#!/perl

# Create a street mispellings hash to map 
# misspelled street names to correct names

%spell = ( 'BRD'   => 'BROAD',
	   'WWOOD' => 'WYNNEWOOD',
	   'WYNWD' => 'WYNNEWOOD' );

# Open a tab-delimited file of crime data
# It is assigned to the CRIMES filehandle variable
open ( CRIMES,      "c:/projects/nicar2000/crimes.txt" );
open ( CRIMESFIXED, ">c:/projects/nicar2000/crimesfixed.txt" );

# Loop thru the file 1 line at a time
while ( $line = <CRIMES> ) {

    # Remove the trailing newline character
    chomp $line;

    # Split the line of text based on tabs (\t)
    # and fill the variables with it
    ( $id, $sector, $ucr_code, $location )
	= split ( /\t/, $line );

    # 400 N BROAD ST
    if ( $location =~ m/^(\d+) ([NSEW]) (.+)$/ ) {
	$block_number = $1;
	$direction    = $2;
	$street       = $3;

    # 400 BROAD ST
    } elsif ( $location =~ m/^(\d+) (.+)$/ ) {
	$block_number = $1;
	$direction    = "";
	$street       = $2;

    # BROAD ST 400
    } elsif ( $location =~ m/^(.+) (\d+)$/ )  {
	$street       = $1;
	$block_number = $2;
	$direction    = "";

    # BROAD ST 400 N
    } elsif ( $location =~ m/^(.+) (\d+) ([NSEW])$/ )  {
	$street       = $1;
	$block_number = $2;
	$direction    = $3;

    } else {
	print "++++ Couldn't parse $location\n";
    }

    # Check for ST|RD|AV etc in the street field
    if ( $street =~ m/^(.+) (ST|AV|RD)$/ ) {
	$street_name = $1;
	$street_type = $2;
    } else {
	$street_name = $street;
	$street_type  = "";
    }

    # Fix mispelled names 
    if ( $spell{$street_name} ) {
	$street_name = $spell{$street_name};
    }

    # Join fields together with tabs
    $output = join ( "\t", $id, $sector, $ucr_code, $location,
		     $block_number, $direction, $street_name, $street_type );

    # Print the line out
    print CRIMESFIXED "$output\n";

}

close CRIMES;
close CRIMESFIXED;