From: TAKAHASHI Masayoshi <maki@...>
Date: 2001-03-07T02:37:16+09:00
Subject: [ruby-talk:12165] Re: NQXML: Pure Ruby XML parser (almost)

Hi,

Jim Menard <jimm@fnord.io.com> wrote:
> NQXML is a pure-Ruby XML parser that I'm writing from scratch. The "NQ"
> stands for "Not Quite" because, among other things, it does not yet support
> UTF-8 or UTF-16, ENTITY or DOCTYPE tags and the DOM parser isn't done yet.
> The tokenizer and SAX parser are working.

I wrote a patch.

 * install.rb: fix bug(? but I cannot install without this fix).
 * tokenizer.rb: ProcessingInstaruction -> ParserError
 * parserTest.rb: not bug, but can parse XML file if it uses
          DocType Decl(ignore 'not implemented' error).

It can parse UTF-8 XML file in XMLParser's archive(index_u8.xml).
# of cource, its UTF-8 support is not perfect (yet).

TAKAHASHI 'Maki' Masayoshi   E-mail: maki@open-news.com


diff -ur nqxml.org/install.rb nqxml/install.rb
--- nqxml.org/install.rb	Wed Mar  7 01:37:20 2001
+++ nqxml/install.rb	Wed Mar  7 01:38:46 2001
@@ -26,6 +26,6 @@
 
 INSTALL_DIR = instdir()
 File.makedirs(INSTALL_DIR + '/' + LIBDIR)
-Find.find(SOURCE_DIR).each { |f|
+Find.find(SOURCE_DIR){ |f|
     File.install(f, INSTALL_DIR + "/" + f, 0644, true) if f =~ /.rb$/
 }
diff -ur nqxml.org/nqxml/tokenizer.rb nqxml/nqxml/tokenizer.rb
--- nqxml.org/nqxml/tokenizer.rb	Wed Mar  7 01:37:19 2001
+++ nqxml/nqxml/tokenizer.rb	Wed Mar  7 02:00:33 2001
@@ -228,7 +228,7 @@
 	def nextEntityOrDoctype
 	    textUpTo('>', true)
 	    skipChar()		# eat '>'
-	    raise ProcessingInstruction.new('not yet implemented;' +
+	    raise ParserError.new('not yet implemented;' +
 					    ' entity skipped')
 	end
 
diff -ur nqxml.org/parseTest.rb nqxml/parseTest.rb
--- nqxml.org/parseTest.rb	Wed Mar  7 01:37:20 2001
+++ nqxml/parseTest.rb	Wed Mar  7 01:59:11 2001
@@ -15,8 +15,14 @@
     begin
 	# Don't do anything, just let parser run through the XML
 	parser.each { | type, name, data | }
-    rescue
-	puts "ERROR: $!\n"
+    rescue NQXML::ParserError
+      if /not yet implemented/ =~ $!
+	retry
+      else
+	puts "ERROR: #{$!}",$@.join("\n")
+      end
+    rescue 
+      puts "ERROR: #{$!}",$@.join("\n")
     end
     puts 'OK'
 end