---------------------------------------------------------------- */
void startup_number();
void startup_grid();
void startup_flow_objects();
void startup_flow_objects_for_image();
void startup_flow_objects_for_matrix();
Ruby cFormat;
#define SDEF(_class_,_name_,_argc_) \
    rb_define_singleton_method(c##_class_,#_name_,(RMethod)_class_##_s_##_name_,_argc_)
#define SDEF2(_name1_,_name2_,_argc_) \
    rb_define_singleton_method(mGridFlow,_name1_,(RMethod)_name2_,_argc_)
STARTUP_LIST(void)
// Ruby's entrypoint.
void Init_gridflow () {
#define FOO(_sym_,_name_) bsym._sym_ = ID2SYM(rb_intern(_name_));
BUILTIN_SYMBOLS(FOO)
#undef FOO
    signal(11,SIG_DFL); // paranoia
    mGridFlow = EVAL("module GridFlow; CObject = ::Object; "
    	"class<<self; attr_reader :bridge_name; end; "
    	"def post_string(s) STDERR.puts s end; "
    	"self end");
    SDEF2("exec",GridFlow_exec,2);
    SDEF2("get_id",GridFlow_get_id,1);
    SDEF2("rdtsc",GridFlow_rdtsc,0);
    SDEF2("profiler_reset2",GridFlow_profiler_reset2,0);
    SDEF2("memcpy_calls",GridFlow_memcpy_calls,0);
    SDEF2("memcpy_bytes",GridFlow_memcpy_bytes,0);
    SDEF2("memcpy_time", GridFlow_memcpy_time,0);
    SDEF2("malloc_calls",GridFlow_malloc_calls,0);
    SDEF2("malloc_bytes",GridFlow_malloc_bytes,0);
    SDEF2("malloc_time", GridFlow_malloc_time,0);
    SDEF2("handle_braces!",GridFlow_handle_braces,1);
    SDEF2("fclass_install",GridFlow_fclass_install,2);
//#define FOO(A) fprintf(stderr,"sizeof("#A")=%d\n",sizeof(A));
//FOO(Dim) FOO(BitPacking) FOO(GridHandler) FOO(GridInlet) FOO(GridOutlet) FOO(GridObject)
//#undef FOO
rb_ivar_set(mGridFlow, SI(@fobjects), rb_hash_new());
    rb_ivar_set(mGridFlow, SI(@fclasses), rb_hash_new());
    rb_ivar_set(mGridFlow, SI(@bsym), PTR2FIX(&bsym));
    rb_define_const(mGridFlow, "GF_VERSION", rb_str_new2(GF_VERSION));
    rb_define_const(mGridFlow, "GF_COMPILE_TIME", rb_str_new2(GF_COMPILE_TIME));
cFObject = rb_define_class_under(mGridFlow, "FObject", rb_cObject);
    EVAL(
\ruby
    module GridFlow
    	class FObject
    	def send_out2(*) end
    	def self.install2(*) end
    	def self.add_creator(name)
    		name=name.to_str.dup
    		GridFlow.fclasses[name]=self
    		GridFlow.add_creator_2 name end
    	end
    end
\end ruby
);
    define_many_methods(cFObject,COUNT(FObject_methods),FObject_methods);
    SDEF(FObject, install, 3);
    SDEF(FObject, new, -1);
    ID gbi = SI(gf_bridge_init);
    if (rb_respond_to(rb_cData,gbi)) rb_funcall(rb_cData,gbi,0);
    Ruby cBitPacking =
    	rb_define_class_under(mGridFlow, "BitPacking", rb_cObject);
    define_many_methods(cBitPacking,
    	ciBitPacking.methodsn,
    	ciBitPacking.methods);
    SDEF(BitPacking,new,-1);
    rb_define_method(rb_cString, "swap32!", (RMethod)String_swap32_f, 0);
    rb_define_method(rb_cString, "swap16!", (RMethod)String_swap16_f, 0);
startup_number();
    startup_grid();
    startup_flow_objects();
    startup_flow_objects_for_image();
    startup_flow_objects_for_matrix();
    if (!EVAL("begin require 'gridflow/base/main.rb'; true\n"
    	"rescue Exception => e; "
    	"STDERR.puts "can't load: #{$!}\n"
    	"backtrace: #{$!.backtrace.join"\n"}\n"
    	"$: = #{$:.inspect}"\n; false end")) return;
    cFormat = EVAL("GridFlow::Format");
    STARTUP_LIST()
    EVAL("h=GridFlow.fclasses; h['#io:window'] = h['#io:quartz']||h['#io:x11']||h['#io:sdl']");
    EVAL("GridFlow.load_user_config");
    signal(11,SIG_DFL); // paranoia
}
void GFStack::push (FObject *o) {
    void *bp = &o; // really. just finding our position on the stack.
