#!/usr/bin/perl
#
# gps2ipod
#
# Version 2.0
#
# Synchronizes geocaches between Garmin GPS and text files in
# the current working directory.  It does the following:
#
# - Download all the waypoints from your GPS
# - Extract all the Geocaches (waypoint ID with prefix GC)
# - Get all the files in the current directory
# - Extract all the Geocache .txt files (GC*.txt)
# - Determine which .txt files no longer correspond to a
#   Geocache waypoint on the GPS, and delete them
# - Determine which Geocaches have already been synced
#   to the current directory, and will remove them from the list
# - Download (slowly) the Geocaching.com page for each cache
# - Transform it into a useful text file
# - Save the text file in the current directory
#
# To synchronize your iPod with your GPS, cd to your iPod's
# "Notes" directory, then run this script.
#
# To avoid the wrath of the Geocaching.com site admins, this
# script implements the following to control bandwidth:
#
# - a 10-second pause between each page request
# - a 30-second pause between each 20 page requests
#
# This mimics the behavior of a user who used the search page
# to get a list of geocaches to inspect.
#
# For questions, comments, and suggestions, contact:
#
# Matthew M. Swann
# swannman@mac.com
#
##############################################################

# Declared at end of file
sub SanitizeHTML;
sub RemoveDuplicates;

# The building blocks of the URL used to fetch the geocache data
# for each waypoint
$URLprefix = "http://www.geocaching.com/seek/cache_details.aspx?wp=";
$URLsuffix = "&Submit6=Find&pf=y&decrypt=y";

# Counter for number of waypoints downloaded from Geocaching.com
# so that we can pause after every 20 to lessen the load
$counter = 0;

# Download all waypoints from the GPS
$waypoints = `gpsbabel -i garmin -f /dev/cu.USA19H821P1.1`;

# Extract all the Geocache IDs into a list
@IDs = $waypoints =~ /\s(GC.+)\//g;

# Get all the files in the current directory
opendir dir, ".";
@files = readdir dir;
closedir dir;

# Make one long string out of all the filenames
$fileblob = join "", @files;

# Then make a new list out of only the Geocache .txt files
# (There MUST be a better way to do this!)
@files = $fileblob =~ /GC.+?\.txt/g;

# Make a new list containing all the Geocaches that exist in the
# current directory but not on the GPS, so we can remove them
my @toRemove;
outer2: foreach $file (@files) {
	# For each file in the dir, look at each Geocache on the GPS
    foreach $ID (@IDs) {
    	# If there's a Geocache that matches the file
        if (($ID . ".txt") eq $file) {
        	# Skip to the next file in the dir, this one is fine
        	next outer2;
        }
    }
    # If we got here, then we didn't find an ID to match the current file
    push @toRemove, $file;
}

# If we found files to remove, take care of them
if (scalar(@toRemove) > 0) {
	# Let the user know what we determined, and give him a few seconds to cancel
	print "Removing " . scalar(@toRemove) . " caches from the current directory" . "\n";
	sleep 5;
	
	# Remove the files
	unlink @toRemove;
	print "Done\n\n";
}

# Remove Geocaches I've already downloaded from the list of IDs
outer: foreach $ID (@IDs) {
	# For each Geocache ID, look at each Geocache .txt file
    foreach $file (@files) {
    	# If there's a file that matches the ID
        if (($ID . ".txt") eq $file) {
        	# Remove the ID from the list
        	$ID = undef;
        	
        	# Skip ahead to the next ID in the loop, since we found
        	# a match (no need to look at the rest of the @files)
        	next outer;
        }
    }
}

# For each waypoint in the list whose ID starts with GC,
# capture the ID, build a URL, and get the HTML from that URL.
foreach $ID (@IDs) {

	# Skip the IDs we deleted (which are now just null entries in the list)
	if ($ID eq "") {
		# Move along, nothing to see here
		next;
	}
	
	print $ID . "\n";
    
    # Use the ID to create the URL for the Geocache's webpage
    my $URL = $URLprefix . $ID . $URLsuffix;
    
    # Use the URL to create the command to fetch the HTML for the webpage
    my $CMD = "curl -s \"" . $URL ."\"";
    
    # Execute the command and store the results in $HTML
    my $HTML = `$CMD`;
    
    # Update the number of times we've accessed a page
    $counter++;
    
    # Extract the CacheName
    $HTML =~ /CacheName\"\>(.+?)\</;
    my $CacheName = $1;
    
    # Extract the CacheSize
    $HTML =~ /CacheOwner.+\>(Size:\s.+?)\</;
    my $CacheSize = $1;
    
    # Extract the ShortDescription and sanitize it
    $HTML =~ /ShortDescription\"\>(.+?)\<\/span/s;
    my $ShortDesc = SanitizeHTML($1);
    
    # Extract the LongDescription and sanitize it
    $HTML =~ /LongDescription\"\>(.+?)\<\/span/s;
    my $LongDesc = SanitizeHTML($1);
    
    # Extract the Hints and sanitize them
    $HTML =~ /Hints\"\>(.+?)\<\/span/;
    my $Hints = SanitizeHTML($1);
    
    # Open a new file named after the Geocache ID
    $filename = ">" . $ID . ".txt";
    open my $file, $filename or die "Can't creat logfile: $!";
    
    # Write our data out to a file
    print $file "<title>" . $ID . "</title>\n";
    print $file $CacheName . "\n";
    print $file $CacheSize . "\n\n";
    print $file $ShortDesc . "\n\n";
    print $file $LongDesc . "\n\n\n\n\n\n\n";
    print $file $Hints;
    
    # Flush the buffer and close the file, which avoids 0-byte files
    # if the user hits Control-C during the call to sleep()
    close $file;
    
    # Pause so Geocaching.com doesn't get mad at us
    if (($counter % 20) == 0) {
        # Pretend that we got to the end of the current query
        # if we've done twenty page loads
        sleep 30;
    } else {
        # Pretend that we're busy reading the page
        sleep 10;
    }
}

# SanitizeHTML()
# Transform HTML tags into equivalent text
################################################################

sub SanitizeHTML {
    # Put the HTML to be sanitized in $raw
    my $raw = shift(@_);
    
    # <p> and <br> become carriage returns
    $raw =~ s/\<p\>/\n/ig;
    $raw =~ s/\<br\>/\n/ig;
   
    # <li> becomes -
    $raw =~ s/\<li\>/- /ig;
    
    # All other tags disappear
    $raw =~ s/\<.+?\>//gs;
    
    # Back where it came from
    return $raw;
}