fast_xs 0.6 → 0.7

data/README

@@ -7,12 +7,10 @@
     http://intertwingly.net/stories/2005/09/28/xchar.rb
     http://intertwingly.net/blog/2005/09/28/XML-Cleansing
 
-  _why also packages a slightly modified version in the latest version
-  of Hpricot (not-yet-released) that also escapes '"' to """ (which
-  is optional).
+  _why also packages an older version with Hpricot (patch in progress).
+  It should be compatible with the latest version of this code.
 
-  This is an exact translation (to the best of my knowledge :) of Sam's
-  original implementation, so it does not escape """.  The version
-  of XML::Builder packaged with Rails also uses Sam's version, and
+  This is an almost exact translation (to the best of my knowledge :) of
+  Sam's original implementation, but it does escape """.
   XML::Builder as packaged in Rails 2.0 will be automatically use
   fast_xs if available.

data/Rakefile

@@ -13,16 +13,16 @@ begin
   rev = Time.at(rev.split("\n")[1].to_i).strftime('%Y%m%d.%H%M%S')
 rescue
 end
-version ||= ENV['VERSION'] || '0.6' + (rev && rev.length > 0 ? ".#{rev}" : '')
+version ||= ENV['VERSION'] || '0.7' + (rev && rev.length > 0 ? ".#{rev}" : '')
 pkg = "#{name}-#{version}"
 bin = "*.{so,o}"
 archlib = "lib/#{::Config::CONFIG['arch']}"
-CLEAN.include ["ext/fast_xs/#{bin}", "lib/**/#{bin}",
-               'ext/fast_xs/Makefile', '**/.*.sw?', '*.gem', '.config',
+CLEAN.include ["ext/**/#{bin}", "lib/**/#{bin}",
+               'ext/**/Makefile', '**/.*.sw?', '*.gem', '.config',
                'pkg']
 rdoc_opts = ['--quiet', '--title', 'fast_xs notes', '--main', 'README',
              '--inline-source']
-pkg_files = %w(CHANGELOG COPYING README Rakefile) +
+pkg_files = %w(COPYING README Rakefile) +
             Dir.glob("{test,lib}/**/*.rb") +
             Dir.glob("ext/**/*.{c,rb,h}")
 
@@ -33,12 +33,12 @@ spec = Gem::Specification.new do |s|
   s.platform = Gem::Platform::RUBY
   s.has_rdoc = true
   s.rdoc_options += rdoc_opts
-  s.extra_rdoc_files = ["README", "CHANGELOG", "COPYING"]
+  s.extra_rdoc_files = ["README", "COPYING"]
   s.summary = "excessively fast escaping"
   s.description = s.summary
   s.author = "Eric Wong"
   s.email = 'normalperson@yhbt.net'
-  s.homepage = 'http://bogonips.org/fast_xs/'
+  s.homepage = 'http://bogomips.org/fast_xs/'
   s.files = pkg_files
   s.require_paths = [archlib, "lib"]
   s.extensions = FileList["ext/**/extconf.rb"].to_a
@@ -74,7 +74,7 @@ Rake::RDocTask.new do |rdoc|
   rdoc.rdoc_dir = 'doc/rdoc'
   rdoc.options += rdoc_opts
   rdoc.main = "README"
-  rdoc.rdoc_files.add ['README', 'CHANGELOG', 'COPYING' ]
+  rdoc.rdoc_files.add ['README', 'COPYING' ]
 end
 
 Rake::GemPackageTask.new(spec) do |p|
@@ -82,7 +82,7 @@ Rake::GemPackageTask.new(spec) do |p|
   p.gem_spec = spec
 end
 
-['fast_xs'].each do |extension|
+%w(fast_xs fast_xs_extra).each do |extension|
   ext = "ext/#{extension}"
   ext_so = "#{ext}/#{extension}.#{Config::CONFIG['DLEXT']}"
   ext_files = FileList[
@@ -114,7 +114,7 @@ task "lib" do
 end
 
 desc "Compiles the Ruby extension"
-task :compile => [:fast_xs] do
+task :compile => %w(fast_xs fast_xs_extra) do
   if Dir.glob(File.join(archlib,"fast_xs.*")).empty?
     STDERR.puts 'Gem failed to build'
     exit(1)
@@ -123,9 +123,9 @@ end
 
 task :install do
   sh %{rake package}
-  sh %{sudo gem install pkg/#{name}-#{version}}
+  sh %{gem install pkg/#{name}-#{version}}
 end
 
