Convert parser to C99

Rakefile

@@ -2,17 +2,50 @@
 
 require 'rubygems'
 require 'rake'
+require 'rake/clean'
+require 'fileutils'
 require 'rake/testtask'
 require 'rubygems/package_task'
 
-task :default => 'test'
+task :default => [:compile, :test]
 
 task :ragel do
-  sh "find . -name '*.rl' | xargs ragel -R -F1"
+  sh "find . -name '*.rl' | xargs ragel -C -G2"
 end
 
+task :compile => [:ragel, :liquid_ext]
 
-Rake::TestTask.new(:test) do |t|
+extension = "liquid_ext"
+ext = "ext/liquid"
+ext_so = "#{ext}/#{extension}.#{RbConfig::CONFIG['DLEXT']}"
+ext_files = FileList[
+  "#{ext}/*.c",
+  "#{ext}/*.h",
+  "#{ext}/*.rl",
+  "#{ext}/extconf.rb",
+  "#{ext}/Makefile",
+  "lib"
+]
+
+task "lib" do
+  directory "lib"
+end
+
+desc "Builds just the #{extension} extension"
+task extension.to_sym => [:ragel, "#{ext}/Makefile", ext_so ]
+
+file "#{ext}/Makefile" => ["#{ext}/extconf.rb"] do
+  Dir.chdir(ext) do ruby "extconf.rb" end
+end
+
+file ext_so => ext_files do
+  Dir.chdir(ext) do
+    sh "make"
+  end
+  cp ext_so, "lib"
+end
+
+Rake::TestTask.new(:test => [:ragel, 'liquid_ext']) do |t|
   t.libs << '.' << 'lib' << 'test'
   t.test_files = FileList['test/liquid/**/*_test.rb']
   t.verbose = false

