source: TI01-discovery/trunk/python/ws2/test_discovery.py @ 1574

# these imports are autogenerated by wsdl2py ...
from DiscoveryService_services_types import *
from DiscoveryService_services import *

#normal imports
import sys
try: #python 2.5
    import xml.etree.ElementTree as ET
except:
    import ElementTree as ET
    
from ETxmlView import loadET,et2text

class ndgSearch:
    ''' Proivdes a client to the NDG discovery services methods exposed by the Discovery.wsdl '''
    
    def __init__(self):
        '''get an instance of the service'''
        #how do we get a different backend provider?
        loc=DiscoveryServiceLocator()
        self.server=loc.getDiscovery()#tracefile=sys.stdout)
        self.documents=None
        self.hits=None
        self.serverSessionID=None
    
    def search(self,term,start=1,howmany=30):
        ''' Get a list of documents for a specific term using a free text search'''
        
        #create a request object
        request=doFullTextSearchRequest()
        request._term=term
        request._start=start
        request._howmany=howmany
        
        #do the search
        response=self.server.doFullTextSearch(request)
        if response._status:
            self.serverSessionID=response._resultId
            self.hits=response._hits
            self.documents=response._documents._document
        self.status=[response._statusMessage,]
        return self.documents

    def getDoc(self,document):
        ''' Return a document from the backend database '''
        
        #create a request object
        request=doPresentRequest()
        request._document=document
        searchResult=self.server.doPresent(request)
        if searchResult._status:
            result=searchResult._document
        else:
            result=searchResult._statusMessage
        return result
                   
    def getAllDocs(self):            
        ''' Parse the list of documents, get them, and load them into a list of tuple strings (name,content) '''
        # we keep the name to make it easy to get the actual document out to avoid xpath issues ...
        
        docs=[]
        self.status=[]
        
        #create a request object
        request=doPresentRequest()
        for doc in self.documents:
            request._document=doc
            result=self.server.doPresent(request)
            if result._status:
                docs.append((doc,result._document))
            else:
                self.status.append((doc,result._statusMessage))
                
        self.docs=docs
        return self.docs
            
            
    def __xmlerror(self,docmessage):
        print docmessage
        return ET.fromstring('<Error><Document>%s</Document><Message>%s</Message></Error>'%docmessage)
            
    def getDocElement(self,document):
        ''' Takes a document path (maybe from a previous call to ndgSearch) and extracts that document
        into an ElementTree instance '''
        #we stick it straight into element tree because we need to use et to get the actual document
        #we want, not the envelope xml elements 
        
        doc=self.getDoc(document)
        path=document.rstrip('.xml')
        try:
            r=loadET(doc)
            #return r.find(path)
            return r
        except:
            return self.__xmlerror((path,doc))
            
    def getAllDocsAsElements(self):
        ''' Get all the documents and load them into a list of ET instances '''
        result=[]
        docs=self.getAllDocs()
        for doc in docs: 
            try:
                r=loadET(doc[1])
                path=doc[0].rstrip('.xml')
                result.append(r)#result.append(r.find(path))
            except:
                result.append(self.__xmlerror((path,doc[1])))
        return result
            
if __name__=="__main__":
    
    s=ndgSearch()
    docs=s.search('temperature')
    print s.status
    for i in docs: print i
    
    #docs=s.getAllDocs()
    #print docs[0]
    #print s.status
    
    doc=s.documents[1]
    r=s.getDocElement(doc)
    print et2text(r)