 task :uninstall => [:clean] do
-  sh %{sudo gem uninstall #{name}}
+  sh %{gem uninstall #{name}}
 end

data/ext/fast_xs/fast_xs.c

@@ -1,7 +1,5 @@
 #include <ruby.h>
 #include <assert.h>
-#include <sys/time.h>
-#include <sys/resource.h>
 #include "ruby_1_9_compat.h"
 #include "fast_xs_type.h"
 #include "gcc.h"
@@ -82,6 +80,8 @@ static size_t escape(char *buf, int n)
 	/* handle ASCII first */
 	if (likely(n < 128)) {
 		if (likely(n >= 0x20 || n == '\t' || n == '\n' || n == '\r')) {
+			if (unlikely(n == '"'))
+				return_const_len("&quot;");
 			if (unlikely(n == '&'))
 				return_const_len("&amp;");
 			if (unlikely(n == '<'))
@@ -144,6 +144,8 @@ static VALUE fast_xs(VALUE self)
 	     tmp++) {
 		int n = NUM2INT(*tmp);
 		if (likely(n < 128)) {
+			if (unlikely(n == '"'))
+				s_len += (sizeof("&quot;") - 2);
 			if (unlikely(n == '&'))
 				s_len += (sizeof("&amp;") - 2);
 			if (unlikely(n == '>' || n == '<'))
@@ -166,158 +168,6 @@ static VALUE fast_xs(VALUE self)
 	return rv;
 }
 
-/*
- * This is coding agnostic, and works on each byte, so some multibyte
- * character sets may not be fully supported (but UTF-8 should be).
- * This is meant to be 100% compatible with the
- * ERB::Util::escape_html and CGI::escapeHTML methods
- */
-static VALUE fast_xs_html(VALUE self)
-{
-	long i;
-	char *s;
-	size_t new_len = RSTRING_LEN(self);
-	char *new_str;
-	VALUE rv;
-
-	for (s = RSTRING_PTR(self), i = RSTRING_LEN(self); --i >= 0; ++s) {
-		if (unlikely(*s == '&'))
-			new_len += (sizeof("&amp;") - 2);
-		else if (unlikely(*s == '<' || *s == '>'))
-			new_len += (sizeof("&gt;") - 2);
-		else if (unlikely(*s == '"'))
-			new_len += (sizeof("&quot;") - 2);
-	}
-
-	rv = rb_str_new(NULL, new_len);
-	new_str = RSTRING_PTR(rv);
-
-#define append_const(buf, x) do { \
-	buf = memcpy(buf, x, sizeof(x) - 1) + sizeof(x) - 1; \
-} while (0)
-
-	for (s = RSTRING_PTR(self), i = RSTRING_LEN(self); --i >= 0; ++s) {
-		if (unlikely(*s == '&'))
-			append_const(new_str, "&amp;");
-		else if (unlikely(*s == '<'))
-			append_const(new_str, "&lt;");
-		else if (unlikely(*s == '>'))
-			append_const(new_str, "&gt;");
-		else if (unlikely(*s == '"'))
-			append_const(new_str, "&quot;");
-		else
-			*new_str++ = *s;
-	}
-
-#undef append_const
-
-	return rv;
-}
-
-#define CGI_URI_OK(x) \
-	((x >= 'a' && x <= 'z') || \
-	 (x >= 'A' && x <= 'Z') || \
-	 (x >= '0' && x <= '9') || \
