bbcodelib 1.0.0

data/ext/bbcode_parser.h

@@ -0,0 +1,159 @@
+/*!
+
+	Abstract:
+		this file is a part of bbcode library
+
+	Author:
+		Igor Franchuk (sprog@online.ru)
+
+	Last Update:
+		$Id: bbcode_parser.h,v 1.1 2007/12/19 19:13:30 lanthruster Exp $
+		Version: 0.01
+*/
+
+#ifndef BBCODE_parseR_H
+#define BBCODE_parseR_H
+#pragma once
+
+#include <string>
+#include <iostream>
+#include <istream>
+#include <stdexcept>
+#include <stack>
+#include <vector>
+#include <map>
+
+#include "bbcode_config.h"
+#include "bbcode_lexer.h"
+#include "bbcode_utils.h"
+
+#pragma warning(disable: 4290) //VS doesn't support exception specifications
+
+namespace{
+
+	using std::string;
+	using std::istream;
+	using std::cin;
+	using std::out_of_range;
+	using std::runtime_error;
+	using std::stack;
+	using std::vector;
+	using std::map;
+	using std::make_pair;
+}
+
+namespace bbcode{
+
+//--------------------------------
+// tag struct
+	struct tag{
+		enum tag_type{opening, closing};
+		typedef map<string, string> attribs;		
+
+		string name;		
+		tag_type type;
+		size_t left_bracket_pos;
+		size_t right_bracket_pos;
+		attribs attr;
+		bool has_closing_tag;
+
+		void clear(void){name = "", type = opening, left_bracket_pos = right_bracket_pos = 0; attr.clear();has_closing_tag = true;}
+		template<class T>
+			bool operator ==(T _t){if(name == _t.name && type == _t.type && left_bracket_pos == _t.left_bracket_pos && 
+									  right_bracket_pos == _t.right_bracket_pos && attr == _t.attr && has_closing_tag == _t.has_closing_tag) return true;
+								   return false;
+								  }
+
+		tag(void):
+			name(""),
+			type(opening),
+			left_bracket_pos(0),
+			right_bracket_pos(0),
+			has_closing_tag(true)
+			{}
+	};
+//--------------------------------
+// this is multipart form data parser class
+
+	class parser{
+
+		typedef vector<tag> tags;
+		typedef map<string, string> supported_tag_storage;
+
+		parser_config _pconf; /* we keep all our configuration in this structure */
+		lexer  _lexer;		  /* lexer, the source of lexems */
+		conds  _current_state;/* parser's current state, defined by cond table */
+		stack<lexeme> _lexems;
+		string _parsed_data; /* used to keep immediate parsed data */	
+		lexeme l;
+		string _im_data;
+
+		supported_tag_storage _supported_tag_storage; //preloaded tags that we support in BBcode library
+		
+		tags  _parsed_tags;
+		tag   _tmp_tag;
+
+		void _bbcode_replacements(void);	//here we're replacing bbcode tags
+		string _tag_replacement(tag& t);	//replace bb tag with HTML tag
+		tags _discarded_tags;//an internal unclosed|overlapped|sentencedtodleletion tag storage, required for U|O recovery
+		void _discarded_tags_cleanup(void);
+
+		string _js_guard(string str); //this is a simple javascript: insert protection
+
+	public:		
+		enum flag{store_data_in_memory=0, store_data_in_file=1};
+
+		//tag schem loader
+
+		void load_supported_tag_schema(const string& schema);
+
+		string current_tag_schema(void) const;
+
+ 		/* class configuration block */
+
+		size_t setf(size_t mode){	size_t old_conf = _pconf._config; _pconf._config |= mode; return old_conf;	}
+
+		void   source_stream(istream& source_stream){	_pconf._source_stream = &source_stream;	}
+		istream*  source_stream(void) const{	return _pconf._source_stream; }
+
+		size_t read_buffer_size(size_t buffer_size)			 throw(out_of_range);
+
+		size_t read_buffer_size( void )	  const	   {	return _pconf._read_buffer_size;	}
+
+		parser_config pconf(parser_config& pconf){ parser_config tconf = _pconf; _pconf = pconf; return tconf;}
+		parser_config pconf(void) const { return _pconf;}	
+		
+
+		/* end of class configuration block */
+
+		/* class functinality */
+		
+		void parse(void);
+		/* return alternative multipart form content */
+		string content(void) const {	return _parsed_data;	}
+
+		/* class configuration block */
+
+		/* constructors */
+
+		parser(void) : _pconf(cin, _DEFAULT_BUFFER_SIZE),
+					   _current_state(C1),
+					   l(TERM),
+					   _im_data(""),
+					   _parsed_data("")
+						{load_supported_tag_schema(_SUPPORTED_BBTAGS);}
+
+		parser(const string& schema) : _pconf(cin, _DEFAULT_BUFFER_SIZE),
+					   _current_state(C1),
+					   l(TERM),
+					   _im_data(""),
+					   _parsed_data("")
+						{load_supported_tag_schema(schema);}
+		/* end of constructors */
+
+		virtual ~parser(){};
+	};
+
+}
+
+#endif

