html2perl

main.cpp

/////////////////////////////////////////////////////////////////////
//
//   $Id: main.cpp,v 1.9 2000/11/10 21:01:09 ricardo Exp $
//
//   Filename:           : main.cpp
//   Date:               : Sat Oct 21 16:52:01 BST 2000
//   Copyright:          : (C) 2000 by Ricardo Gladwell
//   Email               : ricgladwell@netscapeonline.co.uk
//
/////////////////////////////////////////////////////////////////////
//                                                                 //
//   This program is free software; you can redistribute it and/or //
//   modify it under the terms of the GNU General Public License   //
//   as published by the Free Software Foundation; either version  //
//   2 of the License, or (at your option) any later version.      //
//                                                                 //
/////////////////////////////////////////////////////////////////////

#ifdef HAVE_CONFIG_H
#include <config.h>
#endif

/////////////////////////////////////////////////////////////////////
//  D E F I N E   S T A T E M E N T S  //////////////////////////////
/////////////////////////////////////////////////////////////////////

#define ESCAPE_CHARACTERS "\t\"\'\\@%$#.:&!-"

/////////////////////////////////////////////////////////////////////
//  I N C L U D E   S T A T E M E N T S  ////////////////////////////
/////////////////////////////////////////////////////////////////////

#include <unistd.h>
#include <iostream.h>
#include <fstream.h>
#include <unistd.h>
#include <string>

/////////////////////////////////////////////////////////////////////
//  M A I N   F U N C T I O N  //////////////////////////////////////
/////////////////////////////////////////////////////////////////////

int main( int argc, char *argv[] )
{

/////////////////////////////////////////////////////////////////////
//  I N I T I A L I S A T I O N  ////////////////////////////////////
/////////////////////////////////////////////////////////////////////

/////////////////////////////////////////////////////////////////////
//  Variable Declarations

	string *file				= NULL;		// input file
	string buffer;							// buffer
	string escapees( ESCAPE_CHARACTERS );	// non-printable chars

	unsigned int index 			= 0;		// generic index
	int lines_read				= 0;		// lines read
	int lines_written			= 0;		// lines written

	bool cgi					= false;	// cgi mode

	istream *in					= &cin;		// input stream

	char c;									// command line argument	

/////////////////////////////////////////////////////////////////////
//  Get and parse command line options.

	while( ( c = getopt( argc, argv, "i:e:c" ) ) != -1 )
	{

		switch( c )
		{

			case 'i':
			// include additional escape characters
			{
				string argument( optarg );

				for( index = 0; index != argument.length(); index++ )
				{

					char c = argument.at( index );
					if( escapees.find( c ) == string::npos ) escapees += c;

				}

				break;
			}

			case 'e':
			// exclude specified escaped characters
			{
				string argument( optarg );
				string temp = "";

				for( index = 0; index != escapees.length(); index++ )
				{

					char c = escapees.at( index );
					if( argument.find( c ) == string::npos ) temp += c;

				}

				escapees = temp;
				break;
			}

			case 'c':
			// switch cgi mode

				cgi = true;
				break;

			default:

				cerr << "usage: " << argv[0] << " -[iec] [file]" << endl;
				exit( 1 );
            	break;

		}

	}

	if( optind < argc ) {

		file = new string( argv[optind] );

	}

/////////////////////////////////////////////////////////////////////
//  If a file name has been specified set-up input file stream.

	ifstream *fin = NULL;

	if ( file != NULL )
	{
		fin = new ifstream( file->c_str() );
		in = fin;
	}

	if ( !(*in) )
	{
		cerr << argv[0] << ": "
			<< (*file)
			<< ": No such file or directory"
			<< endl;
		return false;
	}

/////////////////////////////////////////////////////////////////////
//  B E G I N   M A I N   C O N T R O L   L O O P  //////////////////
/////////////////////////////////////////////////////////////////////

	if( cgi ) cout << "print \"Content-type: text/html\\n\\n";

	while ( !in->eof() )
	{

/////////////////////////////////////////////////////////////////////
// Read in a line into the string buffer character by characters
// until you hit a newline ('\n') or carriage return ('\r')
// character.

		// empty buffer
		buffer = "";

		// read line or until EOF
		for ( c = in->get(); c != '\n' && c != '\r'; c = in->get() )
		{
			if ( c == -1 ) break;
			buffer += c;
		}

		// increment line counter
		// initialise loop variables
		lines_read++;
		bool empty = true;
		int first_tab = -1;

/////////////////////////////////////////////////////////////////////
//  Eat leading whitespace.

		for( index = 0; index != buffer.length(); index++ )
		{

			if ( buffer[index] == '\t' && first_tab == -1 )
			{
				first_tab = index;
			}

			if ( !isspace( buffer[index] ) )
			{
				if ( first_tab != -1 ) index = first_tab;

				buffer.erase( 0, index );
 				empty = false;

				if( lines_written != 0 || cgi ) cout << "\";" << endl;
				break;
			}

		}

/////////////////////////////////////////////////////////////////////
// If line is empty simply insert an extra newline in the
// terminating newline string.

		if ( empty )
		{
			cout << "\\n";
			continue;
		}

/////////////////////////////////////////////////////////////////////
//  Find all bad characters in the escape character list and add the
// '\' escape character to the front.  When you encounter a tab
// character replace with the string '\\t'.

		for ( index = buffer.find_first_of( escapees, 0 );
			index != string::npos;
				index = buffer.find_first_of( escapees, index + 2 ) )
		{
			if ( buffer[index] == '\t' ) buffer.replace( index, 1, "\\t" );
			else buffer.insert( index, "\\" );
		}

		cout << "print \"" << buffer << "\", \"\\n";
		lines_written++;

	}

	cout << "\";" << endl;

/////////////////////////////////////////////////////////////////////
//  T E R M I N A T I O N  //////////////////////////////////////////
/////////////////////////////////////////////////////////////////////

	if( fin != NULL ) fin->close();
	return 0;

}