-
Notifications
You must be signed in to change notification settings - Fork 0
API for accessing Reliefweb data and sample applications
martinburchell/crisiscamp_reliefweb
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Folders and files
Name | Name | Last commit message | Last commit date | |
---|---|---|---|---|
Repository files navigation
================================================== ReliefWeb light-weight map viewer proof-of-concept Code by Crisis Camp London crew ================================================== Files: rss_php.php is an RSS feed parser, from http://rssphp.net/ mapsearch.php gets latest map details from ReliefWeb via RSS feed. At present we have feed URLs hard-coded as we don't have the details of how to generate the feed URLs automatically. This takes three parameters via HTTP GET request: the country requested, the document type requested, and a keywork search term. The country and document type are currently ignored, and the only keywords recognised are "idp" and "hospitals" - anything else will be ignored and treated as if the box was blank. To save bandwidth and improve performance, we store copies of the maps locally and can generate scrollable tilesets. For each item in the RSS feed, we check whether there is a local copy of the PDF, and get a copy if there isn't. If a copy has previously been made but no tile-set yet generated, we link to the file on our server and tell the user that the tiles haven't yet been generated. If the tiles have been generated, we link to both the tile viewer and the original PDF file. When getting a copy of a file from ReliefWeb, from the RSS feed we get the URL of a landing page that links to the actual PDF file, so we download this HTML file and use a regular expression to extract the URL of the PDF. RSS-urls.txt is a note of the URLs of the RSS feeds which have been copied into mapsearch.php ------- mapfiles (directory) is where the maps copied from ReliefWeb are stored. They are named according to the guid in the RSS feed, and are initially stored in this directory. Once a tile-set have been generated from a file, the tiles and the original PDF are moved into a sub-directory which has the guid as its name. The other files in this directory (documented below) are from Harry Wood, and are also available from his site at http://harrywood.co.uk/maps/haiti/tiledpdf/ mapfiles/cloudmade.html mapfiles/openlayers.html are two tile viewers. They both do much the same thing. mapfiles/googletilecutter.sh is a GPLed Bash script by Ian C. Stevens (http://crazedmonkey.com/) that "Cuts an image into tiles suitable for use with Google Maps." mapfiles/generate.sh is run with the name of a PDF file as its only argument. It uses ImageMagick to covert the PDF to a PNG image, and then runs googletilecutter.sh. It does this at different sizes, for different zoom levels. It then makes all the tiles it has generated world-readable and moves them in the appropriate sub-directory. TODO: This is *VERY* slow, and cannot be run in parallel. mapfiles/generateAll.sh runs generate.sh on each PDF file in a directory, in turn. If you have a lot of PDFs, this can take several days to complete. You have been warned! :)
About
API for accessing Reliefweb data and sample applications
Resources
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published