/*
 * This file is a part of EZC -- Easy templating in C++ library
 * and is distributed under the BSD 3-Clause licence.
 * Author: Tomasz Sowa <t.sowa@ttmath.org>
 */

/* 
 * Copyright (c) 2007-2018, Tomasz Sowa
 * All rights reserved.
 * 
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions are met:
 * 
 *  * Redistributions of source code must retain the above copyright notice,
 *    this list of conditions and the following disclaimer.
 *    
 *  * Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 *    
 *  * Neither the name Tomasz Sowa nor the names of contributors to this
 *    project may be used to endorse or promote products derived
 *    from this software without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
 * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
 * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF
 * THE POSSIBILITY OF SUCH DAMAGE.
 */


#include "patternparser.h"
#include "convert/convert.h"



namespace Ezc
{



PatternParser::PatternParser()
{
	allow_include = true;
	pblocks       = 0;
	include_level_max       = 100;
	delete_white_text_items = false;
	program_mode = false;
	log = nullptr;
}



void PatternParser::Directory(const char * dir, const char * dir2)
{
	directory.clear();
	directory2.clear();

	if( dir )
		PT::UTF8ToWide(dir, directory);

	if( dir2 )
		PT::UTF8ToWide(dir2, directory2);
}



void PatternParser::Directory(const std::string & dir)
{
	PT::UTF8ToWide(dir, directory);
	directory2.clear();
}



void PatternParser::Directory(const std::string & dir, const std::string & dir2)
{
	PT::UTF8ToWide(dir, directory);
	PT::UTF8ToWide(dir2, directory2);
}



void PatternParser::Directory(const wchar_t * dir, const wchar_t * dir2)
{
	directory.clear();
	directory2.clear();

	if( dir )
		directory = dir;

	if( dir2 )
		directory2 = dir2;
}



void PatternParser::Directory(const std::wstring & dir)
{
	directory = dir;
	directory2.clear();
}



void PatternParser::Directory(const std::wstring & dir, const std::wstring & dir2)
{
	directory  = dir;
	directory2 = dir2;
}



void PatternParser::SetBlocks(Blocks & blocks)
{
	pblocks = &blocks;
}


void PatternParser::SetCommentary(const char * com_start, const char * com_stop)
{
	PT::UTF8ToWide(com_start, commentary_start);
	PT::UTF8ToWide(com_stop,  commentary_stop);
}



void PatternParser::SetCommentary(const std::string & com_start, const std::string & com_stop)
{
	PT::UTF8ToWide(com_start, commentary_start);
	PT::UTF8ToWide(com_stop,  commentary_stop);
}



void PatternParser::SetCommentary(const wchar_t * com_start, const wchar_t * com_stop)
{
	commentary_start = com_start;
	commentary_stop  = com_stop;
}



void PatternParser::SetCommentary(const std::wstring & com_start, const std::wstring & com_stop)
{
	commentary_start = com_start;
	commentary_stop  = com_stop;
}


void PatternParser::SetProgramMode(bool program_mode)
{
	this->program_mode = program_mode;
}


void PatternParser::SetLogger(PT::Log * log)
{
	this->log = log;
}


void PatternParser::CreateMsg(std::wstring & out, const wchar_t * type, const wchar_t * arg)
{
	out = commentary_start;
	out += L"Ezc: ";
	out += type;

	if( arg )
	{
		out += ' ';
		out += arg;
	}

	out += commentary_stop;
}



void PatternParser::ParseFile(const std::string & file_name, Pattern & pattern)
{
	ParseFile(file_name.c_str(), pattern);
}




void PatternParser::ParseFile(const char * file_name, Pattern & pattern)
{
	pat = &pattern;
	PT::UTF8ToWide(file_name, pat->item_root.file_name);
	include_level = 0;

	CreateTreeReadIncludeSkipAllowFlag(pat->item_root);
}




void PatternParser::ParseFile(const std::wstring & file_name, Pattern & pattern)
{
	ParseFile(file_name.c_str(), pattern);
}




void PatternParser::ParseFile(const wchar_t * file_name, Pattern & pattern)
{
	pat = &pattern;
	pat->item_root.file_name = file_name;
	include_level  = 0;
	CreateTreeReadIncludeSkipAllowFlag(pat->item_root);
}




void PatternParser::ParseString(const char * str, Pattern & pattern)
{
	PT::UTF8ToWide(str, string_content);
	ParseString(string_content.c_str(), pattern);
	string_content.clear();
}



void PatternParser::ParseString(const std::string & str, Pattern & pattern)
{
	ParseString(str.c_str(), pattern);
}



