Changeset 3069

Timestamp:

29/11/07 10:15:03 (13 years ago)

Author:

lawrence

Message:

Semantic Search support in discovery
(Hopefully this is all the relevant files, my sandbox is in a bit of a
state)

Location:

TI05-delivery/ows_framework/trunk/ows_server/ows_server

Files:

• controllers/discovery.py (modified) (5 diffs)
• controllers/retrieve.py (modified) (1 diff)
• models/DocumentRetrieve.py (modified) (6 diffs)
• models/ndgRetrieve.py (modified) (7 diffs)
• templates/results.kid (modified) (2 diffs)
• templates/semantic.kid (modified) (1 diff)

TI05-delivery/ows_framework/trunk/ows_server/ows_server/controllers/discovery.py

@@ -5 +5 @@
 from ows_server.models import ndgSearch as NS
 from ows_server.models.DocumentRetrieve import DocumentRetrieve as DR
+from ows_server.models.DocumentRetrieve import ndgVocabPOX as VS
 from ows_server.models.DIF import DIF
 from ows_server.templates.renderDiscoverySet import renderDiscoverySet
@@ -12 +13 @@
 from ows_server.lib.mailer import mailHandler
 from ows_server.models.xmlHandler2 import xmlHandler
+debug=0
 
 
@@ -155 +157 @@
         print howmany,hits
         c.state=DiscoveryState(id,searchString,request.environ,hits,searchConstraints,start,howmany)
+        c.querystring=request.environ['QUERY_STRING']
         print c.state
         
@@ -202 +205 @@
                 
                 c.difs=difs
-                if 'searchString' in self.inputs:
-                    c.querystring='searchString=%s&textTarget=%s'%(self.inputs['searchString'],self.inputs['textTarget'])
-                else:
-                    c.querystring=''
+                
                 session['results']=h.current_url()
                 session.save()
@@ -292 +292 @@
     
     def semantic(self):
-        c.constraints=request.environ['QUERY_STRING']
+        self.__setup()
+        vs=VS(proxyServer=self.cf.get('DEFAULT','proxyServer'))
+        if 'searchString' in self.inputs:
+            [broader,narrower]=vs.getRelated(self.inputs['searchString'])
+            c.narrower=narrower
+            c.broader=broader
+            #get a base string for the links to new searches
+            if 'start' in self.inputs: del self.inputs['start']
+            if 'howmany' in self.inputs: del self.inputs['howmany']
+            self.inputs['searchString']='###SEARCHSSTRING###'
+            q='%s/discovery?'%g.server
+            for i in self.inputs: q+='%s=%s&'%(i,self.inputs[i])
+            url=q[0:-1]
+            # and now build the links 
+            c.narrower=[]
+            c.broader=[]
+            for i in narrower:
+                c.narrower.append((i,url.replace('###SEARCHSSTRING###',i)))
+            for i in broader:
+                c.broader.append((i,url.replace('###SEARCHSSTRING###',i)))
+        else:
+            broader,narrower=[],[]
+        
         return render_response('semantic',fragment=True)
     

TI05-delivery/ows_framework/trunk/ows_server/ows_server/controllers/retrieve.py

@@ -100 +100 @@
                     r='content'
             else:
+                name=str(self.uri)
                 if self.format=='xml':
                     c.xml=x.tohtml()

TI05-delivery/ows_framework/trunk/ows_server/ows_server/models/DocumentRetrieve.py

@@ -2 +2 @@
 from ndg_xqueries import *
 from xml.dom import minidom 
+try:
+    from xml.etree import ElementTree as ET
+except:
+    import ElementTree as ET
 import urllib2
 #from stripped_xqueries import strip_de_xquery
@@ -35 +39 @@
             xquery=xquery.replace('TargetCollection',targetCollection)
             xquery=xquery.replace('LocalID',localID)
-            print xquery
+            if debug:print xquery
             id,summary=self.executeQuery(xquery)
             if summary['hits']==1:
@@ -118 +122 @@
         return r
             
-class ndgHTTP:
-    ''' Provides a get method to obtain an xml document from a remote NDG repository '''
-    def __init__(self,remoteHost,proxyServer=None):
-        self.remoteHost=remoteHost
+class genericHTTP(object):
+    ''' Provides a generic HTTP request class '''
+    def __init__(self,proxyServer=None):
         if proxyServer is None:
             proxyHandler=urllib2.ProxyHandler({})
@@ -130 +133 @@
         self.opener=urllib2.build_opener(proxyHandler)
         