    if (n>=GF_STACK_MAX)
    	RAISE("stack overflow (maximum %d FObject activations at once)", GF_STACK_MAX);
    uint64 t = rdtsc();
    if (n) s[n-1].time = t - s[n-1].time;
    s[n].o = o;
    s[n].bp = bp;
    s[n].time = t;
    n++;
}
void GFStack::pop () {
    uint64 t = rdtsc();
    if (!n) RAISE("stack underflow (WHAT?)");
    n--;
    if (s[n].o) s[n].o->total_time += t - s[n].time;
    if (n) s[n-1].time = t - s[n-1].time;
}
uint64 gf_timeofday () {
    timeval t;
    gettimeofday(&t,0);
    return t.tv_sec*1000000+t.tv_usec;
}
--- NEW FILE: grid.c ---
/*
    $Id: grid.c,v 1.1 2005/10/04 02:02:13 matju Exp $
GridFlow
    Copyright (c) 2001,2002,2003,2004 by Mathieu Bouchard
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.
See file ../COPYING for further informations on licensing terms.
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.
*/
#include <stdlib.h>
#include <fcntl.h>
#include <errno.h>
#include <stdio.h>
#include <string.h>
#include <unistd.h>
#include <sys/time.h>
#include "grid.h.fcs"
#include <ctype.h>
/* copied from bridge/puredata.c (sorry: linkage issue) */
struct Pointer : CObject { void *p; Pointer(void *_p) : p(_p) {}};
Ruby Pointer_s_noo (void *ptr) {
    return Data_Wrap_Struct(EVAL("GridFlow::Pointer"), 0, 0, new Pointer(ptr));}
static void *Pointer_gut (Ruby rself) {DGS(Pointer); return self->p;}
//#define TRACE fprintf(stderr,"%s %s [%s:%d]\n",INFO(parent),__PRETTY_FUNCTION__,__FILE__,__LINE__);assert(this);
#define TRACE assert(this);
#define CHECK_TYPE(d) \
    if (NumberTypeE_type_of(d)!=this->nt) RAISE("%s(%s): " \
    	"type mismatch during transmission (got %s expecting %s)", \
    	INFO(parent), \
    	__PRETTY_FUNCTION__, \
    	number_type_table[NumberTypeE_type_of(d)].name, \
    	number_type_table[this->nt].name);
#define CHECK_BUSY(s) \
    if (!dim) RAISE("%s: " #s " not busy",INFO(parent));
#define CHECK_ALIGN(d) \
    {int bytes = number_type_table[nt].size/8; \
    int align = ((long)(void*)d)%bytes; \
    if (align) {L;gfpost("%s(%s): Alignment Warning: %p is not %d-aligned: %d", \
    	INFO(parent), __PRETTY_FUNCTION__, (void*)d,bytes,align);}}
#define CHECK_ALIGN2(d,nt) \
    {int bytes = number_type_table[nt].size/8; \
    int align = ((long)(void*)d)%bytes; \
    if (align) {L;gfpost("Alignment Warning: %p is not %d-aligned: %d", \
    	(void*)d,bytes,align);}}
// **************** Grid ******************************************
#define FOO(S) static inline void NUM(Ruby x, S &y) {y=convert(x,(int32*)0);}
EACH_INT_TYPE(FOO)
#undef FOO
#define FOO(S) \
static inline void NUM(Ruby x, S &y) { \
    if (TYPE(x)==T_FLOAT) y = RFLOAT(x)->value; \
    else if (INTEGER_P(x)) y = convert(x,(S*)0); \
    else RAISE("expected Float (or at least Integer)");}
EACH_FLOAT_TYPE(FOO)
#undef FOO
void Grid::init_from_ruby_list(int n, Ruby *a, NumberTypeE nt) {
    Ruby delim = SYM(#);
    for (int i=0; i<n; i++) {
    	if (a[i] == delim) {
    		STACK_ARRAY(int32,v,i);
    		if (i!=0 && TYPE(a[i-1])==T_SYMBOL) nt=NumberTypeE_find(a[--i]);