void PatternParser::ParseString(const wchar_t * str, Pattern & pattern)
{
	pat = &pattern;
	itext = str;
	include_level = 0;

	pat->item_root.Clear();
	CreateTreeContainer(pat->item_root);
}



void PatternParser::ParseString(const std::wstring & str, Pattern & pattern)
{
	ParseString(str.c_str(), pattern);
}









void PatternParser::AllowInclude(bool allow)
{
	allow_include = allow;
}



void PatternParser::DeleteWhiteTextItems(bool del)
{
	delete_white_text_items = del;
}



void PatternParser::SetIncludeMax(int include_max)
{
	include_level_max = include_max;
}







bool PatternParser::HasFileAtBeginning(const wchar_t * path, const wchar_t * file)
{
	for(; *path && *file; ++path, ++file)
	{
		if( *path != *file )
			return false;
	}

	if( *file != 0 )
		return false;

	// "\" is from a dos path syntax
	if( *path==0 || *path=='\\' || *path=='/' )
		return true;

return false;
}


/*
	".." is not allowed in the file path
	you cannot go up from your template directory
*/
bool PatternParser::IsFileCorrect(const wchar_t * name)
{
	while( *name )
	{
		if( HasFileAtBeginning(name, L"..") )
			return false;

		// looking for the next slash of backslash
		while( *name && *name!='\\' && *name!='/' )
			name += 1;

		// skipping the slash (or backslash)
		if( *name )
			name += 1;
	}
	
return true;
}




/*
	'name' must be a relative path - without a slash or backslash
*/
void PatternParser::ReadFile(const std::wstring & name, std::wstring & result)
{
	ReadFile(name.c_str(), result);
}


/*
	'name' must be a relative path - without a slash or backslash
*/
void PatternParser::ReadFile(const wchar_t * name, std::wstring & result)
{
	if( !IsFileCorrect(name) )
	{
		CreateMsg(result, L"incorrect file name: ", name);
	}
	else
	{
		result.clear();

		if( !ReadFileFromDir(directory, name, result) )
			if( !ReadFileFromDir(directory2, name, result) )
				CreateMsg(result, L"can't open: ", name);
	}
}



bool PatternParser::ReadFileFromDir(const std::wstring & dir, const wchar_t * name, std::wstring & result)
{
	if( dir.empty() )
		return false;

	file_name = dir;
	file_name += '/';
	file_name += name;
	PT::WideToUTF8(file_name, afile_name);

	std::ifstream file(afile_name.c_str());

	if( !file )
	{
		file_name.clear();
		afile_name.clear();
		return false;
	}

	if( log )
	{
		if( include_level <= 1 )
			(*log) << PT::Log::log4 << "Ezc: reading pattern: " << afile_name << PT::Log::logend;
		else
			(*log) << PT::Log::log4 << "     including pattern: " << afile_name << PT::Log::logend;
	}

	ReadFile(file, result);

	file_name.clear();
	afile_name.clear();

return true;
}



void PatternParser::ReadFile(std::ifstream & file, std::wstring & result)
{
	PT::UTF8ToWide(file, result);
}



int PatternParser::ReadCharInText()
{
	if( *itext==0 || *itext=='[' )
		return -1;

	if( *itext == '\\' )
	{
		if( *(itext+1)=='\\' || *(itext+1)=='['  || *(itext+1)==']' )
			++itext;
	}

return *(itext++);
}


bool PatternParser::IsWhite(wchar_t c)
{
	// 13 (\r) is from a dos file at the end of a line (\r\n)
	// 160 is a non-breaking space

	if( c==' ' || c=='\t' || c==13 || c==160 || c==10 )
		return true;

return false;
}



void PatternParser::SkipWhite()
{
	while( IsWhite(*itext) )
		++itext;
}


void PatternParser::SkipOneStatement()
{
size_t count = 1;

	while( *itext != 0 && count > 0 )
	{
		if( *itext == '[' )
			count += 1;
		else
		if( *itext == ']' )
			count -= 1;

		itext += 1;
	}
}


void PatternParser::CheckWhiteAndDelete(std::wstring & s)
{
size_t i;
	
	if( s.empty() )
		return;

	for(i=0 ; i<s.size() && IsWhite(s[i]) ; ++i);

	if( i == s.size() )
		s.clear(); // the whole string consists of white characters
}


void PatternParser::CheckFunctionIsNumber(Item::Function & function)
{
	if( IsPositiveNumber(function.name) )
		function.arg = wcstol(function.name.c_str(), 0, 10);
}