ext/liquid/Makefile

@@ -0,0 +1,213 @@
+
+SHELL = /bin/sh
+
+# V=0 quiet, V=1 verbose.  other values don't work.
+V = 0
+Q1 = $(V:1=)
+Q = $(Q1:0=@)
+n=$(NULLCMD)
+ECHO1 = $(V:1=@$n)
+ECHO = $(ECHO1:0=@echo)
+
+#### Start of system configuration section. ####
+
+srcdir = .
+topdir = /Users/tobi/.rbenv/versions/1.9.3-p194-perf/include/ruby-1.9.1
+hdrdir = /Users/tobi/.rbenv/versions/1.9.3-p194-perf/include/ruby-1.9.1
+arch_hdrdir = /Users/tobi/.rbenv/versions/1.9.3-p194-perf/include/ruby-1.9.1/$(arch)
+VPATH = $(srcdir):$(arch_hdrdir)/ruby:$(hdrdir)/ruby
+prefix = $(DESTDIR)/Users/tobi/.rbenv/versions/1.9.3-p194-perf
+exec_prefix = $(prefix)
+rubylibprefix = $(libdir)/$(RUBY_BASE_NAME)
+bindir = $(exec_prefix)/bin
+sbindir = $(exec_prefix)/sbin
+libexecdir = $(exec_prefix)/libexec
+datarootdir = $(prefix)/share
+datadir = $(datarootdir)
+sysconfdir = $(prefix)/etc
+sharedstatedir = $(prefix)/com
+localstatedir = $(prefix)/var
+includedir = $(prefix)/include
+oldincludedir = $(DESTDIR)/usr/include
+docdir = $(datarootdir)/doc/$(PACKAGE)
+infodir = $(datarootdir)/info
+htmldir = $(docdir)
+dvidir = $(docdir)
+pdfdir = $(docdir)
+psdir = $(docdir)
+libdir = $(exec_prefix)/lib
+localedir = $(datarootdir)/locale
+mandir = $(datarootdir)/man
+ridir = $(datarootdir)/$(RI_BASE_NAME)
+sitedir = $(rubylibprefix)/site_ruby
+vendordir = $(rubylibprefix)/vendor_ruby
+rubyhdrdir = $(includedir)/$(RUBY_BASE_NAME)-$(ruby_version)
+sitehdrdir = $(rubyhdrdir)/site_ruby
+vendorhdrdir = $(rubyhdrdir)/vendor_ruby
+rubylibdir = $(rubylibprefix)/$(ruby_version)
+archdir = $(rubylibdir)/$(arch)
+sitelibdir = $(sitedir)/$(ruby_version)
+sitearchdir = $(sitelibdir)/$(sitearch)
+vendorlibdir = $(vendordir)/$(ruby_version)
+vendorarchdir = $(vendorlibdir)/$(sitearch)
+
+NULLCMD = :
+
+CC = /usr/bin/gcc-4.2
+CXX = g++
+LIBRUBY = $(LIBRUBY_A)
+LIBRUBY_A = lib$(RUBY_SO_NAME)-static.a
+LIBRUBYARG_SHARED = 
+LIBRUBYARG_STATIC = -l$(RUBY_SO_NAME)-static
+OUTFLAG = -o 
+COUTFLAG = -o 
+
+RUBY_EXTCONF_H = 
+cflags   =  $(optflags) $(debugflags) $(warnflags)
+optflags = -O3
+debugflags = -ggdb
+warnflags = -Wextra -Wno-unused-parameter -Wno-parentheses -Wno-long-long -Wno-missing-field-initializers -Wpointer-arith -Wwrite-strings -Wdeclaration-after-statement -Wshorten-64-to-32 -Wimplicit-function-declaration
+CFLAGS   = -fno-common $(cflags)  -pipe $(ARCH_FLAG)
+INCFLAGS = -I. -I$(arch_hdrdir) -I$(hdrdir)/ruby/backward -I$(hdrdir) -I$(srcdir)
+DEFS     = 
+CPPFLAGS =  -I'/Users/tobi/.rbenv/versions/1.9.3-p194-perf/include'  -D_XOPEN_SOURCE -D_DARWIN_C_SOURCE $(DEFS) $(cppflags)
+CXXFLAGS = $(CFLAGS) $(cxxflags)
+ldflags  = -L. -L'/Users/tobi/.rbenv/versions/1.9.3-p194-perf/lib'  -L/usr/local/lib
+dldflags = -Wl,-undefined,dynamic_lookup -Wl,-multiply_defined,suppress -Wl,-flat_namespace
+ARCH_FLAG = 
+DLDFLAGS = $(ldflags) $(dldflags) $(ARCH_FLAG)
+LDSHARED = $(CC) -dynamic -bundle
+LDSHAREDXX = $(CXX) -dynamic -bundle
+AR = ar
+EXEEXT = 
+
+RUBY_BASE_NAME = ruby
+RUBY_INSTALL_NAME = ruby
+RUBY_SO_NAME = ruby
+arch = x86_64-darwin12.0.0
+sitearch = $(arch)
+ruby_version = 1.9.1
+ruby = /Users/tobi/.rbenv/versions/1.9.3-p194-perf/bin/ruby