    		for (int j=0; j<i; j++) v[j] = convert(a[j],(int32*)0);
    		init(new Dim(i,v),nt);
    		CHECK_ALIGN2(this->data,nt);
    		if (a[i] != delim) i++;
    		i++; a+=i; n-=i;
    		goto fill;
    	}
    }
    if (n!=0 && TYPE(a[0])==T_SYMBOL) {
    	nt = NumberTypeE_find(a[0]);
    	a++, n--;
    }
    init(new Dim(n),nt);
    CHECK_ALIGN2(this->data,nt);
    fill:
    int nn = dim->prod();
    n = min(n,nn);
#define FOO(type) { \
    Pt<type> p = (Pt<type>)*this; \
    if (n==0) CLEAR(p,nn); \
    else { \
    	for (int i=0; i<n; i++) NUM(a[i],p[i]); \
    	for (int i=n; i<nn; i+=n) COPY(p+i,p,min(n,nn-i)); }}
    TYPESWITCH(nt,FOO,)
#undef FOO
}
void Grid::init_from_ruby(Ruby x) {
    if (TYPE(x)==T_ARRAY) {
    	init_from_ruby_list(rb_ary_len(x),rb_ary_ptr(x));
    } else if (INTEGER_P(x) || FLOAT_P(x)) {
    	init(new Dim(),int32_e);
    	CHECK_ALIGN2(this->data,nt);
    	((Pt<int32>)*this)[0] = INT(x);
    } else {
    	rb_funcall(
    	EVAL("proc{|x| raise "can't convert to grid: #{x.inspect}"}"),
    	SI(call),1,x);
    }
}
// **************** GridInlet *************************************
// must be set before the end of GRID_BEGIN phase, and so cannot be changed
// afterwards. This is to allow some optimisations. Anyway there is no good reason
// why this would be changed afterwards.
void GridInlet::set_factor(int factor) {
    if(!dim) RAISE("huh?");
    if(factor<=0) RAISE("%s: factor=%d should be >= 1",INFO(parent),factor);
    if (dim->prod() && dim->prod() % factor)
    	RAISE("%s: set_factor: expecting divisor",INFO(parent));
    if (factor > 1) {
    	buf=new Grid(new Dim(factor), nt);
    	bufi=0;
    } else {
    	buf=0;
    }
}
static Ruby GridInlet_begin_1(GridInlet *self) {
#define FOO(T) self->gh->flow(self,-1,Pt<T>()); break;
    TYPESWITCH(self->nt,FOO,)
#undef FOO
    return Qnil;
}
static Ruby GridInlet_begin_2(GridInlet *self) {
    self->dim = 0; // hack
    return (Ruby) 0;
}
bool GridInlet::supports_type(NumberTypeE nt) {
#define FOO(T) return !! gh->flow_##T;
    TYPESWITCH(nt,FOO,return false)
#undef FOO
}
Ruby GridInlet::begin(int argc, Ruby *argv) {TRACE;
    if (!argc) return PTR2FIX(this);
    GridOutlet *back_out = (GridOutlet *) Pointer_gut(argv[0]);
    nt = (NumberTypeE) INT(argv[1]);
    argc-=2, argv+=2;
    PROF(parent) {
    if (dim) {
    	gfpost("%s: grid inlet conflict; aborting %s in favour of %s",
    		INFO(parent), INFO(sender), INFO(back_out->parent));
    	abort();
    }
    sender = back_out->parent;
    if ((int)nt<0 || (int)nt>=(int)number_type_table_end)
    	RAISE("%s: inlet: unknown number type",INFO(parent));
    if (!supports_type(nt))
    	RAISE("%s: number type %s not supported here",
    		INFO(parent), number_type_table[nt].name);
    STACK_ARRAY(int32,v,argc);
    for (int i=0; i<argc; i++) v[i] = NUM2INT(argv[i]);
    P<Dim> dim = this->dim = new Dim(argc,v);
    dex=0;
    buf=0;
    int r = rb_ensure(
    	(RMethod)GridInlet_begin_1,(Ruby)this,
    	(RMethod)GridInlet_begin_2,(Ruby)this);
    if (!r) {abort(); goto hell;}