-	 (x == '.' || x == '-' || x == '_'))
-
-static inline VALUE _xs_uri_encode(VALUE self, const unsigned int space_to_plus)
-{
-	long i;
-	char *s;
-	size_t new_len = RSTRING_LEN(self);
-	char *new_str;
-	VALUE rv;
-
-	for (s = RSTRING_PTR(self), i = RSTRING_LEN(self); --i >= 0; ++s) {
-		if (likely(CGI_URI_OK(*s)) || (space_to_plus && *s == ' '))
-			continue;
-		new_len += 2;
-	}
-
-	rv = rb_str_new(NULL, new_len);
-	new_str = RSTRING_PTR(rv);
-
-	for (s = RSTRING_PTR(self), i = RSTRING_LEN(self); --i >= 0; ++s) {
-		if (likely(CGI_URI_OK(*s)))
-			*new_str++ = *s;
-		else if (space_to_plus && *s == ' ')
-			*new_str++ = '+';
-		else {
-			static const char cgi_digitmap[] = "0123456789ABCDEF";
-			new_str[2] = cgi_digitmap[(*s % 16)];
-			new_str[1] = cgi_digitmap[((*s/16) % 16)];
-			new_str[0] = '%';
-			new_str += 3;
-		}
-	}
-
-	return rv;
-}
-
-/*
- * Compatible with ERB::Util::url_encode / ERB::Util::u, this iterates
- * through each byte, so multibyte character sets may not supported (but
- * UTF-8 should be).
- */
-static VALUE fast_xs_url(VALUE self)
-{
-	return _xs_uri_encode(self, 0);
-}
-
-/*
- * Compatible with CGI::escape(), this iterates through each byte, so
- * multibyte character sets may not supported (but UTF-8 should be).
- */
-static VALUE fast_xs_cgi(VALUE self)
-{
-	return _xs_uri_encode(self, 1);
-}
-
-static VALUE _uxs_uri(VALUE self, const unsigned int plus_to_space)
-{
-	char *s, *new_str;
-	long i;
-	size_t new_len = RSTRING_LEN(self);
-	VALUE rv;
-
-	for (s = RSTRING_PTR(self), i = RSTRING_LEN(self);
-	     --i >= 0;
-	     ++s) {
-		if (unlikely(*s == '%') &&
-		    likely(is_hex(s[1])) &&
-		    likely(is_hex(s[2]))) {
-			new_len -= 2;
-			s += 2;
-			i -= 2;
-		}
-	}
-
-	rv = rb_str_new(NULL, new_len);
-	new_str = RSTRING_PTR(rv);
-
-	for (s = RSTRING_PTR(self), i = RSTRING_LEN(self);
-	     --i >= 0;
-	     ++s, ++new_str) {
-		if (plus_to_space && unlikely(*s == '+'))
-			*new_str = ' ';
-		else if (unlikely(*s == '%') &&
-		         likely(is_hex(s[1])) &&
-			 likely(is_hex(s[2]))) {
-			*new_str = hexpair_to_int(s[1], s[2]);
-			s += 2;
-			i -= 2;
-		} else
-			*new_str = *s;
-	}
-
-	return rv;
-}
-
-static VALUE fast_uxs_cgi(VALUE self)
-{
-	return _uxs_uri(self, 1);
-}
-
 void Init_fast_xs(void)
 {
 	assert(cp_1252[159 - 128] == 376); /* just in case I skipped a line */
@@ -329,8 +179,4 @@ void Init_fast_xs(void)
 	rb_global_variable(&C_fmt);
 
