Changeset 164

Timestamp:

04/09/2008 22:32:36 (5 months ago)

Author:

stepheneb

Message:

ragel 6.1 fixes for hpricot_scan.java.rl; update jruby object access

Applied the updates to the ragel code for 6.1 to hpricot_scan.java.rl.

Updated decrated method used to access IRubyObject object:

new IRubyObject[]{runtime.newSymbol("buffer_size")}

becomes:

runtime.getCurrentContext(),new IRubyObject[]{runtime.newSymbol("buffer_size")}

The include statements in the main ragel code didn't
work with paths starting with ext/ so I made them
into relative path references which did work.

Updates to the Rakefile to support easy
compilation of the C ragel source code
using different ragel compilation strategies.
This is the default setting:

DEFAULT_RAGEL_C_CODE_GENERATION = "really_fast goto_driven"

which is -G2

The result is no test failures in C and only
one in Java: test_latin1_entities.

Location:

trunk

Files:

• Rakefile (modified) (4 diffs)
• ext/hpricot_scan/hpricot_scan.java.rl (modified) (9 diffs)
• ext/hpricot_scan/hpricot_scan.rl (modified) (1 diff)

trunk/Rakefile

@@ -22 +22 @@
       Dir.glob("{bin,doc,test,lib,extras}/**/*") + 
       Dir.glob("ext/**/*.{h,java,c,rb,rl}") + 
-      %w[ext/hpricot_scan/hpricot_scan.c ext/hpricot_scan/HpricotScanService.java] # needed because it's generated later
+      %w[ext/hpricot_scan/hpricot_scan.c ext/hpricot_scan/HpricotScanService.java] # needed because they are generated later
+RAGEL_C_CODE_GENERATION_STYLES = {
+  "table_driven" => 'T0',
+  "faster_table_driven" => 'T1',
+  "flat_table_driven" => 'F0',
+  "faster_flat_table_driven" => 'F1',
+  "goto_driven" => 'G0',
+  "faster_goto_driven" => 'G1',
+  "really_fast goto_driven" => 'G2'
+  # "n_way_split_really_fast_goto_driven" => 'P<N>'
+}
+DEFAULT_RAGEL_C_CODE_GENERATION = "really_fast goto_driven"
 SPEC =
   Gem::Specification.new do |s|
@@ -114 +125 @@
 task :hpricot_scan => [:ragel]
 
-desc "returns the ragel version"
+desc "Determines the Ragel version and displays it on the console along with the location of the Ragel binary."
 task :ragel_version do
   @ragel_v = `ragel -v`[/(version )(\S*)/,2].to_f
+  puts "Using ragel version: #{@ragel_v}, location: #{`which ragel`}"
+  @ragel_v
 end
 
@@ -122 +135 @@
 task :ragel => [:ragel_version] do
   if @ragel_v >= 6.1
-    sh %{ragel ext/hpricot_scan/hpricot_scan.rl -G2 -o ext/hpricot_scan/hpricot_scan.c}
+    @ragel_c_code_generation_style = RAGEL_C_CODE_GENERATION_STYLES[DEFAULT_RAGEL_C_CODE_GENERATION]
+    sh %{ragel ext/hpricot_scan/hpricot_scan.rl -#{@ragel_c_code_generation_style} -o ext/hpricot_scan/hpricot_scan.c}
   else
     STDERR.puts "Ragel 6.1 or greater is required."
@@ -129 +143 @@
 end
 
-desc "Generates the Java scanner code with Ragel."
+# Java only supports the table-driven code 
+# generation style at this point. 
+desc "Generates the Java scanner code using the Ragel table-driven code generation style."
 task :ragel_java => [:ragel_version] do
   if @ragel_v >= 6.1
-    sh %{ragel -J ext/hpricot_scan/hpricot_scan.java.rl -G2 -o ext/hpricot_scan/HpricotScanService.java}
+    puts "compiling with ragel version #{@ragel_v}"
+    sh %{ragel -J -o ext/hpricot_scan/HpricotScanService.java ext/hpricot_scan/hpricot_scan.java.rl}    
   else
     STDERR.puts "Ragel 6.1 or greater is required."

