sl@0: #
sl@0: # Copyright (c) 2009 Nokia Corporation and/or its subsidiary(-ies).
sl@0: # All rights reserved.
sl@0: # This component and the accompanying materials are made available
sl@0: # under the terms of the License "Eclipse Public License v1.0"
sl@0: # which accompanies this distribution, and is available
sl@0: # at the URL "http://www.eclipse.org/legal/epl-v10.html".
sl@0: #
sl@0: # Initial Contributors:
sl@0: # Nokia Corporation - initial contribution.
sl@0: #
sl@0: # Contributors:
sl@0: #
sl@0: # Description:
sl@0: #
sl@0: # UnicodeMaxDecompose.pl
sl@0: #
sl@0: # Adds maximal decompositions of the character and maximal decompositions of
sl@0: # its folded varient to the Unicode data.
sl@0: #
sl@0: # Added as the fourth field after the 'Symbain:' marker in the following format:
sl@0: #
sl@0: # Symbian:<grapheme-role>;<excluded>;<folded>;<max-decomposition>;<folded-decomposition>
sl@0: # where each of <max-decomposition> and <folded-decomposition> are strings
sl@0: # of hex numbers separated by spaces, representing the complete decomposition
sl@0: # of the character and its folded equivalent respectively.
sl@0: #
sl@0: # Usage:
sl@0: # perl -w UnicodeMaxDecompose.pl < <output-of-UnicodeAddFolded>
sl@0: 
sl@0: use strict;
sl@0: 
sl@0: if (scalar(@ARGV) != 0)
sl@0: 	{
sl@0: 	print (STDERR "Usage:\nperl -w UnicodeMaxDecompose.pl < <output-of-UnicodeAddFolded>\n");
sl@0: 	exit 1;
sl@0: 	}
sl@0: 
sl@0: my %StatedDecomposition = ();
sl@0: my %CompleteDecomposition = ();
sl@0: 
sl@0: sub Decompose
sl@0: 	{
sl@0: 	my ($code) = @_;
sl@0: 	return unless exists $StatedDecomposition{$code};
sl@0: 	my $stated = $StatedDecomposition{$code};
sl@0: 	delete $StatedDecomposition{$code};
sl@0: 	my @complete = ();
sl@0: 	foreach my $hexelt ( split(' ', $stated) )
sl@0: 		{
sl@0: 		if ($hexelt)
sl@0: 			{
sl@0: 			Decompose($hexelt);
sl@0: 			if (exists $CompleteDecomposition{$hexelt})
sl@0: 				{
sl@0: 				push @complete, $CompleteDecomposition{$hexelt};
sl@0: 				}
sl@0: 			else
sl@0: 				{
sl@0: 				push @complete, $hexelt;
sl@0: 				}
sl@0: 			}
sl@0: 		}
sl@0: 	$CompleteDecomposition{$code} = join(' ', @complete);
sl@0: 	}
sl@0: 
sl@0: my %Folded = ();
sl@0: my %LineToCode = ();
sl@0: my @RawLine = ();
sl@0: 
sl@0: my $lineNo = 0;
sl@0: while (my $line = <STDIN>)
sl@0: 	{
sl@0: 	chomp $line;
sl@0: 	$lineNo++;
sl@0: 	# Split into fields: make sure trailing null strings are not
sl@0: 	# deleted by adding a dummy final field
sl@0: 	my @attribute = split(/;/, $line.';dummy');
sl@0: 	# Delete the dummy field
sl@0: 	pop @attribute;
sl@0: 	die ("Line $lineNo is missing 'Symbian:' entries. Has UnicodeAddFolded been run?")
sl@0: 		if (scalar(@attribute) == 16);
sl@0: 	if (scalar(@attribute) == 17)
sl@0: 		{
sl@0: 		die ("Line $lineNo is missing 'Symbian:' entries. Has UnicodeAddFolded been run?")
sl@0: 			if ($attribute[15] !~ /^[ \t]*symbian:/i);
sl@0: 		my $code = $attribute[0];
sl@0: 		die("First attribute '$code' not a valid Unicode codepoint at line $lineNo")
sl@0: 			unless ($code =~ /^1?[0-9a-fA-F]{4,5}$/ && hex($code) < 0x110000);
sl@0: 		my $decomposition = $attribute[5];
sl@0: 		die("Decomposition '$decomposition' at line $lineNo is not a valid Unicode decomposition.")
sl@0: 			unless $decomposition =~ /^[ \t]*(<.*>[ \t]*[0-9a-fA-F])?[0-9a-fA-F \t]*$/;
sl@0: 		my $folded = $attribute[16];
sl@0: 		die ("'$folded' not a valid string of hex values at line $lineNo.")
sl@0: 			unless $folded =~ /[0-9a-fA-F \t]*/;
sl@0: 		# Store all decompositions that  have no tag and at least one value
sl@0: 		if ($decomposition =~ /^[ \t]*[0-9a-fA-F]/)
sl@0: 			{
sl@0: 			$StatedDecomposition{$code} = $decomposition;
sl@0: 			}
sl@0: 		if ($folded =~ /[0-9a-fA-F]/)
sl@0: 			{
sl@0: 			$Folded{$code} = $folded;
sl@0: 			}
sl@0: 		$LineToCode{$lineNo-1} = $code;
sl@0: 		}
sl@0: 	elsif ($line !~ /^[ \t]*$/)
sl@0: 		{
sl@0: 		die 'Do not understand line '.$lineNo;
sl@0: 		}
sl@0: 	$RawLine[$lineNo-1] = $line;
sl@0: 	}
sl@0: 
sl@0: # Completely decompose all strings in the %StatedDecomposition
sl@0: foreach my $code (keys %StatedDecomposition)
sl@0: 	{
sl@0: 	Decompose($code);
sl@0: 	}
sl@0: 
sl@0: # Now decompose all the folded versions
sl@0: foreach my $code (keys %Folded)
sl@0: 	{
sl@0: 	my @result = ();
sl@0: 	foreach my $hexelt (split(' ', $Folded{$code}))
sl@0: 		{
sl@0: 		if (exists $CompleteDecomposition{$hexelt})
sl@0: 			{
sl@0: 			push @result, split(' ', $CompleteDecomposition{$hexelt});
sl@0: 			}
sl@0: 		else
sl@0: 			{
sl@0: 			push @result, $hexelt;
sl@0: 			}
sl@0: 		}
sl@0: 	$Folded{$code} = join(' ', @result);
sl@0: 	}
sl@0: 
sl@0: # Now output all the results
sl@0: for (my $i = 0; $i != scalar(@RawLine); $i++)
sl@0: 	{
sl@0: 	print $RawLine[$i];
sl@0: 	if (exists $LineToCode{$i})
sl@0: 		{
sl@0: 		my $code = $LineToCode{$i};
sl@0: 		print ';';
sl@0: 		my $decomp = '';
sl@0: 		$decomp = $CompleteDecomposition{$code}
sl@0: 			if exists $CompleteDecomposition{$code};
sl@0: 		print $decomp.';';
sl@0: 		if (exists $Folded{$code})
sl@0: 			{
sl@0: 			print $Folded{$code}
sl@0: 			}
sl@0: 		else
sl@0: 			{
sl@0: 			# If there is no folded value, but there is a decomposition
sl@0: 			# sequence, the character must fold to the decomposition
sl@0: 			# sequence too.
sl@0: 			print $decomp;
sl@0: 			}
sl@0: 		}
sl@0: 	print "\n";
sl@0: 	}