 	rb_define_method(rb_cString, "fast_xs", fast_xs, 0);
-	rb_define_method(rb_cString, "fast_xs_html", fast_xs_html, 0);
-	rb_define_method(rb_cString, "fast_xs_cgi", fast_xs_cgi, 0);
-	rb_define_method(rb_cString, "fast_uxs_cgi", fast_uxs_cgi, 0);
-	rb_define_method(rb_cString, "fast_xs_url", fast_xs_url, 0);
 }

data/ext/fast_xs_extra/extconf.rb

@@ -0,0 +1,5 @@
+require 'mkmf'
+have_header('assert.h') or exit
+find_header('gcc.h', '../fast_xs')
+dir_config('fast_xs_extra')
+create_makefile('fast_xs_extra')

data/ext/fast_xs_extra/fast_xs_extra.c

@@ -0,0 +1,165 @@
+#include <ruby.h>
+#include <assert.h>
+#include "ruby_1_9_compat.h"
+#include "fast_xs_type.h"
+#include "gcc.h"
+
+/*
+ * This is coding agnostic, and works on each byte, so some multibyte
+ * character sets may not be fully supported (but UTF-8 should be).
+ * This is meant to be 100% compatible with the
+ * ERB::Util::escape_html and CGI::escapeHTML methods
+ */
+static VALUE fast_xs_html(VALUE self)
+{
+	long i;
+	char *s;
+	size_t new_len = RSTRING_LEN(self);
+	char *new_str;
+	VALUE rv;
+
+	for (s = RSTRING_PTR(self), i = RSTRING_LEN(self); --i >= 0; ++s) {
+		if (unlikely(*s == '&'))
+			new_len += (sizeof("&amp;") - 2);
+		else if (unlikely(*s == '<' || *s == '>'))
+			new_len += (sizeof("&gt;") - 2);
+		else if (unlikely(*s == '"'))
+			new_len += (sizeof("&quot;") - 2);
+	}
+
+	rv = rb_str_new(NULL, new_len);
+	new_str = RSTRING_PTR(rv);
+
+#define append_const(buf, x) do { \
+	buf = memcpy(buf, x, sizeof(x) - 1) + sizeof(x) - 1; \
+} while (0)
+
+	for (s = RSTRING_PTR(self), i = RSTRING_LEN(self); --i >= 0; ++s) {
+		if (unlikely(*s == '&'))
+			append_const(new_str, "&amp;");
+		else if (unlikely(*s == '<'))
+			append_const(new_str, "&lt;");
+		else if (unlikely(*s == '>'))
+			append_const(new_str, "&gt;");
+		else if (unlikely(*s == '"'))
+			append_const(new_str, "&quot;");
+		else
+			*new_str++ = *s;
+	}
+
+#undef append_const
+
+	return rv;
+}
+
+#define CGI_URI_OK(x) \
+	((x >= 'a' && x <= 'z') || \
+	 (x >= 'A' && x <= 'Z') || \
+	 (x >= '0' && x <= '9') || \
+	 (x == '.' || x == '-' || x == '_'))
+
+static inline VALUE _xs_uri_encode(VALUE self, const unsigned int space_to_plus)
+{
+	long i;
+	char *s;
+	size_t new_len = RSTRING_LEN(self);
+	char *new_str;
+	VALUE rv;
+
+	for (s = RSTRING_PTR(self), i = RSTRING_LEN(self); --i >= 0; ++s) {
+		if (likely(CGI_URI_OK(*s)) || (space_to_plus && *s == ' '))
+			continue;
+		new_len += 2;
+	}
+
+	rv = rb_str_new(NULL, new_len);
+	new_str = RSTRING_PTR(rv);
+
+	for (s = RSTRING_PTR(self), i = RSTRING_LEN(self); --i >= 0; ++s) {
+		if (likely(CGI_URI_OK(*s)))
+			*new_str++ = *s;
+		else if (space_to_plus && *s == ' ')
+			*new_str++ = '+';
+		else {
+			static const char cgi_digitmap[] = "0123456789ABCDEF";
+			new_str[2] = cgi_digitmap[(*s % 16)];
+			new_str[1] = cgi_digitmap[((*s/16) % 16)];
+			new_str[0] = '%';
+			new_str += 3;
+		}
+	}
+
+	return rv;
+}
+
+/*
+ * Compatible with ERB::Util::url_encode / ERB::Util::u, this iterates
+ * through each byte, so multibyte character sets may not supported (but
+ * UTF-8 should be).
+ */
+static VALUE fast_xs_url(VALUE self)
+{
+	return _xs_uri_encode(self, 0);
+}
+
+/*
+ * Compatible with CGI::escape(), this iterates through each byte, so
+ * multibyte character sets may not supported (but UTF-8 should be).
+ */
+static VALUE fast_xs_cgi(VALUE self)
+{
+	return _xs_uri_encode(self, 1);
+}
+
+static VALUE _uxs_uri(VALUE self, const unsigned int plus_to_space)
+{
+	char *s, *new_str;
+	long i;
+	size_t new_len = RSTRING_LEN(self);
+	VALUE rv;
+
+	for (s = RSTRING_PTR(self), i = RSTRING_LEN(self);
+	     --i >= 0;
+	     ++s) {
+		if (unlikely(*s == '%') &&
+		    likely(is_hex(s[1])) &&
+		    likely(is_hex(s[2]))) {
+			new_len -= 2;
+			s += 2;
+			i -= 2;
+		}
+	}
+
+	rv = rb_str_new(NULL, new_len);
+	new_str = RSTRING_PTR(rv);
+
+	for (s = RSTRING_PTR(self), i = RSTRING_LEN(self);
+	     --i >= 0;
+	     ++s, ++new_str) {
+		if (plus_to_space && unlikely(*s == '+'))
+			*new_str = ' ';
+		else if (unlikely(*s == '%') &&
+		         likely(is_hex(s[1])) &&
+			 likely(is_hex(s[2]))) {
+			*new_str = hexpair_to_int(s[1], s[2]);
+			s += 2;
+			i -= 2;
+		} else
+			*new_str = *s;
+	}
+
+	return rv;
+}
+
+static VALUE fast_uxs_cgi(VALUE self)
+{
+	return _uxs_uri(self, 1);
+}
+
+void Init_fast_xs_extra(void)
+{
+	rb_define_method(rb_cString, "fast_xs_html", fast_xs_html, 0);
+	rb_define_method(rb_cString, "fast_xs_cgi", fast_xs_cgi, 0);
+	rb_define_method(rb_cString, "fast_uxs_cgi", fast_uxs_cgi, 0);
+	rb_define_method(rb_cString, "fast_xs_url", fast_xs_url, 0);
+}