-    def uriget(self,uri):
-        n=ndgObject(uri)
-        return self.get(n.repository,n.schema,n.localID)
-
-    def get(self,repository,schema,localID,format='NDG-B0',targetCollection=None):
-        #last two arguments for compatibiltiy with API, not for use in restful get ...
-        #TODO what about security? Probably means we need to get the headers of our responses sorted ...
-        url='%s/view/%s__%s__%s?format=raw&outputSchema=%s'%(self.remoteHost,repository,schema,localID,format)
-        #http://docs.python.org/lib/urllib2-examples.html
+    def get(self,url):
         request=urllib2.Request(url)
         response='Error obtaining remote file: '
@@ -156 +151 @@
             return f.read()
         else:
-            raise IOError(response)
+            raise IOError(response)  
+            
+class ndgHTTP(genericHTTP):
+    ''' Provides a get method to obtain an xml document from a remote NDG repository '''
+    def __init__(self,remoteHost,proxyServer=None):
+        self.remoteHost=remoteHost
+        genericHTTP.__init__(self,proxyServer)
+        
+    def uriget(self,uri):
+        n=ndgObject(uri)
+        return self.get(n.repository,n.schema,n.localID)
+
+    def get(self,repository,schema,localID,format='NDG-B0',targetCollection=None):
+        #last two arguments for compatibiltiy with API, not for use in restful get ...
+        #TODO what about security? Probably means we need to get the headers of our responses sorted ...
+        url='%s/view/%s__%s__%s?format=raw&outputSchema=%s'%(self.remoteHost,repository,schema,localID,format)
+        #http://docs.python.org/lib/urllib2-examples.html
+        return genericHTTP.get(self,url)
         
     def setSecurity(self,location,usercode,password):
@@ -162 +174 @@
         pass
     
+class ndgVocabPOX(genericHTTP):
+    ''' Provides a POX interface to the vocab server '''
+    def __init__(self,path="http://vocab.ndg.nerc.ac.uk/axis2/services/vocab/",proxyServer=None):
+        genericHTTP.__init__(self,proxyServer)
+        self.path="http://vocab.ndg.nerc.ac.uk/axis2/services/vocab/"
+        self.ns="http://vocab.ndg.nerc.ac.uk/"
+    def getRelated(self,subject):
+        ''' Get a related record '''
+        url='%sgetRelatedRecordByCriteria?subjectText=%s&predicate=255&inferences=True&objectList=%slist/P211/current'%(self.path,subject,self.ns)
+        self.url=url
+        doc=genericHTTP.get(self,url)
+        x=ET.fromstring(doc)
+        b=x.findall('*/{urn:vocab/types}broadMatch')
+        n=x.findall('*/{urn:vocab/types}narrowMatch')
+        self.broader=[(i.find('{urn:vocab/types}entryTerm').text or '') for i in b]
+        self.narrower=[(i.find('{urn:vocab/types}entryTerm').text or '') for i in n]
+        return [self.broader,self.narrower]

TI05-delivery/ows_framework/trunk/ows_server/ows_server/models/ndgRetrieve.py

@@ -100 +100 @@
     def testSOAP(self):
         ''' test link to SOAP web service for DIF'''
-
         self.setup()
-        doc='neodc.nerc.ac.uk__DIF__NEODC_NEXTMAP'
+        doc='neodc.nerc.ac.uk__DIF__dataent_11707799142713994'
         uri=ndgObject.ndgObject(doc)
         status,xml=ndgRetrieve(uri,self.c)
@@ -109 +108 @@
     def testOriSOAP(self):
         ''' test link to SOAP web service for DIF (original version)'''
-
         self.setup()
-        doc='neodc.nerc.ac.uk__DIF__NEODC_NEXTMAP'
+        doc='neodc.nerc.ac.uk__DIF__dataent_11707799142713994'
         uri=ndgObject.ndgObject(doc)
         status,xml=ndgRetrieve(uri,self.c,output='original')
@@ -120 +118 @@
                 users will need to ensure that the document is actually 
                 in your exist database! '''
-        
         self.setup()
         doc='badc.nerc.ac.uk__NDG-B0__activity_activity_cwvc'
-        doc='neodc.nerc.ac.uk__NDG-B1__NEODC_NCAVEO_LCM2000'
+        doc='neodc.nerc.ac.uk__NDG-B1__dataent_11924879127625221'
         uri=ndgObject.ndgObject(doc)
         status,xml=ndgRetrieve(uri,self.c)
@@ -130 +127 @@
     def testNoSOAP(self):
         ''' Test that we get a sensible error if a DIF doesn't exist '''
-        
         self.setup()
         doc='badc.nerc.ac.uk__DIF__blah'
@@ -139 +135 @@
     def testNoB(self):
         ''' Test that we get a sensible error if a NDG-B doc doesn't exist '''
-        
         self.setup()
         doc='badc.nerc.ac.uk__NDG-B0__blah'
@@ -159 +154 @@
         ''' Attempt a remote get '''
         self.setup()
-        doc='www.npm.ac.uk__NDG-B1__data_entity.692'
+        doc='www.npm.ac.uk__NDG-B1__data_entity.363'
         uri=ndgObject.ndgObject(doc)
         status,xml=ndgRetrieve(uri,self.c,remote=1)
-        if not status: print xml
+        if not status: print xml #ie print error message if there is one!
         self.assertEqual(status,1)
         
