// tstring.cpp
//
// This file is a part of Unicode NSIS.
//
// Copyright (C) 2007-2009 Jim Park
//
// Licensed under the zlib/libpng license (the "License");
// you may not use this file except in compliance with the License.
//
// This software is provided 'as-is', without any expressed or implied
// warranty.
//
// Provides TSTRING support.

#ifdef _UNICODE

#include "tstring.h"
#include "validateunicode.h"
#include <vector>

// Simple RAII for C-styled FILE pointers.
class ScopedFile
{
	public:
		ScopedFile(FILE* file) : m_file(file) {}

		~ScopedFile()
		{
			if (this->m_file != NULL)
			{
				fflush(this->m_file);
				fclose(this->m_file);
			}
		}

		operator FILE*(){ return this->m_file; }

		operator bool() { return this->m_file != NULL; }

	private:
		FILE* m_file;
};

FILE* FileOpenUnicodeText(const TCHAR* file, const TCHAR* mode)
{
	extern FILE *g_output;
	CValidateUnicode::FILE_TYPE ftype = CValidateUnicode::UTF_16LE;

	// If we are reading an existing file, check to see what type of file it
	// is first.
	if (_tcsstr(mode, _T("w+")) ||
	    _tcsstr(mode, _T("r")))
	{
		ScopedFile fp(_tfopen(file, _T("rb")));

		if (fp)
		{
			fseek(fp, 0, SEEK_END);
			size_t fileSize = ftell(fp);
			if (fileSize == 0)
			{
			   // Empty files are treated as UTF-8.
			   ftype = CValidateUnicode::UTF_8;
			}
			else
			{
			   std::vector<unsigned char> buffer(fileSize);
			   fseek(fp, 0, SEEK_SET);
			   fread(&buffer[0], sizeof(unsigned char), fileSize, fp);

			   ftype = CValidateUnicode::CheckBOM(&buffer[0], buffer.size());

			   switch (ftype)
			   {
				   case CValidateUnicode::UTF_8:
				   case CValidateUnicode::UTF_16LE:
				   case CValidateUnicode::UTF_16BE:
					   break;
				   case CValidateUnicode::UTF_32LE:
				   case CValidateUnicode::UTF_32BE:
					   _ftprintf(g_output, _T("File '%s' has a BOM marked as %s which is not supported at this time.\n"),
							   file, CValidateUnicode::TypeToName(ftype));
					   exit(-1);
					   break;
				   case CValidateUnicode::UNKNOWN:
					   // If unknown, let's see if it's not just UTF_8 without a BOM.
					   if (CValidateUnicode::ValidateUTF8(&buffer[0], buffer.size()) == 2)
					   {
                           // contains UTF-8 characters sequences
						   _ftprintf(g_output, _T("File '%s' has no BOM but seems to be UTF-8.\n"), file);
						   ftype = CValidateUnicode::UTF_8;
                       }
					   break;
				   default:
					   _ftprintf(g_output, _T("CValidateUnicode::CheckBOM() for file '%s' returned an unknown return value: %d\n"),
							   file, ftype);
					   exit(-1);
					   break;
			   }
			}			   
		}
	}

	tstring strMode(mode);

	switch (ftype)
	{
		case CValidateUnicode::UTF_8:
			strMode.append(_T(", ccs=UTF-8"));
			break;
		case CValidateUnicode::UTF_16LE:
			strMode.append(_T(", ccs=UTF-16LE"));
			break;
		default:
			// Looks like fopen() doesn't support other encodings of Unicode.
			strMode.append(_T(", ccs=UNICODE"));
			break;
	}

	return _tfopen(file, strMode.c_str());
}

#endif