+RUBY = $(ruby)
+RM = rm -f
+RM_RF = $(RUBY) -run -e rm -- -rf
+RMDIRS = rmdir -p
+MAKEDIRS = mkdir -p
+INSTALL = /usr/bin/install -c
+INSTALL_PROG = $(INSTALL) -m 0755
+INSTALL_DATA = $(INSTALL) -m 644
+COPY = cp
+
+#### End of system configuration section. ####
+
+preload = 
+
+libpath = . $(libdir)
+LIBPATH =  -L. -L$(libdir)
+DEFFILE = 
+
+CLEANFILES = mkmf.log
+DISTCLEANFILES = 
+DISTCLEANDIRS = 
+
+extout = 
+extout_prefix = 
+target_prefix = 
+LOCAL_LIBS = 
+LIBS =  -lc  -lpthread -ldl -lobjc 
+SRCS = liquid_ext.c
+OBJS = liquid_ext.o
+TARGET = liquid_ext
+DLLIB = $(TARGET).bundle
+EXTSTATIC = 
+STATIC_LIB = 
+
+BINDIR        = $(bindir)
+RUBYCOMMONDIR = $(sitedir)$(target_prefix)
+RUBYLIBDIR    = $(sitelibdir)$(target_prefix)
+RUBYARCHDIR   = $(sitearchdir)$(target_prefix)
+HDRDIR        = $(rubyhdrdir)/ruby$(target_prefix)
+ARCHHDRDIR    = $(rubyhdrdir)/$(arch)/ruby$(target_prefix)
+
+TARGET_SO     = $(DLLIB)
+CLEANLIBS     = $(TARGET).bundle 
+CLEANOBJS     = *.o  *.bak
+
+all:    $(DLLIB)
+static: $(STATIC_LIB)
+.PHONY: all install static install-so install-rb
+.PHONY: clean clean-so clean-rb
+
+clean-rb-default::
+clean-rb::
+clean-so::
+clean: clean-so clean-rb-default clean-rb
+		@-$(RM) $(CLEANLIBS) $(CLEANOBJS) $(CLEANFILES)
+
+distclean-rb-default::
+distclean-rb::
+distclean-so::
+distclean: clean distclean-so distclean-rb-default distclean-rb
+		@-$(RM) Makefile $(RUBY_EXTCONF_H) conftest.* mkmf.log
+		@-$(RM) core ruby$(EXEEXT) *~ $(DISTCLEANFILES)
+		@-$(RMDIRS) $(DISTCLEANDIRS) 2> /dev/null || true
+
+realclean: distclean
+install: install-so install-rb
+
+install-so: $(RUBYARCHDIR)
+install-so: $(RUBYARCHDIR)/$(DLLIB)
+$(RUBYARCHDIR)/$(DLLIB): $(DLLIB)
+	@-$(MAKEDIRS) $(@D)
+	$(INSTALL_PROG) $(DLLIB) $(@D)
+install-rb: pre-install-rb install-rb-default
+install-rb-default: pre-install-rb-default
+pre-install-rb: Makefile
+pre-install-rb-default: Makefile
+pre-install-rb-default:
+	$(ECHO) installing default liquid_ext libraries
+$(RUBYARCHDIR):
+	$(Q) $(MAKEDIRS) $@
+
+site-install: site-install-so site-install-rb
+site-install-so: install-so
+site-install-rb: install-rb
+
+.SUFFIXES: .c .m .cc .mm .cxx .cpp .C .o
+
+.cc.o:
+	$(ECHO) compiling $(<)
+	$(Q) $(CXX) $(INCFLAGS) $(CPPFLAGS) $(CXXFLAGS) $(COUTFLAG)$@ -c $<
+
+.mm.o:
+	$(ECHO) compiling $(<)
+	$(Q) $(CXX) $(INCFLAGS) $(CPPFLAGS) $(CXXFLAGS) $(COUTFLAG)$@ -c $<
+
+.cxx.o:
+	$(ECHO) compiling $(<)
+	$(Q) $(CXX) $(INCFLAGS) $(CPPFLAGS) $(CXXFLAGS) $(COUTFLAG)$@ -c $<
+
+.cpp.o:
+	$(ECHO) compiling $(<)
+	$(Q) $(CXX) $(INCFLAGS) $(CPPFLAGS) $(CXXFLAGS) $(COUTFLAG)$@ -c $<
+
+.C.o:
+	$(ECHO) compiling $(<)
+	$(Q) $(CXX) $(INCFLAGS) $(CPPFLAGS) $(CXXFLAGS) $(COUTFLAG)$@ -c $<
+
+.c.o:
+	$(ECHO) compiling $(<)
+	$(Q) $(CC) $(INCFLAGS) $(CPPFLAGS) $(CFLAGS) $(COUTFLAG)$@ -c $<
+
+.m.o:
+	$(ECHO) compiling $(<)
+	$(Q) $(CC) $(INCFLAGS) $(CPPFLAGS) $(CFLAGS) $(COUTFLAG)$@ -c $<
+
+$(DLLIB): $(OBJS) Makefile
+	$(ECHO) linking shared-object $(DLLIB)
+	@-$(RM) $(@)
+	$(Q) $(LDSHARED) -o $@ $(OBJS) $(LIBPATH) $(DLDFLAGS) $(LOCAL_LIBS) $(LIBS)
+
+
+
+$(OBJS): $(hdrdir)/ruby.h $(hdrdir)/ruby/defines.h $(arch_hdrdir)/ruby/config.h