@@ -169 +164 @@
         self.setup()
         doc='badc.nerc.ac.uk__NumSim__HadAM2_CodeBase'
+        doc='badc.nerc.ac.uk__NumSim__HadGEM1_Control_Experiment'
         uri=ndgObject.ndgObject(doc)
         status,xml=ndgRetrieve(uri,self.c)
-        print status,xml
+        self.assertEqual(status,1)
+        
+    def testHTTPget(self):
+        ''' Try a remote get with no frills'''
+        s='http://www.google.co.uk'
+        x=DocumentRetrieve.genericHTTP(proxyServer='http://wwwcache3.rl.ac.uk:8080/')
+        y=x.get(s)
+        
+    def testVocabServer(self):
+        ''' Test the vocabServer '''
+        v=DocumentRetrieve.ndgVocabPOX(proxyServer="http://wwwcache3.rl.ac.uk:8080/")
+        print v.getRelated('rain')
+        
+        
         
         

TI05-delivery/ows_framework/trunk/ows_server/ows_server/templates/results.kid

@@ -23 +23 @@
             <div id="context">
                 ${c.state.hits} results for ${c.state.searchString} (${XML(c.state.constraints)})</div>
-            <div id="SemanticSearchBox"/>
-            
+            <div id="SemanticSearchBox"><div id="SemanticSearchResults">
+                <div py:if="c.state.searchString!=''">
+                Not found what you are looking for? Try ${XML(h.link_to_remote("Alternative Searches",dict(update="SemanticSearchBox",
+                        url='%s?%s'%(h.url_for(action='semantic'),c.querystring))))}.
+                </div></div>
+            </div>
             <?python
             n,p=c.state.getNext()
@@ -32 +36 @@
             ?>
             <div class="resultsBar"> 
-                 <div class="resultsLeft">
-                    ${XML(h.link_to_remote("Semantic Search",dict(update="SemanticSearchBox",
-                            url='%s?%s'%(h.url_for(action='semantic'),c.querystring))))}
-                    <span py:replace="helpIcon('semser_help')"/>
-                 </div>
-                 <div>
-                 <a href="${c.state.constrainedurl}"> Refine Search</a><span py:replace="helpIcon('refser_help')"/>
-                | Found ${c.state.hits} | Showing ${c.state.offset}-$upper 
-                  <span py:if="p!=[]"> | <a href="$purl"> Previous ${p[1]}</a> </span>
-                  <span py:if="n!=[]"> | <a href="$nurl"> Next ${n[1]}</a></span> | </div>
-                <div id="semser_help" class="hidden">
-                    <div class="helptxt"><p>
-                    The semantic search option allows you to see how many hits a search using your current parameters would return if you
-                    used a slightly different vocabulary in your search term.
-                    </p></div>
-                </div>
-                <div id="refser_help" class="hidden">
+                <div>
+                    <a href="${c.state.constrainedurl}"> Refine Search</a><span py:replace="helpIcon('refser_help')"/>
+                    | Found ${c.state.hits} | Showing ${c.state.offset}-$upper 
+                    <span py:if="p!=[]"> | <a href="$purl"> Previous ${p[1]}</a> </span>
+                    <span py:if="n!=[]"> | <a href="$nurl"> Next ${n[1]}</a></span> | 
+                </div><div id="refser_help" class="hidden">
                     <div class="helptxt"><p>
                     The refined search option allows you to constrain your search by returning you to the advanced search page, but retaining your current search parameters

TI05-delivery/ows_framework/trunk/ows_server/ows_server/templates/semantic.kid

@@ -2 +2 @@
     <body>
         <div id="SemanticSearchResults">
-         Semantic Search not yet implemented ($c.constraints)
+            <span py:if="c.narrower!=[] or c.broader!=[]">You could also try </span>
+            <span py:if="c.narrower !=[]"> 
+                narrower search(s): <span py:for="i,url in c.narrower"><a href="$url">$i</a>&nbsp;</span>
+            </span>
+            <span py:if="c.narrower !=[] and c.broader!=[]">;</span>
+            <span py:if="c.broader !=[]"> 
+                broader search(s): <span py:for="i,url in c.broader"><a href="$url">$i</a>&nbsp;</span>
+            </span>
+            <span py:if="c.narrower==[] and c.broader==[]">No alternatives available</span>
         </div>
     </body>