bool PatternParser::IsNameChar(wchar_t c)
{
	return ((c>='a' && c<='z') ||
			(c>='A' && c<='Z') ||
			(c>='0' && c<='9') ||
			c=='_' || c=='-' || c=='.' || c=='#');
}



bool PatternParser::IsDigit(wchar_t c)
{
	return (c>='0' && c<='9');
}



bool PatternParser::IsPositiveNumber(const std::wstring & str)
{
size_t i;

	for(i=0 ; i<str.size() ; ++i)
		if( !IsDigit(str[i]) )
			return false;

return true;
}


// reading an expression name or a function name
bool PatternParser::ReadName(std::wstring & name)
{
	name.clear();

	while( IsNameChar(*itext) )
	{
		name += *itext;
		++itext;
	}

return !name.empty();
}




// string can have a quote character (escaped with a backslash) e.g. "sample text \"with quotes\""
// use \\ to insert one backslash
bool PatternParser::ReadString(std::wstring & str)
{
	str.clear();
	SkipWhite();

	// string is signed by its first quote character (")
	if( *itext != '\"' )
		return false;

	++itext;

	while( *itext && *itext!='\"' )
	{
		if( itext[0]=='\\' && itext[1]=='\"' )
		{
			str += '\"';
			itext += 2;
		}
		else
		if( itext[0]=='\\' && itext[1]=='\\' )
		{
			str += '\\';
			itext += 2;
		}
		else
		{
			str += *itext;
			itext += 1;
		}
	}

	if( *itext == '\"' )
	{
		++itext;
		return true;
	}
	else
	{
		return false;
	}
}





bool PatternParser::ReadParamString(Item::Function & function)
{
	Item::Function & fun = function.AddNewParam();
	fun.is_function = false;

return ReadString(fun.name);
}


/*
	returns true if it correctly reads all parameters
*/
bool PatternParser::ReadParams(Item::Function & function)
{
bool res = true;
bool check_next_param = true;

	while( check_next_param && res )
	{
		SkipWhite();

		if( *itext == '[' )
		{
			res = ReadFunction(function.AddNewParam(), true);
		}
		else
		if( *itext == '\"' )
		{
			res = ReadParamString(function);
		}
		else
		if( *itext != ']' && *itext != 0 )
		{
			res = ReadFunction(function.AddNewParam(), false);
		}
		else
		{
			// *itext is equal to ']'
			check_next_param = false;
			res = true;
		}
	}

return res;
}





bool PatternParser::ReadFunction(Item::Function & function, bool with_params)
{
	bool res = false;
	SkipWhite();
	function.Clear();

	if( *itext == '[' )
	{
		itext += 1;
		res = ReadFunction(function, true);

		SkipWhite();

		if( *itext == ']' )
		{
			itext += 1;
		}
		else
		{
			SkipOneStatement();
			res = false;
		}
	}
	else
	{
		res = ReadName(function.name);

		if( res )
		{
			if( *itext == ':' )
			{
				itext += 1;
				ReadName(function.postfix); // we allow the postfix to be empty
			}

			if( with_params )
				res = ReadParams(function);
		}
	}

	if( res )
	{
		// IMPROVE ME
		// this will be called more than once for nested functions
		CheckFunctionIsNumber(function);
	}

	function.is_function = res;

	return res;
}



/*
	returns true if a function has been correctly read
*/
bool PatternParser::ReadFunction(Item & item)
{
	bool function_read_correctly = ReadFunction(item.function, true);

	if( !function_read_correctly )
	{
		item.function.Clear();
		item.type = Item::item_err;
	}

	return function_read_correctly;
}



void PatternParser::CreateTreeReadItemDirectiveCheckEnding(Item & item)
{
	SkipWhite();

	if( *itext == ']' )
	{
		itext += 1;
	}
	else
	{
		item.type = Item::item_err;
		item.function.Clear();
		SkipOneStatement();
	}
}




void PatternParser::ReadNormalStatement(Item & item)
{
	item.type          = Item::item_function;
	item.has_function  = true;

	if( !ReadFunction(item.function, true) )
	{
		item.type = Item::item_err;
		item.function.Clear();
	}
}



void PatternParser::ReadDirectiveIf(Item & item)
{
	item.type = Item::item_if;
	ReadFunction(item);
}



void PatternParser::ReadDirectiveEnd(Item & item)
{
	item.type = Item::item_end;
}



void PatternParser::ReadDirectiveElse(Item & item)
{
	item.type = Item::item_else;
}



void PatternParser::ReadDirectiveFor(Item & item)
{
	item.type = Item::item_for;
	ReadFunction(item);
}



void PatternParser::ReadDirectiveComment(Item & item)
{
	item.type = Item::item_comment;
	