    this->dim = dim;
    back_out->callback(this);
    hell:;} // PROF
    return Qnil;
}
template <class T> void GridInlet::flow(int mode, int n, Pt<T> data) {TRACE;
    CHECK_BUSY(inlet);
    CHECK_TYPE(*data);
    CHECK_ALIGN(data);
    PROF(parent) {
    if (this->mode==0) {dex += n; return;} // ignore data
    if (n==0) return; // no data
    switch(mode) {
    case 4:{
    	int d = dex + bufi;
    	if (d+n > dim->prod()) {
    		gfpost("grid input overflow: %d of %d from [%s] to [%s]",
    			d+n, dim->prod(), INFO(sender), 0);
    		n = dim->prod() - d;
    		if (n<=0) return;
    	}
    	int bufn = factor();
    	if (buf && bufi) {
    		Pt<T> bufd = (Pt<T>)*buf;
    		int k = min(n,bufn-bufi);
    		COPY(bufd+bufi,data,k);
    		bufi+=k; data+=k; n-=k;
    		if (bufi==bufn) {
    			int newdex = dex+bufn;
    			if (this->mode==6) {
    				Pt<T> data2 = ARRAY_NEW(T,bufn);
    				COPY(data2,bufd,bufn);
    				CHECK_ALIGN(data2);
    				gh->flow(this,bufn,data2);
    			} else {
    				CHECK_ALIGN(bufd);
    				gh->flow(this,bufn,bufd);
    			}
    			dex = newdex;
    			bufi = 0;
    		}
    	}
    	int m = (n/bufn)*bufn;
    	if (m) {
    		int newdex = dex + m;
    		if (this->mode==6) {
    			Pt<T> data2 = ARRAY_NEW(T,m);
    			COPY(data2,data,m);
    			CHECK_ALIGN(data2);
    			gh->flow(this,m,data2);
    		} else {
    			gh->flow(this,m,data);
    		}
    		dex = newdex;
    	}
    	data += m;
    	n -= m;
    	if (buf && n>0) COPY((Pt<T>)*buf+bufi,data,n), bufi+=n;
    }break;
    case 6:{
    	assert(!buf);
    	int newdex = dex + n;
    	gh->flow(this,n,data);
    	if (this->mode==4) delete[] (T *)data;
    	dex = newdex;
    }break;
    case 0: break; // ignore data
    default: RAISE("%s: unknown inlet mode",INFO(parent));
    }} // PROF
}
void GridInlet::end() {TRACE;
    assert(this);
    if (!dim) RAISE("%s: inlet not busy",INFO(parent));
    if (dim->prod() != dex) {
    	gfpost("incomplete grid: %d of %d from [%s] to [%s]",
    		dex, dim->prod(), INFO(sender), INFO(parent));
    }
    PROF(parent) {
#define FOO(T) gh->flow(this,-2,Pt<T>());
    TYPESWITCH(nt,FOO,)
#undef FOO
    } // PROF
    dim=0;
    buf=0;
    dex=0;
}
template <class T> void GridInlet::from_grid2(Grid *g, T foo) {TRACE;
    nt = g->nt;
    dim = g->dim;
    int n = g->dim->prod();
    gh->flow(this,-1,Pt<T>());
    if (n>0 && this->mode!=0) {
    	Pt<T> data = (Pt<T>)*g;
    	CHECK_ALIGN(data);
    	int size = g->dim->prod();
    	if (this->mode==6) {
    		Pt<T> d = data;
    		data = ARRAY_NEW(T,size);  // problem with int64,float64 here.
    		COPY(data,d,size);
    		CHECK_ALIGN(data);
    		gh->flow(this,n,data);
    	} else {
    		int ntsz = number_type_table[nt].size;
    		int m = GridOutlet::MAX_PACKET_SIZE/*/ntsz*//factor();
    		if (!m) m++;
    		m *= factor();
    		while (n) {
    			if (m>n) m=n;
    			CHECK_ALIGN(data);
    			gh->flow(this,m,data);
    			data+=m; n-=m; dex+=m;
    		}			
    	}
    }
    gh->flow(this,-2,Pt<T>());
    //!@#$ add error handling.