data/ext/bbcode_utils.cpp

@@ -0,0 +1,143 @@
+#include "bbcode_utils.h"
+
+namespace bbcode{
+
+	//--------------------------------
+	// upper case block
+	string upper_case(string s){		
+		transform(s.begin(), s.end(), s.begin(), capitalize);
+		return s;
+	}
+	unsigned char capitalize(unsigned char c){		
+		return toupper(c);
+	}
+	//--------------------------------
+	// form_parcer, helper functions
+
+	string get_module_msg(const string& str){
+			return _THIS_MODULE_STR_IDENT + " " + str;
+	}
+
+	//--------------------------------
+	// form_parcer, helper functions
+	string random_string(size_t length){
+		string res;
+		string dictionary = _RND_GENERATOR_DICTIONARY;
+		if(dictionary.empty()) throw runtime_error(get_module_msg("Empty _RND_GENERATOR_DICTIONARY"));
+		for(size_t i = 0; i < length; i++){
+			random_shuffle(dictionary.begin(), dictionary.end());
+			//res.push_back(dictionary[0]);
+			res.append(1, dictionary[0]);
+		}
+		return res;
+	}
+
+	// ------------------------------------------------------------------------------------------------------------------
+	//! Templates specializations
+	// ------------------------------------------------------------------------------------------------------------------
+
+	// ------------------------------------------------------------------------------------------------------------------
+	//! string to string spec (two arg)
+	template<> string stream_cast(const string &value, cast_manip manip)
+	{
+		return value;
+	}
+
+	// ------------------------------------------------------------------------------------------------------------------
+	//! char* to string spec (two arg)
+	template<> string stream_cast(const char* value, cast_manip manip)
+	{
+		return string(value);
+	}
+
+	// ------------------------------------------------------------------------------------------------------------------
+	//! char* to string spec (one arg)
+	template<> string stream_cast(const string &value)
+	{
+		return value;
+	}
+
+	// ------------------------------------------------------------------------------------------------------------------
+	//! char* to string spec (one arg)
+	template<> string stream_cast(const char* value)
+	{
+		return string(value);
+	}
+
+	// ------------------------------------------------------------------------------------------------------------------
+	//! Split string
+	strings split_string(const string &str, const string &delimiter, const bool include_empty)
+	{
+		strings result;
+		string sub = "";
+		std::size_t dpos = 0, lastdpos = 0, dsize = delimiter.size();
+		do {
+			dpos = str.find(delimiter, lastdpos);
+			sub = str.substr(lastdpos, dpos - lastdpos);
+			if (!sub.empty() || include_empty) result.push_back(sub);
+			lastdpos = dpos + dsize;
+		} while (dpos != string::npos);
+		return result;
+	}
+
+	// ------------------------------------------------------------------------------------------------------------------
+	//! Splits string so that length of any substring is no more than max_sub_length
+	strings split_string(const string &str, const string &delimiter, bool include_empty, size_t max_sub_length)
+	{
+		strings result = split_string(str, delimiter, include_empty);
+		if (max_sub_length > 0) {
+			strings new_res;
+			for (strings::iterator it = result.begin(); it != result.end(); it++) {
+				size_t len = it->length();
+				if (len > max_sub_length) {
+					size_t count = len / max_sub_length + 1;
+					strings subsplit;
+					for (size_t i = 0; i < count; i++) {
+						subsplit.push_back(it->substr(i * max_sub_length, max_sub_length));
+					}
+					new_res.insert(new_res.end(), subsplit.begin(), subsplit.end());
+				}
+				else {
+					new_res.push_back(*it);
+				}
+			}
+			return new_res;
+		}
+		return result;
+	}
+	//----------------------------------------------------------
+	//! functor for find_if algorithm
+	//! trim_left, trim_right
+	class trim_findfunct {
+		const string _char_list;
+	public:
+		trim_findfunct(const string char_list):_char_list(char_list)  { }
+		bool operator () ( const char ch) const {
+			return string::npos == _char_list.find(ch);
+		}
+	};
+	//----------------------------------------------------------
+	//! erase head spaces
+	string trim_left( const string &str, const string &char_list ) {
+		string ret = "";
+		string::const_iterator it = find_if(str.begin(), str.end(), trim_findfunct(char_list) );
+		copy( it, str.end(), back_inserter(ret) );
+		return ret;
+	}
+
+	//----------------------------------------------------------
+	//! erase trail spaces
+	string trim_right( const string &str, const string &char_list ) {
+		string ret = "";
+		string::const_reverse_iterator it = find_if(str.rbegin(), str.rend(), trim_findfunct(char_list) );
+		copy(str.begin(), str.begin() + (str.rend() - it) , back_inserter(ret));
+		return ret;
+	}
+
+	//----------------------------------------------------------
+	//! erase head and trail spaces
+	string trim( const string &str, const string &char_list ) {
+		return trim_left(trim_right(str,char_list),char_list);
+	}	
+
+}