data/lib/fast_xs_monkey_patcher.rb

@@ -1,4 +1,5 @@
 require 'fast_xs'
+require 'fast_xs_extra'
 
 class CGI
   def CGI::escapeHTML(value); value.fast_xs_html; end

data/test/test_xml_escaping.rb

@@ -13,6 +13,7 @@ class TestXmlEscaping < Test::Unit::TestCase
     assert_equal '&amp;', '&'.fast_xs              # ampersand
     assert_equal '&lt;',  '<'.fast_xs              # left angle bracket
     assert_equal '&gt;',  '>'.fast_xs              # right angle bracket
+    assert_equal '&quot;', '"'.fast_xs             # double quote
   end
 
   def test_invalid

metadata

@@ -1,43 +1,29 @@
 --- !ruby/object:Gem::Specification 
-rubygems_version: 0.9.4.7
-specification_version: 2
 name: fast_xs
 version: !ruby/object:Gem::Version 
-  version: "0.6"
-date: 2008-02-20 00:00:00 -08:00
-summary: excessively fast escaping
-require_paths: 
-- lib/i486-linux
-- lib
-email: normalperson@yhbt.net
-homepage: http://bogonips.org/fast_xs/
-rubyforge_project: 
-description: excessively fast escaping
+  version: "0.7"
+platform: ruby
+authors: 
+- Eric Wong
 autorequire: 
-default_executable: 
 bindir: bin
-has_rdoc: true
-required_ruby_version: !ruby/object:Gem::Requirement 
-  requirements: 
-  - - ">="
-    - !ruby/object:Gem::Version 
-      version: "0"
-  version: 
-required_rubygems_version: !ruby/object:Gem::Requirement 
-  requirements: 
-  - - ">="
-    - !ruby/object:Gem::Version 
-      version: "0"
-  version: 
-platform: ruby
-signing_key: 
 cert_chain: []
 
-post_install_message: 
-authors: 
-- Eric Wong
+date: 2009-02-07 00:00:00 -08:00
+default_executable: 
+dependencies: []
+
+description: excessively fast escaping
+email: normalperson@yhbt.net
+executables: []
+
+extensions: 
+- ext/fast_xs/extconf.rb
+- ext/fast_xs_extra/extconf.rb
+extra_rdoc_files: 
+- README
+- COPYING
 files: 
-- CHANGELOG
 - COPYING
 - README
 - Rakefile
@@ -46,12 +32,15 @@ files:
 - test/test_erb_util_module_overrides.rb
 - lib/fast_xs_monkey_patcher.rb
 - ext/fast_xs/fast_xs.c
+- ext/fast_xs_extra/fast_xs_extra.c
 - ext/fast_xs/extconf.rb
+- ext/fast_xs_extra/extconf.rb
+- ext/fast_xs/gcc.h
 - ext/fast_xs/ruby_1_9_compat.h
 - ext/fast_xs/fast_xs_type.h
-- ext/fast_xs/gcc.h
-test_files: []
-
+has_rdoc: true
+homepage: http://bogomips.org/fast_xs/
+post_install_message: 
 rdoc_options: 
 - --quiet
 - --title
@@ -59,15 +48,27 @@ rdoc_options:
 - --main
 - README
 - --inline-source
-extra_rdoc_files: 
-- README
-- CHANGELOG
-- COPYING
-executables: []
-
-extensions: 
-- ext/fast_xs/extconf.rb
+require_paths: 
+- lib/x86_64-linux
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement 
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      version: "0"
+  version: 
+required_rubygems_version: !ruby/object:Gem::Requirement 
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      version: "0"
+  version: 
 requirements: []
 
-dependencies: []
+rubyforge_project: 
+rubygems_version: 1.3.1
+signing_key: 
+specification_version: 2
+summary: excessively fast escaping
+test_files: []
 

data/CHANGELOG

@@ -1,8 +0,0 @@
-= 0.2
-=== 6th December, 2007
-* First real release
-* added fast_xs_html and fast_xs_cgi methods
-
-= 0.1
-=== 3rd October, 2007
-* initial implementation, not officially released