trunk/ext/hpricot_scan/hpricot_scan.java.rl

@@ -21 +21 @@
 
        public void ELE(IRubyObject N) {
-         if (tokend > tokstart || text) {
+         if (te > ts || text) {
            IRubyObject raw_string = runtime.getNil();
            ele_open = false; text = false;
-           if (tokstart != -1 && N != cdata && N != sym_text && N != procins && N != comment) { 
-             raw_string = runtime.newString(new String(buf,tokstart,tokend-tokstart));
+           if (ts != -1 && N != cdata && N != sym_text && N != procins && N != comment) { 
+             raw_string = runtime.newString(new String(buf,ts,te-ts));
            } 
            rb_yield_tokens(N, tag[0], attr, raw_string, taint);
@@ -79 +79 @@
              mark = mark_aval;
            }
-           if(mark > tokstart) {
+           if(mark > ts) {
              if(N == tag) {
-               mark_tag  -= tokstart;
+               mark_tag  -= ts;
              } else if(N == akey) {
-               mark_akey -= tokstart;
+               mark_akey -= ts;
              } else if(N == aval) {
-               mark_aval -= tokstart;
+               mark_aval -= ts;
              }
            }
@@ -95 +95 @@
              attr = RubyHash.newHash(runtime);
            }
-           ((RubyHash)attr).aset(K,V);
+           ((RubyHash)attr).op_aset(runtime.getCurrentContext(),K,V);
+           // ((RubyHash)attr).aset(K,V);
          }
        }
@@ -115 +116 @@
            if(ele_open) { 
              ele_open = false; 
-             if(tokstart > -1) { 
-               mark_tag = tokstart; 
+             if(ts > -1) { 
+               mark_tag = ts; 
              } 
            } else {
@@ -184 +185 @@
   }
 
-  include hpricot_common "ext/hpricot_scan/hpricot_common.rl";
+  include hpricot_common "hpricot_common.rl";
 
 }%%
@@ -210 +211 @@
 int cs, act, have = 0, nread = 0, curline = 1, p=-1;
 boolean text = false;
-int tokstart=-1, tokend;
+int ts=-1, te;
+int eof=-1;
 char[] buf;
 Ruby runtime;
@@ -300 +302 @@
     if ( done && ele_open ) {
       ele_open = false;
-      if(tokstart > -1) {
-        mark_tag = tokstart;
-        tokstart = -1;
+      if(ts > -1) {
+        mark_tag = ts;
+        ts = -1;
         text = true;
       }
     }
 
-    if(tokstart == -1) {
+    if(ts == -1) {
       have = 0;
-      /* text nodes have no tokstart because each byte is parsed alone */
+      /* text nodes have no ts because each byte is parsed alone */
       if(mark_tag != -1 && text) {
         if (done) {
@@ -322 +324 @@
       mark_tag = 0;
     } else {
-      have = pe - tokstart;
-      System.arraycopy(buf,tokstart,buf,0,have);
+      have = pe - ts;
+      System.arraycopy(buf,ts,buf,0,have);
       SLIDE(tag);
       SLIDE(akey);
       SLIDE(aval);
-      tokend = (tokend - tokstart);
-      tokstart = 0;
+      te = (te - ts);
+      ts = 0;
     }
   }
@@ -359 +361 @@
 public static void Init_hpricot_scan(Ruby runtime) {
   RubyModule mHpricot = runtime.defineModule("Hpricot");
-  mHpricot.getMetaClass().attr_accessor(new IRubyObject[]{runtime.newSymbol("buffer_size")});
+  mHpricot.getMetaClass().attr_accessor(runtime.getCurrentContext(),new IRubyObject[]{runtime.newSymbol("buffer_size")});
   CallbackFactory fact = runtime.callbackFactory(HpricotScanService.class);
   mHpricot.getMetaClass().defineMethod("scan",fact.getSingletonMethod("__hpricot_scan",IRubyObject.class));

trunk/ext/hpricot_scan/hpricot_scan.rl

@@ -109 +109 @@
   }
 
-  include hpricot_common "ext/hpricot_scan/hpricot_common.rl";
+  include hpricot_common "hpricot_common.rl";
 
 }%%