	// skipping the comment
	while( *itext && *itext!=']' )
		++itext;
}



void PatternParser::ReadDirectiveInclude(Item & item)
{
	if( ReadString(item.file_name) )
		item.type = Item::item_include;
	else
		item.type = Item::item_err;
}



void PatternParser::ReadDirectiveDef(Item & item)
{
	item.type = Item::item_def;
	ReadFunction(item);
}


void PatternParser::ReadDirectiveFilter(Item & item)
{
	item.type = Item::item_filter;
	ReadFunction(item);
}


void PatternParser::ReadDirectiveEzc(Item & item)
{
	item.type = Item::item_ezc;
	ReadFunction(item);
}


/*
	[out] is a shorthand for [ezc out]
*/
void PatternParser::ReadDirectiveOut(Item & item)
{
	item.type          = Item::item_ezc;
	item.has_function  = true;

	item.function.Clear();
	item.function.name = L"out";
	item.function.is_function = true;
	
	if( !ReadParams(item.function) )
	{
		item.type = Item::item_err;
		item.function.Clear();
	}
}


void PatternParser::ReadDirectiveBlock(Item & item)
{
	item.type = Item::item_block;
	ReadFunction(item);

	// only one function without arguments
	if( !item.function.parameters.empty() )
		item.type = Item::item_err;
}


void PatternParser::ReadDirectiveReturn(Item & item)
{
	item.type = Item::item_return;
	ReadFunction(item);
}





void PatternParser::CreateTreeReadItemDirective(Item & item)
{
std::wstring name;

	++itext;
	SkipWhite();
	const wchar_t * old_itext = itext;

	ReadName(name);
	
	if     ( name == L"if" )		ReadDirectiveIf(item);
	else if( name == L"end" )		ReadDirectiveEnd(item);
	else if( name == L"else" )		ReadDirectiveElse(item);
	else if( name == L"for" )		ReadDirectiveFor(item);
	else if( name == L"include" )	ReadDirectiveInclude(item);
	else if( name == L"def" )		ReadDirectiveDef(item);
	else if( name == L"filter" )	ReadDirectiveFilter(item);
	else if( name == L"ezc" )		ReadDirectiveEzc(item);
	else if( name == L"out" )		ReadDirectiveOut(item);
	else if( name == L"block" )		ReadDirectiveBlock(item);
	else if( name == L"return" )	ReadDirectiveReturn(item);
	else if( name == L"#" )			ReadDirectiveComment(item);
	else if( *old_itext == '[' || !name.empty() )
	{
		itext = old_itext;
		ReadNormalStatement(item);
	}

	CreateTreeReadItemDirectiveCheckEnding(item);
}



void PatternParser::CreateTreeReadItemText(Item & item)
{
int c;
	
	while( (c = ReadCharInText()) != -1 )
		item.text += static_cast<wchar_t>(c);

	if( delete_white_text_items )
		CheckWhiteAndDelete(item.text);

	item.type = Item::item_text;
}


void PatternParser::CreateTreeReadDirectiveExpression(Item & item, bool is_statement)
{
	int brackets_counter = is_statement ? 1 : 0;

	while( *itext )
	{
		wchar_t c = *itext;

		if( c == 10 || c == 13 )
			c = ' ';

		if( c == ';' && brackets_counter == 0 )
		{
			itext += 1;
			return; // end of normal expression (not in a statement such as 'for' or 'if')
		}

		if( c == '(' )
		{
			brackets_counter += 1;
		}

		if( c == ')' )
		{
			brackets_counter -= 1;

			if( is_statement && brackets_counter == 0 )
			{
				itext += 1;
				return; // end of statement expression
			}
		}

		if( !IsWhite(c) || item.text.empty() || !IsWhite(item.text.back()) )
			item.text += c;

		itext += 1;
	}
}


bool PatternParser::CreateTreeCheckProgramDirective(Item & item)
{
	const wchar_t * old_itext = itext;

	if( PT::IsSubStringNoCasep(L"if", itext) )
	{
		itext += 2;
		SkipWhite();

		if( *itext == '(' )
		{
			itext += 1;
			item.type = Item::item_if;
			return true;
		}
	}

	if( PT::IsSubStringNoCasep(L"while", itext) )
	{
		itext += 5;
		SkipWhite();

		if( *itext == '(' )
		{
			itext += 1;
			item.type = Item::item_for;
			return true;
		}
	}

	itext = old_itext;
	return false;
}


bool PatternParser::CreateTreeReadExpression(Item & item)
{
	SkipWhite();

	if( *itext == 0 )
		return false;