data/ext/bbcode_utils.h

@@ -0,0 +1,144 @@
+/*!
+	Abstract:
+		this file is a part of bbcode library
+
+	Author:
+		Igor Franchuk (sprog@online.ru)
+
+	Last Update:
+		$Id: bbcode_utils.h,v 1.1 2007/12/19 19:13:30 lanthruster Exp $
+		Version: 0.01
+*/
+
+#ifndef BBCODE_UTILS_H
+#define BBCODE_UTILS_H
+
+#include <string>
+#include <vector>
+#include <sstream>
+#include <istream>
+#include <iostream>
+#include <algorithm>
+#include <functional>
+#include <time.h>
+#include <stdexcept>
+#include "bbcode_config.h"
+
+namespace{
+	using std::string;
+	using std::istream;
+	using std::transform;
+	using std::uppercase;
+	using std::cin;
+	using std::cout;
+	using std::stringstream;
+	using std::random_shuffle;
+	using std::runtime_error;	
+	using std::vector;
+	using std::find_if;
+}
+
+namespace bbcode{	
+	typedef vector<string> strings;
+//--------------------------------
+// helper functions
+	unsigned char capitalize(unsigned char c);
+	string upper_case(string s); /* only for Latin-1, locale is not used */
+	string random_string(size_t length); /* returns random string sequence */
+
+//--------------------------------
+// helper templates
+
+		template<class T, class C>
+		T myswap(T& p1, C& p2){
+			T old = p1;
+			p1 = p2;
+			return old;
+		};
+
+	//! Type T to Q std stream conversion
+	enum cast_manip { cm_none, cm_dec, cm_hex, cm_oct, cm_boolalpha };
+
+	//! Cast implementation
+	template<class T, class Q>
+		inline T stream_cast_impl(const Q &value, cast_manip manip)
+	{
+		std::stringstream buf;
+		T res;
+		switch (manip)
+		{
+		case cm_dec: buf << std::dec; break;
+		case cm_hex: buf << std::hex; break;
+		case cm_oct: buf << std::oct; break;
+		case cm_boolalpha: buf << std::boolalpha; break;
+		}
+		buf << value;
+		buf >> res;
+		return res;
+	}
+
+	//! For cast with two arguments
+	template<class T, class Q>
+		T stream_cast(const Q &value, cast_manip manip)
+	{
+		return stream_cast_impl<T>(value, manip);
+	}
+
+	//! char* spec (two arg)
+	template<class T> T stream_cast(const char* value, cast_manip manip)
+	{
+		return stream_cast_impl<T>(string(value), manip);
+	}
+	//! For one argument casting
+	template<class T, class Q> T stream_cast(const Q &value)
+	{
+		return stream_cast_impl<T>(value, cm_none);
+	}
+	//! char* spec (one arg)
+	template<class T> T stream_cast(const char* value)
+	{
+		return stream_cast_impl<T>(string(value));
+	}
+
+	//! size_t hack for WIN32
+#ifdef WIN32
+	//! size_t spec (one arg)
