wiki:Clients

Version 36 (modified by Dimitar Misev, 7 years ago) ( diff )

—

  1. Clients
    1. Clients coming with rasdaman
      1. Rasql query language access
      2. python
      3. Geo clients
      4. raswct (rasdaman Web Client Toolkit)
    2. Third-party clients
      1. R (data analysis)
      2. NASA WorldWind (Web-based virtual globe)
      3. OpenLayers and Leaflet (Web-based map navigation)
      4. ncWMS (Web-based climate data navigation)
      5. QGIS (Web GIS)
      6. MapServer (Web GIS server)
      7. GeoServer (Web GIS server)
      8. GDAL (image processing library)
      9. THREDDS (scientific data cataolog)
      10. ESRI ArcGIS (GIS tool)

Clients

It is important for data users to get convenient access to rasdaman functionality, for manifold reasons: users want to stay in their well-known environments (such as QGIS, ESRI, or R); users have additional functionality in their tools which they want to apply on data retrieved from rasdaman; users do not want to learn yet another data language. To this end, a large and increasing number of tools can access rasdaman today, thereby enabling these with flexible, scalable server-side processing of n-D "Big Array Data". They largely fall into two categories, rasdaman-provided and "external" third-party tools. Below they are listed and explained.

rasdaman can interface with a large and increasing number of tools. Some clients allow direct query submission, such as rasql and directql. For rapid development of browser clients the raswct toolkit is provided.

Geo tools supported include OpenLayers, QGIS, ncWMS, THREDDS, GDAL, MapServer, !ArcGIS; soon: GeoServer. Geo standards supported are OGC WMS, WCS, WCPS, and WPS; hence, all clients conforming to these standards can additionally be used for interfacing. All geo semantics is implemented through the petascope component (​source) which is implemented as a Java application layer offering OGC Web services.

Clients coming with rasdaman

Rasql query language access

The following clients allow to submit rasql queries:

  • rasql is a command-line utility for sending queries and saving results in local files. Invocation example (see rasql --help for details): 
    rasql -q 'select encode( mr, "png" ) from mr' --out file
    • To see a list of all collections with arrays in the database: 
      rasql -q 'select c from RAS_COLLECTIONNAMES as c' --out string
  • rasql queries can also be invoked via a Web service by default listening at localhost:8080/rasdaman/rasql. To this end, the corresponding servlet needs to be started (see Installation Guide).
    • Syntax: 
      http://{service}/{path}/rasdaman/rasql?params

Example: 
http://www.acme.com/rasdaman?query=select%20rgb.red+rgb.green%20from%20rgb&username=rasguest&password=rasguest
    • This servlet endpoint accepts KVP requests with the following parameters:
      • query=q where q is a valid rasql query, appropriately escaped as per http pecification.
      • username=u where u is the user name for logging into rasdaman (current: rasguest for retrieval query (select) and rasadmin for the editable query (insert, update, delete,…).
      • password=p where p is the password for logging into rasdaman (current: rasguest for username: rasguest and rasadmin for username: rasadmin).
    • To upload data to server, it is needed to attach the data in HTTP message's body and send the POST request to the Servlet endpoint with user have the write permission (e.g: rasadmin). In Linux, one tool can support this feature is cURL. Syntax: 
      curl -F 'image=@{/Path_To_Image}' '{http://Rasql_Servlet_Endpoint?username=rasadmin&password=rasadmin&query={insert/update query}'
      • Here is an example: 
        Collection: test_mr_test1

Insert file to collection:
curl -F 'image=@/home/rasdaman/rasdaman/systemtest/testcases_services/test_all_wcst_import/test_data/wcps_mr/mr_1.png' 'http://localhost:8080/rasdaman/rasql?username=rasadmin&password=rasadmin&query=insert%20into%20test_mr_test1%20values%20decode(%241)' 