	if( *itext == '{' )
	{
		item.type = Item::item_container;
		itext += 1;
		return true;
	}

	if( *itext == '}' )
	{
		item.type = Item::item_end;
		itext += 1;
		return true;
	}

	if( CreateTreeCheckProgramDirective(item) )
	{
		CreateTreeReadDirectiveExpression(item, true);
	}
	else
	{
		item.type = Item::item_function;
		CreateTreeReadDirectiveExpression(item, false);
	}

	return true;
}


bool PatternParser::CreateTreeReadItem(Item & item)
{
	item.Clear();

	if( program_mode )
	{
		return CreateTreeReadExpression(item);
	}
	else
	{
		if( *itext == '[' )
		{
			CreateTreeReadItemDirective(item);
			return true;
		}
		else
		if( *itext )
		{
			CreateTreeReadItemText(item);
			return true;
		}
	}

// the end of the string
return false;
}





void PatternParser::CreateTreeReadInclude(Item & item)
{
	if( allow_include )
	{
		CreateTreeReadIncludeSkipAllowFlag(item);
	}
	else
	{
		if( log )
		{
			(*log) << PT::Log::log2 << "Ezc: \"include\" directive is not allowed" << PT::Log::logend;
		}
	}
}




void PatternParser::CreateTreeReadIncludeSkipAllowFlag(Item & item)
{
	if( item.file_name.empty() )
		return;

	if( include_level > include_level_max )
	{
		if( log )
		{
			(*log) << PT::Log::log1 << "Ezc: \"include\" directive has reached the maximum level" << PT::Log::logend;
		}

		return;
	}

	++include_level;
	std::wstring file_text; // this temporary object must not be global (includes can be nested)
	ReadFile(item.file_name, file_text);
	
	const wchar_t * itext_old = itext;
	itext = file_text.c_str();

	item.Clear();
	CreateTreeContainer(item);

	itext = itext_old;
	--include_level;
}



void PatternParser::CreateTreeReadIf(Item & item)
{
	Item * pitem = item.AddItem();
	pitem->Clear();

	if( program_mode )
		CreateTree(*pitem);
	else
		CreateTreeContainer(*pitem);

	if( program_mode )
	{
		SkipWhite();

		if( PT::IsSubStringNoCasep(L"else", itext) )
		{
			itext += 4;
			pitem = item.AddItem();
			CreateTree(*pitem);
		}
	}
	else
	{
		if( pitem->LastItemType() == Item::item_else )
		{
			pitem->DeleteLastItem();

			pitem = item.AddItem();
			CreateTreeContainer(*pitem);
		}
	}

	if( pitem->LastItemType() == Item::item_end )
		pitem->DeleteLastItem();
}



void PatternParser::CreateTreeReadBlock(Item & item)
{
	Item item_block;

	CreateTreeContainer(item_block);

	if( item_block.LastItemType() == Item::item_end )
		item_block.DeleteLastItem();

	//if( pblocks && item.functions.size()==1 )
	if( pblocks )
		pblocks->Insert(item.function.name, item_block);
}



void PatternParser::CreateTreeReadFor(Item & item)
{
	Item * pitem = item.AddItem();

	if( program_mode )
		CreateTree(*pitem);
	else
		CreateTreeContainer(*pitem);

	if( pitem->LastItemType() == Item::item_end )
		pitem->DeleteLastItem();
}



bool PatternParser::CreateTree(Item & item)
{
	do
	{
		if( !CreateTreeReadItem(item) )
		{
			return false;
		}

		if( item.type == Item::item_block )
			CreateTreeReadBlock(item);
	}
	while( item.type == Item::item_comment || item.type == Item::item_block );

	// such container can be read in program mode
	if( item.type == Item::item_container )
		CreateTreeContainer(item);

	if( item.type == Item::item_if )
		CreateTreeReadIf(item);

	// CHECK ME is it correct to check item_filter and item_ezc here and call CreateTreeReadFor?
	if( item.type == Item::item_for		||
		item.type == Item::item_filter	||
		item.type == Item::item_ezc )
		CreateTreeReadFor(item);

	if( item.type == Item::item_include )
		CreateTreeReadInclude(item);

	return true;
}


void PatternParser::CreateTreeContainer(Item & item)
{
	bool item_read_correctly;
	Item * pitem;

	item.type = Item::item_container;

	do
	{
		pitem = item.AddItem();
		item_read_correctly = CreateTree(*pitem);
	}
	while( item_read_correctly && pitem->type != Item::item_end && pitem->type != Item::item_else);

	if( !item_read_correctly )
		item.DeleteLastItem();
}





} // namespace Ezc