eaudeweb / natura2000db Goto Github PK

http://biodiversitate.mmediu.ro/rio/natura2000

A database and search interface for natura2000 sites in Romania. For every protected area the database contains a document with a complex schema, that is indexed in Solr. The database can be searched via a faceted search interface through the web app.

The website also contains an interactive map with point and polygon editor, for use by environment experts, to mark industrial and other man-made sites within a protected area.

The application is implemented in Flask, with Backbone on the front-end, with solr acting as index and also primary data store. It runs on any Unix-like or Windows OS with Python 2.6 or 2.7 and Java 1.5 for Solr.

http://biodiversitate.mmediu.ro/rio/natura2000/map/

The map displays several layers :SCIs and SPAs and other protected areas by default; administrative borders, water bodies, and infrastructure can be turned on. Clicking on the map shows information about the features at those coordinates in a box in the top-right corner of the map. There is also a measurement functionality (distances and areas) at the top part of the left-side menu. The base layer can be switched between OpenStreetMap, Bing maps, and Google maps.

There is an additional editable layer that is tied to the user's gmail account if they log in. Users can create points and polygons by entering coordinates in the Stereo70 projection. For instance, lng=500000 lat=500000 is near the center of Romania. A user must explicitly save the layer (from the cog-wheel menu next to the point/polygon creation functions) for it to be persisted.

1. Install Python >= 2.6 and virtualenv; activate the virtualenv:

   virtualenv -p python2.6 sandbox
   echo '*' > sandbox/.gitignore
   . sandbox/bin/activate
   

2. Install dependencies:

   pip install -r requirements-dev.txt
   

3. Optionally create a local configuration file:

   mkdir instance
   echo "DEBUG = True" > instance/settings.py
   

4. Run the server:

   ./manage.py runserver
   

Documents are stored and indexed in a Solr database. Solr is installed separately but the configuration is provided in the solr directory.

1. Download Solr somewhere outside the project repository:

   curl -O http://www.eu.apache.org/dist/lucene/solr/3.5.0/apache-solr-3.5.0.tgz
   tar xzf apache-solr-3.5.0.tgz
   

2. Start Solr with our configuration. Assuming __solrkit__ is the apache-solr-3.5.0 folder unpacked above, and __repo__ is the project repository, run:

   cd __solrkit__/example
   java -Dsolr.solr.home=__repo__/solr -jar start.jar
   

A minimal buildout is provided in the buildout folder. It's mostly self-contained except for the NaayaBundles-CHMEU package and the database file.

Get the NaayaBundles-CHMEU package:

mkdir buildout/src
cd buildout/src
svn co https://svn.eionet.europa.eu/repositories/Naaya/trunk/eggs/NaayaBundles-CHMEU

In the buildout directory, bootstrap and run the buildout:

cp buildout-example.cfg buildout.cfg
python2.4 bootstrap.py -d
bin/buildout

Finally, copy the database file (Data.fs) to var/filestorage/.

To run the application with Zope, first start Solr, then the manage.py server, and then zope (by running bin/zope-instance fg from within the buildout directory). The application should be accessible at http://localhost:8080/chm_ro/rio/natura2000/.

1. Dump documents to json. This assumes the Access database is loaded in MySQL on localhost, in the database rio:

   pip install -r migrations/requrements.txt
   ./manage.py accessdb_mjson > instance/access.mjson
   

2. Load json to Solr:

   ./manage.py import_mjson < instance/access.mjson