Update file to collection:
curl -F 'image=@/home/rasdaman/rasdaman/systemtest/testcases_services/test_all_wcst_import/test_data/wcps_mr/mr_1.png' 'http://localhost:8080/rasdaman/rasql?username=rasadmin&password=rasadmin&query=update%20test_mr_test1%20set%20test_mr_test1%20assign%20decode(%241)'
  • directql is not strictly a client, but a rasdaman server linked directly with the rasql utility (see above). Main purpose is for simplified debugging of the server by allowing command-line invocation of the server engine. (​source )
  • rview is a deprecated visual query interface for desktop use (executable). It is still part of the rasdaman package, but not maintained any longer and will soon be phased out. See below.

python

  • There is a ​RasdaPy Python package that allows to connect to rasdaman, send queries and get back results as Python or numpy array.
  • An OWSLib fork provides support for WCPS; this ​tutorial provides examples.

Geo clients

raswct (rasdaman Web Client Toolkit)

This JavaScript library allows, by way of input and output widgets, to easily compose a Web client. Simple versions need only HTML, but advanced interfaces can be created. One example for the use of raswct is ​standards.rasdaman.org. See documentation for details.

Third-party clients

R (data analysis)

RRasdaman is an R package providing a database interface for rasdaman, i.e.: for submitting rasdaman queries from within R. It is submitted to R and part of this toolset. See installation/setup guide and tutorial for more information.

NASA WorldWind (Web-based virtual globe)

documentation tbd

OpenLayers and Leaflet (Web-based map navigation)

documentation tbd

ncWMS (Web-based climate data navigation)

documentation tbd

QGIS (Web GIS)

This plugin enables the direct download, processing, integration and usage of Coverages from an OGC compliant WCPS 1.0 enabled servers within the QGIS software (v2.0+ supported). The plugin has been submitted to the official QGIS plugin repository, approval is pending; in the meantime it can be installed manually:

  • get the rasdaman source tree: 
    git clone git://rasdaman.org/rasdaman.git
  • the plugin can be found in directory rasdaman/applications/qgis-wcps
  • copy the qgis-wcps/QgsWcpsClient1 directory to your qgis-python-plugin directory, usually located in $HOME/.qgis2/python/plugins: 
    mkdir -p $HOME/.qgis2/python/plugins/
cp -r rasdaman/applications/qgis-wcps/QgsWcpsClient1 $HOME/.qgis2/python/plugins/
  • now you can start QGIS and access the QgsWcpsClient1 plug-in via the Qgis Menu → Plugins → WcpsClient1 → WCPS 1.0 Client

MapServer (Web GIS server)

documentation tbd

GeoServer (Web GIS server)

documentation tbd

GDAL (image processing library)

To compile GDAL with rasdaman support:

  1. Get latest GDAL source
  2. Compile and install (in the /home/rasdaman/gdal) 
    ./configure --prefix=/home/rasdaman/gdal --mandir=/home/rasdaman/gdal/share/man --includedir=/home/rasdaman/gdal/include/gdal 
  --with-threads --with-grass=no --with-hide-internal-symbols=yes --with-libtiff=internal --with-geotiff=internal --with-jasper 
  --with-netcdf --with-xerces --with-geos --with-sqlite3 --with-curl --with-pg --with-ogdi --with-python --with-odbc 
  --with-static-proj4=yes --with-spatialite=/usr --with-cfitsio=no --with-epsilon=/usr --with-rasdaman=/home/rasdaman/install
make
make install
    • Note: rasdaman should be compiled with -fPIC, so make sure to specify --with-pic when running ./configure on rasdaman.
  3. Setup environment 
    export GDAL_HOME=/home/rasdaman/gdal
export PATH=$PATH:$GDAL_HOME/bin
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$GDAL_HOME/lib

THREDDS (scientific data cataolog)

documentation tbd

ESRI ArcGIS (GIS tool)

documentation tbd

Note: See TracWiki for help on using the wiki.