cccmap

A mini-project to make maps work better for complementary currency networks, first by showing their own, and secondly by publishing a csv file for the public, suitable for converting into a map. This should help efforts to collate data from all platforms for study and to testify to the growth of complementary currencies.

1. Share your site data for others (Such as the Credit Commons Collective) to aggregate and process.

Construct and (ideally) file_put_contents a file called geo.csv in your web root with the following columns:

Fieldname	Description	Example
url	the address for use in a link. the 'http' is not needed, and usually not the www	mysite.com
latitude	a floating point number between -90 and +90	45.93836
longitude	a floating point number between -180 and +180	6.827263
WKT	Well Known Text - (If lat & lon not supplied), an address string we can attempt to geocode
title	the name of the exchange	Anytown Hour Bank
description	Or slogan. one sentence max	We do things for each other!
logo	absolute url of the exchange's logo, smaller is better	mysite.com/logo.png
active_members	number of members in the exchange deemed 'active' e.g. have logged in or traded	99
year_transactions	Transactions recorded in the last year. Try to exclude automated and mass transactions	999
ads	Number of EXTANT offers + wants	999

There are several examples of all these files in this repository.
Remember there are php functions to help you generate and write csv files.
N.B. How do we ensure UTF8?

This format is subject to change and all maps being aggregated will be notified.

Offical rendering is on http://creditcommons.net/map N.B. Real world changes won't appear on maps immediately because both the sites, and the aggregator are likely to be keeping caches.

2. publish your own geojson file (optional)

Geojson is a standard format for publishing points on a map and info about them. You can can click on a map to learn the coordinates here: http://latlong.net or you can do it programmatically like this:

// create a free account on mapbox.com to get the $access_token
function geocode($placename) {
  $wkt = urlencode($string);
  $url = implode('/', [
    'https://api.mapbox.com',
    'geocoding',
    'v5',
    'mapbox.places',
    urlencode($string) .'.json'
  ]);
  $access_token = 'xxxxxx';
  $result = file_get_contents($url."?access_token=$access_token");
  $obj = json_decode($result);
  if (is_object($obj)) {
    if (count($obj->features)) {
      return $obj->features[0]->geometry->coordinates;
    }
    else {
      echo "<br />Geocoding '$string' failed: ". print_r($obj);
    }
  }

Then build an array of your communities and iterate through, populating the points. Note that the longitude goes before latitude in this format.

foreach ($myData as $site) {
  $allPoints[] = [
    'type' => 'Feature',
    'geometry' => [
      'type' => 'Point',
      'coordinates' => [$site['longitude'],  $site['latitude']]
    ],
    'properties' => [
      "title" => $site['name'], //plain text
      "description" => "blah blah" //html with more details. N.B. How do we ensure UTF8?
    ]
  ];
}
$geoJson = [//this is in the wrong place
  'type' => 'FeatureCollection',
  'features' => $allPoints
];
//This file can be viewed automatically in a github repository
file_put_contents('my_geo.json', json_encode($geoJson));//maybe best at the web root?

See the one at http://communityforge.net/geo.json, which is rebuilt regularly Ensure the list is kept up to date either with cron or with caching strategy. Next you probably want to render these point on your own site. This can be done again with the help of mapbox.com and the access token you created above.

3 Render the geojson on your site.

See index.php for an example of how to do this. You may need a mapbox access token from http://mapbox.com

and the points should now render on the map. Note that this map is centred on lat +40 and lon 0, and zoomed to 4.
You might have to add an apache directive (in .htaccess perhaps) to execute php on a page with an html extension.
AddType application/x-httpd-php .html Nginx is a bit more complicated.

To see the code we used to aggregate all the csv files and present a consistent look and feel, contact us.