View source for WCS GetCoverage for Grid returning CSV Example

<center>'''WCS python demos by [[User:Hoijarvi|Hoijarvi]]''' </center>
<center>For questions related to this page e-mail [[User:Hoijarvi|Hoijarvi]]</center> 

== Simple WCS Python Example ==
This example uses popular python scripting language to call Datafed WCS via SOAP. See WCS page [[SOAPifying_WCS]]

=== Getting python ===

Windows users: go to http://www.activestate.com and download Python 2.4 or later.
You can also go straight to [http://www.activestate.com/store/languages/register.plex?id=ActivePython download]

The main site for python is http://www.python.org but I prefer to install via the activestate package. Especially Mark Hammonds PythonWin is a nice IDE.

=== Getting Demo Script ===

Download http://datafed.net/demo/soapdemo.txt and save it with extension '''.py'''

type from command line '''python soapdemo.py''' and watch it run.

There are no parameters, the script is hard coded for demo purposes.

=== Defining the Query ===

After some declarations there is the WCS query wrapped in SOAP envelope. 

The triple quotes allow multiline string constants in python. 

We have now variable '''example_soap_in'''

<pre>
#   this WCS query queries USA temperatures 2005-12-06T12:00:00 at 1350 m height.
#   The Envelope and Body nodes are standard SOAP wrapping.
#   The actual query is the http-post query defined by WCS schema.

example_soap_in = """
<soap:Envelope xmlns:soap="http://schemas.xmlsoap.org/soap/envelope/">
    <soap:Body>
        <wcs:GetCoverage version="1.0.0" service="WCS"
                xmlns:gml="http://www.opengis.net/gml" xmlns:wcs="http://www.opengis.net/wcs">
            <!-- datafed dataset_abbr.param_abbr defines the coverage name -->
            <wcs:sourceCoverage>THREDDS_GFS.T</wcs:sourceCoverage> 
            <wcs:domainSubset>
                <wcs:spatialSubset>
                    <!--
                        This element queries the USA at 1350 m elevation.
                        If the dataset has no elevation, only lat and lon limits
                        are needed. 
                    -->
                    <gml:Envelope srsName="WGS84(DD)">
                        <!-- lon_min lat_min elev_min -->
                        <gml:pos>-130 24 1350</gml:pos>
                        <!-- lon_max lat_max elev_max -->
                        <gml:pos>-65 50 1350</gml:pos>
                    </gml:Envelope>
                    <gml:Grid dimension="3">
                        <gml:limits>
                            <!--
                                grid size. Datafed does not support resizing, so
                                these numbers have no meaning for grid datasets.
                            -->
                            <gml:GridEnvelope>
                                <gml:low>0 0 0</gml:low>
                                <gml:high>99 99 0</gml:high>
                            </gml:GridEnvelope>
                        </gml:limits>
                        <gml:axisName>lat</gml:axisName>
                        <gml:axisName>lon</gml:axisName>
                        <gml:axisName>elev</gml:axisName>
                    </gml:Grid>
                </wcs:spatialSubset>
                <wcs:temporalSubset>
                    <!--
                        query data for one time only.
                        Datafed currently supports also time ranges,
                        but only for stored time interval.
                        Enumerated times are not yet supported.                        
                    -->
                    <gml:timePosition>2005-12-06T12:00:00</gml:timePosition>
                </wcs:temporalSubset>
            </wcs:domainSubset>
            <wcs:output>
                <!--
                    For test purposes, use Comma Separated Values.
                    NetCDF and GML are also supported.
                -->
                <wcs:format>CSV</wcs:format>
            </wcs:output>
        </wcs:GetCoverage>
    </soap:Body>
</soap:Envelope>
"""
</pre>

=== Making http-post request ===

Function '''query_datafed_wcs''' makes a WCS query to Datafed. Remove the tracing print statements if you use this for real.

It does it by doing all the low level http stuff: open connection, send headers, send data, inspect return, parse return xml, get the real data with another http-get request.

<pre>
# This function performs WCS SOAP query to datafed
# INPUT: the soap query, similar to example_soap_in
# output: the data stream

def query_datafed_wcs(soap_in_text):
    # open connection to std port 80
    c = httplib.HTTPConnection("webapps.datafed.net", 80)
    c.connect()
    try:
        # the Web Service Definition Language definition is at
        # http://webapps.datafed.net/WCS.asmx?WSDL
        c.putrequest("POST", "/WCS.asmx")
        c.putheader("soapAction", "http://datafed.net/WCS/GetCoverage")
        c.putheader("content-type", "text/xml")
        c.putheader("content-length", repr(len(soap_in_text)))
        c.endheaders()
        c.send(soap_in_text)
        r = c.getresponse()
        print 'response status '   repr(r.status)   ' '   repr(r.reason)

        # 200 means OK. Anything other is a failure.
        if r.status <> 200:
            raise 'http-error', repr(r.status)   ' '   repr(r.reason)

        x = xml.dom.minidom.parse(r)
        print x.toxml()

        # We have the output. It does not contain data, just
        # metadata and an uri pointer to the real data.
        try:
            # The uri is written into 'data' node, so let's look for it.
            for u in x.getElementsByTagNameNS("http://datafed.net/xs/wcs", "data"):
                print 'fetching data '   u.firstChild.nodeValue
                return urllib2.urlopen(urllib2.Request(u.firstChild.nodeValue))
            # We did not find the uri, so the result is an error            
            raise 'soap-error', x
        finally:
            # due to circular references, you have to have the next line.
            # Otherwise memory will leak. Other python implementations
            # like Jython or IronPython rely on platform provided
            # garbage collection, and don't need it. 
            x.unlink()
    finally:
        #closing expensive resources ASAP is a good practice
        c.close()
</pre>

=== Running the script ===

Python has no main program, but if invoked from command line the special variable '''__name__'''  is set to '''__main__'''. So if you issue command '''python soapdemo.py''' then the following section will be executed.

Alternative way to run this is to 
* start python
* type '''import soapdemo''' without the '''.py''' extension
* type '''csv = soapdemo.query_datafed_wcs(soapdemo.example_soap_in)''' You now have http stream named csv.
* print it by '''print csv.read()'''

<pre>
#
#   If you run this script from command line, the next will be executed
#
if __name__ == "__main__":
    print 'querying datafed'
    csv = query_datafed_wcs(example_soap_in)
    print "here's the data"
    print ''
    # we have the csv stream, so let's look at it
    print csv.read()
</pre>