ext/liquid/extconf.rb

@@ -0,0 +1,6 @@
+require 'mkmf'
+
+dir_config("liquid_ext")
+have_library("c", "main")
+
+create_makefile("liquid_ext")

ext/liquid/liquid_ext.rl

@@ -0,0 +1,159 @@
+/*
+  Parser for context#[] method. Generated through ragel from parser.rl 
+  Only modify parser.rl. Run rake ragel afterwards to generate this file. 
+*/
+
+#include <ruby.h>
+
+%%{
+  machine fsm;
+
+  action mark {
+    mark = p;
+  }
+
+  action lookup {
+    EMIT("lookup", Qnil)
+  }
+
+  action call {
+    EMIT("call", Qnil)
+  }
+  action range {
+    EMIT("range", Qnil)
+  }
+
+  constants = ( "true" | "false" | "nil" | "null" );
+
+  # strings 
+  string = "\"" any* "\"" | "'" any* "'";
+
+  # nothingness
+  nil = "nil" | "null" ;
+
+  # numbers
+  integer = ('+'|'-')? digit+;
+  float = ('+'|'-')? digit+ '.' digit+;
+
+  # simple values
+  primitive = (
+
+    integer       >mark %{ 
+      EMIT("id", rb_funcall(rb_cObject, rb_intern("Integer"), 1, rb_str_new(mark, p - mark))); 
+    } |
+
+    float         >mark %{ 
+      EMIT("id", rb_funcall(rb_cObject, rb_intern("Float"), 1, rb_str_new(mark, p - mark))) 
+    } |
+
+    nil           %{ EMIT("id", Qnil) } | 
+    "true"        %{ EMIT("id", Qtrue) } |
+    "false"       %{ EMIT("id", Qfalse) } |
+    
+    string        >mark %{ EMIT("id", rb_str_new(mark + 1, p - mark - 2)) } 
+
+  );
+
+  entity = (
+    ((alpha [A-Za-z0-9_\-]*) - (constants)) >mark %{ 
+      EMIT("id", rb_str_new(mark, p - mark))
+      EMIT("lookup", Qnil) 
+    }
+  );
+
+  # Because of recursion we cannot immediatly resolve the content of this in 
+  # the current grammar. We simply re-invoke the parser here to descend into 
+  # the substring
+  recur = (
+    (any+ - ']') >mark %{
+      VALUE body = rb_str_new(mark, p - mark);
+      liquid_context_parse_impl(body, tokens);
+    }
+  );
+
+  expr = (
+    entity |
+    primitive |    
+    "(" (primitive | entity) ".." (primitive | entity) <: ")"  %range |
+    "[" recur "]" %lookup    
+  );
+
+  hash_accessors  = (
+    "[" recur "]" %call |
+
+    ".first"  %{ 
+      EMIT("buildin", rb_str_new2("first"))
+    } |
+
+    ".last"  %{ 
+      EMIT("buildin", rb_str_new2("last"))
+    } |
+
+    ".size"  %{ 
+      EMIT("buildin", rb_str_new2("size"))      
+    } |
+
+    "." ((alpha [A-Za-z0-9_\-]*) - ("first"|"last"|"size"))  >mark  %{ 
+      EMIT("id", rb_str_new(mark, p - mark))
+      EMIT("call", Qnil) 
+    }
+  );
+
+  main := ( 
+    
+    expr <: (hash_accessors)*
+
+  );
+    
+}%%
+
+%% write data nofinal;
+
+// def self.emit(sym, data, tokens) 
+//   puts "emitting: #{sym} -> #{data.inspect}" if $VERBOSE
+//   tokens.push [sym, data]
+// end
+
+#define EMIT(sym, data) rb_ary_push(tokens, rb_ary_new3(2, ID2SYM(rb_intern(sym)), data)); 
+
+
+void liquid_context_parse_impl(VALUE text, VALUE tokens)
+{
+  char *p;
+  char *pe;
+  char *eof;
+  char *mark;
+  int cs, res = 0;
+
+  if (RSTRING_LEN(text) <= 0) {
+    return;
+  }
+  
+  mark = p = RSTRING_PTR(text); 
+  eof = pe = RSTRING_PTR(text) + RSTRING_LEN(text);    
+
+  %% write init;
+  %% write exec;
+}
+
+VALUE liquid_context_parse(VALUE self, VALUE text) {
+  VALUE tokens;
+
+  //printf("text: %s\n", RSTRING_PTR(text));
+  
+  //Check_Type(text, T_STRING);
+
+  tokens = rb_ary_new();
+  liquid_context_parse_impl(text, tokens);
+  return tokens;
+}
+
+static VALUE rb_Liquid;
+static VALUE rb_Parser;
+
+void Init_liquid_ext()
+{
+  rb_Liquid = rb_define_module("Liquid");
+  rb_Parser = rb_define_class_under(rb_Liquid, "Parser", rb_cObject);
+  rb_define_singleton_method(rb_Parser, "parse", liquid_context_parse, 1);  
+}