    // rescue; abort(); end ???
    dim = 0;
    dex = 0;
}
void GridInlet::from_grid(Grid *g) {TRACE;
    if (!supports_type(g->nt))
    	RAISE("%s: number type %s not supported here",
    		INFO(parent), number_type_table[g->nt].name);
#define FOO(T) from_grid2(g,(T)0);
    TYPESWITCH(g->nt,FOO,)
#undef FOO
}
/* **************** GridOutlet ************************************ */
void GridOutlet::begin(int woutlet, P<Dim> dim, NumberTypeE nt) {TRACE;
    int n = dim->count();
    this->nt = nt;
    this->dim = dim;
    Ruby a[n+4];
    a[0] = INT2NUM(woutlet);
    a[1] = bsym._grid;
    a[2] = Pointer_s_noo(this);
    a[3] = INT2NUM(nt);
    for(int i=0; i<n; i++) a[4+i] = INT2NUM(dim->get(i));
    parent->send_out(COUNT(a),a);
    frozen=true;
    if (!dim->prod()) {end(); return;}
    int32 lcm_factor = 1;
    for (uint32 i=0; i<inlets.size(); i++) lcm_factor = lcm(lcm_factor,inlets[i]->factor());
    if (nt != buf->nt) {
    	// biggest packet size divisible by lcm_factor
    	int32 v = (MAX_PACKET_SIZE/lcm_factor)*lcm_factor;
    	if (v==0) v=MAX_PACKET_SIZE; // factor too big. don't have a choice.
    	buf=new Grid(new Dim(v),nt);
    }
}
// send modifies dex; send_direct doesn't
template <class T>
void GridOutlet::send_direct(int n, Pt<T> data) {TRACE;
    assert(data); assert(frozen);
    CHECK_BUSY(outlet); CHECK_TYPE(*data); CHECK_ALIGN(data);
    for (; n>0; ) {
    	int pn = min(n,MAX_PACKET_SIZE);
    	for (uint32 i=0; i<inlets.size(); i++) inlets[i]->flow(4,pn,data);
    	data+=pn, n-=pn;
    }
}
void GridOutlet::flush() {TRACE;
    if (!bufi) return;
#define FOO(T) send_direct(bufi,(Pt<T>)*buf);
    TYPESWITCH(buf->nt,FOO,)
#undef FOO
    bufi = 0;
}
template <class T, class S>
static void convert_number_type(int n, Pt<T> out, Pt<S> in) {
    for (int i=0; i<n; i++) out[i]=(T)in[i];
}
//!@#$ buffering in outlet still is 8x faster...?
//!@#$ should use BitPacking for conversion...?
// send modifies dex; send_direct doesn't
template <class T>
void GridOutlet::send(int n, Pt<T> data) {TRACE;
    assert(data); assert(frozen);
    if (!n) return;
    CHECK_BUSY(outlet); CHECK_ALIGN(data);
    if (NumberTypeE_type_of(*data)!=nt) {
    	int bs = MAX_PACKET_SIZE;
#define FOO(T) { \
    STACK_ARRAY(T,data2,bs); \
    for (;n>=bs;n-=bs,data+=bs) { \
    	convert_number_type(bs,data2,data); send(bs,data2);} \
    convert_number_type(n,data2,data); \
    send(n,data2); }
    	TYPESWITCH(nt,FOO,)
#undef FOO
    } else {
    	dex += n;
    	assert(dex <= dim->prod());
    	if (n > MIN_PACKET_SIZE || bufi + n > MAX_PACKET_SIZE) flush();
    	if (n > MIN_PACKET_SIZE) {
    		send_direct(n,data);
    	} else {
    		COPY((Pt<T>)*buf+bufi,data,n);
    		bufi += n;
    	}