+	template<class T> T stream_cast(const size_t value)
+	{
+		return stream_cast_impl<T>(static_cast<unsigned int>(value), cm_none);
+	}
+	//! size_t spec (one arg)
+	template<class T> T stream_cast(const size_t value, cast_manip manip)
+	{
+		return stream_cast_impl<T>(static_cast<unsigned int>(value), manip);
+	}
+#endif
+
+	//! string to string spec (two arg)
+	template<> string stream_cast(const string &value, cast_manip manip);
+	//! string to string spec (one arg)
+	template<> string stream_cast(const string &value);
+	//! char* to string spec (two arg)
+	template<> string stream_cast(const char* value, cast_manip manip);
+	//! char* to string spec (one arg)
+	template<> string stream_cast(const char* value);
+
+//--------------------------------
+// helper functions
+
+	string get_module_msg(const string& str);
+
+	//! Splits string
+	strings split_string(const string &str, const string &delimiter, const bool include_empty = true);
+	//! Splits string so that length of any substring is no more than max_sub_length
+	strings split_string(const string &str, const string &delimiter, bool include_empty, size_t max_sub_length);
+	//! erase head spaces
+	string trim_left( const string &str, const string &char_list=" \t\n\r\0\x0B");
+	//! erase trail spaces
+	string trim_right( const string &str, const string &char_list=" \t\n\r\0\x0B" );
+	//! erase head and trail spaces
+	string trim( const string &str, const string &char_list=" \t\n\r\0\x0B" );
+
+
+}
+#endif 

data/ext/extconf.rb

@@ -0,0 +1,13 @@
+# Loads mkmf which is used to make makefiles for Ruby extensions
+require 'mkmf'
+
+# Give it a name
+extension_name = 'bbcode'
+
+# The destination
+dir_config(extension_name)
+
+have_library("stdc++")
+
+# Do the work
+create_makefile(extension_name)

data/lib/bbcodelib.rb

@@ -0,0 +1,6 @@
+$:.unshift(File.dirname(__FILE__)) unless
+  $:.include?(File.dirname(__FILE__)) || $:.include?(File.expand_path(File.dirname(__FILE__)))
+
+module Bbcodelib
+  VERSION = '1.0.0'
+end

data/script/console

@@ -0,0 +1,10 @@
+#!/usr/bin/env ruby
+# File: script/console
+irb = RUBY_PLATFORM =~ /(:?mswin|mingw)/ ? 'irb.bat' : 'irb'
+
+libs =  " -r irb/completion"
+# Perhaps use a console_lib to store any extra methods I may want available in the cosole
+# libs << " -r #{File.dirname(__FILE__) + '/../lib/console_lib/console_logger.rb'}"
+libs <<  " -r #{File.dirname(__FILE__) + '/../lib/bbcodelib.rb'}"
+puts "Loading bbcodelib gem"
+exec "#{irb} #{libs} --simple-prompt"