ext/liquid/mkmf.log

@@ -0,0 +1,22 @@
+have_library: checking for main() in -lc... -------------------- yes
+
+"/usr/bin/gcc-4.2 -o conftest -I/Users/tobi/.rbenv/versions/1.9.3-p194-perf/include/ruby-1.9.1/x86_64-darwin12.0.0 -I/Users/tobi/.rbenv/versions/1.9.3-p194-perf/include/ruby-1.9.1/ruby/backward -I/Users/tobi/.rbenv/versions/1.9.3-p194-perf/include/ruby-1.9.1 -I. -I'/Users/tobi/.rbenv/versions/1.9.3-p194-perf/include'  -D_XOPEN_SOURCE -D_DARWIN_C_SOURCE  -I'/Users/tobi/.rbenv/versions/1.9.3-p194-perf/include'   -O3 -ggdb -Wextra -Wno-unused-parameter -Wno-parentheses -Wno-long-long -Wno-missing-field-initializers -Wpointer-arith -Wwrite-strings -Wdeclaration-after-statement -Wshorten-64-to-32 -Wimplicit-function-declaration  -pipe conftest.c  -L. -L/Users/tobi/.rbenv/versions/1.9.3-p194-perf/lib -L. -L'/Users/tobi/.rbenv/versions/1.9.3-p194-perf/lib'  -L/usr/local/lib     -lruby-static  -lpthread -ldl -lobjc "
+checked program was:
+/* begin */
+1: #include "ruby.h"
+2: 
+3: int main() {return 0;}
+/* end */
+
+"/usr/bin/gcc-4.2 -o conftest -I/Users/tobi/.rbenv/versions/1.9.3-p194-perf/include/ruby-1.9.1/x86_64-darwin12.0.0 -I/Users/tobi/.rbenv/versions/1.9.3-p194-perf/include/ruby-1.9.1/ruby/backward -I/Users/tobi/.rbenv/versions/1.9.3-p194-perf/include/ruby-1.9.1 -I. -I'/Users/tobi/.rbenv/versions/1.9.3-p194-perf/include'  -D_XOPEN_SOURCE -D_DARWIN_C_SOURCE  -I'/Users/tobi/.rbenv/versions/1.9.3-p194-perf/include'   -O3 -ggdb -Wextra -Wno-unused-parameter -Wno-parentheses -Wno-long-long -Wno-missing-field-initializers -Wpointer-arith -Wwrite-strings -Wdeclaration-after-statement -Wshorten-64-to-32 -Wimplicit-function-declaration  -pipe conftest.c  -L. -L/Users/tobi/.rbenv/versions/1.9.3-p194-perf/lib -L. -L'/Users/tobi/.rbenv/versions/1.9.3-p194-perf/lib'  -L/usr/local/lib     -lruby-static -lc  -lpthread -ldl -lobjc "
+checked program was:
+/* begin */
+1: #include "ruby.h"
+2: 
+3: /*top*/
+4: int main() {return 0;}
+5: int t() { void ((*volatile p)()); p = (void ((*)()))main; return 0; }
+/* end */
+
+--------------------
+

lib/liquid.rb

@@ -62,7 +62,7 @@ require 'liquid/standardfilters'
 require 'liquid/condition'
 require 'liquid/module_ex'
 require 'liquid/utils'
-require 'liquid/parser'
+require 'liquid_ext'
 