    	if (dex==dim->prod()) end();
    }
}
template <class T>
void GridOutlet::give(int n, Pt<T> data) {TRACE;
    assert(data); CHECK_BUSY(outlet); assert(frozen);
    assert(dex+n <= dim->prod()); CHECK_ALIGN(data);
    if (NumberTypeE_type_of(*data)!=nt) {
    	send(n,data);
    	delete[] (T *)data;
    	return;
    }
    if (inlets.size()==1 && inlets[0]->mode == 6) {
    	// this is the copyless buffer passing
    	flush();
    	inlets[0]->flow(6,n,data);
    	dex += n;
    } else {
    	flush();
    	send_direct(n,data);
    	dex += n;
    	delete[] (T *)data;
    }
    if (dex==dim->prod()) end();
}
void GridOutlet::callback(GridInlet *in) {TRACE;
    CHECK_BUSY(outlet); assert(!frozen);
    int mode = in->mode;
    assert(mode==6 || mode==4 || mode==0);
    inlets.push_back(in);
}
\class GridObject < FObject
//!@#$ does not handle types properly
//!@#$ most possibly a big hack
template <class T>
void GridObject_r_flow(GridInlet *in, int n, Pt<T> data) {
    GridObject *self = in->parent;
    uint32 i;
    for (i=0; i<self->in.size(); i++) if (in==self->in[i].p) break;
    if (i==self->in.size()) RAISE("inlet not found?");
    if (n==-1) {
    	rb_funcall(self->rself,SI(send_in),2,INT2NUM(i),SYM(rgrid_begin));
    } else if (n>=0) {
    	Ruby buf = rb_str_new((char *)((uint8 *)data),n*sizeof(T));
    	rb_funcall(self->rself,SI(send_in),3,INT2NUM(i),SYM(rgrid_flow),buf);
    } else {
    	rb_funcall(self->rself,SI(send_in),2,INT2NUM(i),SYM(rgrid_end));
    }
}
\def Symbol inlet_nt (int inln) {
    if (inln<0 || inln>=(int)in.size()) RAISE("bad inlet number");
    P<GridInlet> inl = in[inln];
    if (!inl) RAISE("no such inlet #%d",inln);
    if (!inl->dim) return Qnil;
    return number_type_table[inl->nt].sym;
}
\def Array inlet_dim (int inln) {
    if (inln<0 || inln>=(int)in.size()) RAISE("bad inlet number");
    P<GridInlet> inl = in[inln];
    if (!inl) RAISE("no such inlet #%d",inln);
    if (!inl->dim) return Qnil;
    int n=inl->dim->count();
    Ruby a = rb_ary_new2(n);
    for(int i=0; i<n; i++) rb_ary_push(a,INT2NUM(inl->dim->v[i]));
    return a;
}
\def void inlet_set_factor (int inln, int factor) {
    if (inln<0 || inln>=(int)in.size()) RAISE("bad inlet number");
    P<GridInlet> inl = in[inln];
    if (!inl) RAISE("no such inlet #%d",inln);
    if (!inl->dim) RAISE("inlet #%d not active",inln);
    inl->set_factor(factor);
}
\def void send_out_grid_begin (int outlet, Array buf, NumberTypeE nt=int32_e) {
    if (outlet<0) RAISE("bad outlet number");
    int n = rb_ary_len(buf);
    Ruby *p = rb_ary_ptr(buf);
    STACK_ARRAY(int32,v,n);
    for (int i=0; i<n; i++) v[i] = convert(p[i],(int32*)0);
    out = new GridOutlet(this,outlet,new Dim(n,v),nt); // valgrind says leak?