data/script/destroy

@@ -0,0 +1,14 @@
+#!/usr/bin/env ruby
+APP_ROOT = File.expand_path(File.join(File.dirname(__FILE__), '..'))
+
+begin
+  require 'rubigen'
+rescue LoadError
+  require 'rubygems'
+  require 'rubigen'
+end
+require 'rubigen/scripts/destroy'
+
+ARGV.shift if ['--help', '-h'].include?(ARGV[0])
+RubiGen::Base.use_component_sources! [:rubygems, :newgem, :newgem_theme, :test_unit]
+RubiGen::Scripts::Destroy.new.run(ARGV)

data/script/generate

@@ -0,0 +1,14 @@
+#!/usr/bin/env ruby
+APP_ROOT = File.expand_path(File.join(File.dirname(__FILE__), '..'))
+
+begin
+  require 'rubigen'
+rescue LoadError
+  require 'rubygems'
+  require 'rubigen'
+end
+require 'rubigen/scripts/generate'
+
+ARGV.shift if ['--help', '-h'].include?(ARGV[0])
+RubiGen::Base.use_component_sources! [:rubygems, :newgem, :newgem_theme, :test_unit]
+RubiGen::Scripts::Generate.new.run(ARGV)

data/test/test_bbcodelib.rb

@@ -0,0 +1,11 @@
+require File.dirname(__FILE__) + '/test_helper.rb'
+
+class TestBbcodelib < Test::Unit::TestCase
+
+  def setup
+  end
+  
+  def test_truth
+    assert true
+  end
+end

data/test/test_helper.rb

@@ -0,0 +1,3 @@
+require 'stringio'
+require 'test/unit'
+require File.dirname(__FILE__) + '/../lib/bbcodelib'

metadata

@@ -0,0 +1,89 @@
+--- !ruby/object:Gem::Specification 
+name: bbcodelib
+version: !ruby/object:Gem::Version 
+  version: 1.0.0
+platform: ruby
+authors: 
+- Stephen Blackstone
+autorequire: 
+bindir: bin
+cert_chain: []
+
+date: 2009-09-29 00:00:00 -04:00
+default_executable: 
+dependencies: 
+- !ruby/object:Gem::Dependency 
+  name: hoe
+  type: :development
+  version_requirement: 
+  version_requirements: !ruby/object:Gem::Requirement 
+    requirements: 
+    - - ">="
+      - !ruby/object:Gem::Version 
+        version: 2.3.3
+    version: 
+description: Provides a fast, efficent BBcode parser for ruby via the bbcodelib package at sourceforge
+email: 
+- sblackstone@gmail.com
+executables: []
+
+extensions: 
+- ext/extconf.rb
+extra_rdoc_files: 
+- History.txt
+- Manifest.txt
+files: 
+- History.txt
+- Manifest.txt
+- README.rdoc
+- Rakefile
+- bbcodelib.gemspec
+- ext/Makefile
+- ext/bbcode.cpp
+- ext/bbcode_config.h
+- ext/bbcode_lexer.cpp
+- ext/bbcode_lexer.h
+- ext/bbcode_parser.cpp
+- ext/bbcode_parser.h
+- ext/bbcode_utils.cpp
+- ext/bbcode_utils.h
+- ext/extconf.rb
+- lib/bbcodelib.rb
+- script/console
+- script/destroy
+- script/generate
+- test/test_bbcodelib.rb
+- test/test_helper.rb
+has_rdoc: true
+homepage: http://rubyforge.org/projects/bbcodelib/
+licenses: []
+
+post_install_message: PostInstall.txt
+rdoc_options: 
+- --main
+- README.rdoc
+require_paths: 
+- lib
+- ext
+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: []
+
+rubyforge_project: bbcodelib
+rubygems_version: 1.3.5
+signing_key: 
+specification_version: 3
+summary: A ruby binding for the bbcodelib library (sourceforge)
+test_files: 
+- test/test_bbcodelib.rb
+- test/test_helper.rb