 # Load all the tags of the standard library
 #

lib/liquid/parser.rl

@@ -1,126 +0,0 @@
-# Parser for context#[] method. Generated through ragel from parser.rl 
-# Only modify parser.rl. Run rake ragel afterwards to generate this file. 
-#
-#VERBOSE=true
-
-%%{
-  machine fsm;
-
-  action mark {
-    mark = p
-  }
-
-  action lookup {
-    emit(:lookup, :instruction, nil, tokens)
-  }
-
-  action call {
-    emit(:call, :instruction, nil, tokens)
-  }
-  action range {
-    emit(:range, :instruction, nil, tokens)
-  }
-
-  constants = ( "true" | "false" | "nil" | "null" );
-
-  # strings 
-  string = "\"" any* "\"" | "'" any* "'";
-
-  # nothingness
-  nil = "nil" | "null" ;
-
-  # numbers
-  integer = ('+'|'-')? digit+;
-  float = ('+'|'-')? digit+ '.' digit+;
-
-  # simple values
-  primitive = (
-
-    integer       >mark %{ emit(:id, :integer, Integer(data[mark..p-1]), tokens) } |
-
-    float         >mark %{ emit(:id, :float, Float(data[mark..p-1]), tokens) } |
-
-    nil           %{ emit(:id, :nil, nil, tokens) } | 
-    "true"        %{ emit(:id, :bool, true, tokens) } |
-    "false"       %{ emit(:id, :bool, false, tokens)} |
-    
-    string        >mark %{ emit(:id, :string, data[mark+1..p-2], tokens) } 
-
-  );
-
-  entity = (
-    ((alpha [A-Za-z0-9_\-]*) - (constants)) >mark %{ 
-      emit(:id, :label, data[mark..p-1], tokens) 
-      emit(:lookup, :variable, nil, tokens) 
-    }
-  );
-
-  # Because of recursion we cannot immediatly resolve the content of this in 
-  # the current grammar. We simply re-invoke the parser here to descend into 
-  # the substring
-  recur = (
-    (any+ - ']') >mark %{       
-      self.parse(data[mark..p-1], tokens)
-    }
-  );
-
-  expr = (
-    entity |
-    primitive |    
-    "(" (primitive | entity) ".." (primitive | entity) <: ")"  %range |
-    "[" recur "]" %lookup    
-  );
-
-  hash_accessors  = (
-    "[" recur "]" %call |
-
-    ".first"  %{ 
-      emit(:buildin, :symbol, "first", tokens)       
-    } |
-
-    ".last"  %{ 
-      emit(:buildin, :symbol, "last", tokens)       
-    } |
-
-    ".size"  %{ 
-      emit(:buildin, :symbol, "size", tokens)       
-    } |
-
-    "." ((alpha [A-Za-z0-9_\-]*) - ("first"|"last"|"size"))  >mark  %{ 
-      emit(:id, :label, data[mark..p-1], tokens) 
-      emit(:call, :variable, nil, tokens) 
-    }
-  );
-
-  main := ( 
-    
-    expr <: (hash_accessors)*
-
-  );
-    
-}%%
-# % fix syntax highlighting
-
-
-module Liquid
-  module Parser
-    %% write data;
-
-    def self.emit(sym, type, data, tokens) 
-      puts "emitting: #{type} #{sym} -> #{data.inspect}" if $VERBOSE
-      tokens.push [sym, data]
-    end
-
-    def self.parse(data, tokens = [])      
-      puts "--> self.parse with #{data.inspect}, #{tokens.inspect}" if $VERBOSE
-
-      eof = data.length  
-
-      %% write init;
-      %% write exec;   
-
-      puts "<-- #{tokens.inspect}" if $VERBOSE
-      return tokens 
-    end
-  end
-end