}
template <class T>
void send_out_grid_flow_2(P<GridOutlet> go, Ruby s, T bogus) {
    int n = rb_str_len(s) / sizeof(T);
    Pt<T> p = rb_str_pt(s,T);
    go->send(n,p);
}
\def void send_out_grid_flow (int outlet, String buf, NumberTypeE nt=int32_e) {
    if (outlet<0) RAISE("bad outlet number");
#define FOO(T) send_out_grid_flow_2(out,argv[1],(T)0);
    TYPESWITCH(nt,FOO,)
#undef FOO
}
// install_rgrid(Integer inlet, Boolean multi_type? = true)
static Ruby GridObject_s_install_rgrid(int argc, Ruby *argv, Ruby rself) {
    if (argc<1 || argc>2) RAISE("er...");
    IEVAL(rself,"@handlers||=[]");
    Ruby handlers = rb_ivar_get(rself,SI(@handlers));
    GridHandler *gh = new GridHandler;
    bool mt = argc>1 ? argv[1]==Qtrue : 0; /* multi_type? */
    if (mt) {
#define FOO(S) gh->flow_##S = GridObject_r_flow;
EACH_NUMBER_TYPE(FOO)
#undef FOO
    } else {
#define FOO(S) gh->flow_##S = 0;
EACH_NUMBER_TYPE(FOO)
#undef FOO
    }
    gh->flow_int32 = GridObject_r_flow;
    //IEVAL(rself,"self.class_eval { def _0_grid(*a) ___grid(0,*a) end }");
    rb_funcall(handlers,SI([]=),2,INT2NUM(INT(argv[0])),PTR2FIX(gh));
    return Qnil;
}
static Ruby GridObject_s_instance_methods(int argc, Ruby *argv, Ruby rself) {
    static const char *names[] = {"grid","list","float"};
    Ruby list = rb_class_instance_methods(argc,argv,rself);
    Ruby handlers = rb_ivar_get(rself,SI(@handlers));
    if (handlers==Qnil) return list;
    for (int i=0; i<rb_ary_len(handlers); i++) {
    	Ruby ghp = rb_ary_ptr(handlers)[i];
    	if (ghp==Qnil) continue;
    	GridHandler *gh = FIX2PTR(GridHandler,ghp);
    	char buf[256];
    	for (int j=0; j<COUNT(names); j++) {
    		sprintf(buf,"_%d_%s",i,names[j]);
    		rb_ary_push(list,rb_str_new2(buf));
    	}
    }
    return list;
}
// this does auto-conversion of list/float to grid
// this also (will) do grid inputs for ruby stuff.
\def Ruby method_missing (...) {
    {
    if (argc<1) RAISE("not enough arguments");
    if (!SYMBOL_P(argv[0])) RAISE("expected symbol");
    const char *name = rb_sym_name(argv[0]);
    char *endp;
    if (*name++!='_') goto hell;
    int i = strtol(name,&endp,10);
    if (name==endp) goto hell;
    if (*endp++!='_') goto hell;
    if (strcmp(endp,"grid")==0) {
    	Ruby handlers = rb_ivar_get(rb_obj_class(rself),SI(@handlers));
    	if (TYPE(handlers)!=T_ARRAY) {
    		rb_p(handlers);
    		RAISE("gridhandler-list missing (maybe forgot install_rgrid ?)"
    		" while trying to receive on inlet %d",i);
    	}
    	if (i>=rb_ary_len(handlers)) RAISE("BORK");
    	GridHandler *gh = FIX2PTR(GridHandler, rb_ary_ptr(handlers)[i]);
    	if (in.size()<=(uint32)i) in.resize(i+1);
    	if (!in[i]) in[i]=new GridInlet((GridObject *)this,gh);
    	return in[i]->begin(argc-1,argv+1);
    }
    // we call the grid method recursively to ask it its GridInlet*
    // don't do this before checking the missing method is exactly that =)
    char foo[42];
    sprintf(foo,"_%d_grid",i);
    P<GridInlet> inl = FIX2PTR(GridInlet,rb_funcall(rself,rb_intern(foo),0));
    if (strcmp(endp,"list" )==0) return inl->from_ruby_list(argc-1,argv+1), Qnil;
    if (strcmp(endp,"float")==0) return inl->from_ruby     (argc-1,argv+1), Qnil;
    }
    hell: return rb_call_super(argc,argv);
}
\classinfo {
    IEVAL(rself,"install 'GridObject',0,0");
    // define in Ruby-metaclass
    rb_define_singleton_method(rself,"instance_methods",(RMethod)GridObject_s_instance_methods,-1);
    rb_define_singleton_method(rself,"install_rgrid",(RMethod)GridObject_s_install_rgrid,-1);
    rb_enable_super(rb_singleton_class(rself),"instance_methods");
}
\end class GridObject
Ruby cGridObject;
void startup_grid () {
    \startall
    cGridObject = rb_const_get(mGridFlow,SI(GridObject));
}
// never call this. this is a hack to make some things work.
// i'm trying to circumvent either a bug in the compiler or i don't have a clue. :-(
void make_gimmick () {
    GridOutlet foo(0,0,0);
#define FOO(S) foo.give(0,Pt<S>());
EACH_NUMBER_TYPE(FOO)
#undef FOO
}