[#1547][plugins,search] Add dataset search plugin interface

The IPackageController now also offers the `before_search` and
`after_search` methods so extensions can modify the search parameters
and the results coming from the standard Solr search.

Parameters starting with `ext_` will be added to an `extras` property
which will have no effect in the default search but will be forwarded
to the extensions as part of the search parameters.

Extensions willing to modify the standard search form can do so
implementing IGenshiStreamFilter and injecting their extra fields
in the `dataset-search-ext` div.

ckan/controllers/package.py

@@ -141,11 +141,15 @@ def pager_url(q=None, page=None):
 
         try:
             c.fields = []
+            search_extras = {}
             for (param, value) in request.params.items():
                 if not param in ['q', 'page'] \
                         and len(value) and not param.startswith('_'):
-                    c.fields.append((param, value))
-                    q += ' %s: "%s"' % (param, value)
+                    if not param.startswith('ext_'):
+                        c.fields.append((param, value))
+                        q += ' %s: "%s"' % (param, value)
+                    else:
+                        search_extras[param] = value
 
             context = {'model': model, 'session': model.Session,
                        'user': c.user or c.author}
@@ -155,6 +159,7 @@ def pager_url(q=None, page=None):
                 'facet.field':g.facets,
                 'rows':limit,
                 'start':(page-1)*limit,
+                'extras':search_extras
             }
 
             query = get_action('package_search')(context,data_dict)

ckan/lib/search/query.py

@@ -12,7 +12,8 @@
 
 VALID_SOLR_PARAMETERS = set([
     'q', 'fl', 'fq', 'rows', 'sort', 'start', 'wt', 'qf',
-    'facet', 'facet.mincount', 'facet.limit', 'facet.field'
+    'facet', 'facet.mincount', 'facet.limit', 'facet.field',
+    'extras' # Not used by Solr, but useful for extensions
 ])
 
 # for (solr) package searches, this specifies the fields that are searched 

ckan/logic/action/get.py

@@ -52,7 +52,7 @@ def package_list(context, data_dict):
     user = context["user"]
     api = context.get("api_version", '1')
     ref_package_by = 'id' if api == '2' else 'name'
-    
+
     check_access('package_list', context, data_dict)
 
     query = model.Session.query(model.PackageRevision)
@@ -115,7 +115,7 @@ def group_list(context, data_dict):
     if order_by not in set(('name', 'packages')):
         raise ValidationError('"order_by" value %r not implemented.' % order_by)
     all_fields = data_dict.get('all_fields',None)
-   
+
     check_access('group_list',context, data_dict)
 
     query = model.Session.query(model.Group).join(model.GroupRevision)
@@ -153,7 +153,7 @@ def group_list_authz(context, data_dict):
 
     query = Authorizer().authorized_query(user, model.Group, model.Action.EDIT)
     groups = set(query.all())
-    
+
     if available_only:
         package = context.get('package')
         if package:
@@ -307,7 +307,7 @@ def package_relationships_list(context, data_dict):
         rel = None
 
     check_access('package_relationships_list',context, data_dict)
-    
+
     # TODO: How to handle this object level authz?
     relationships = Authorizer().\
                     authorized_package_relationships(\
@@ -660,6 +660,10 @@ def package_search(context, data_dict):
 
     check_access('package_search', context, data_dict)
 
+    # check if some extension needs to modify the search params
+    for item in PluginImplementations(IPackageController):
+        data_dict = item.before_search(data_dict)
+
     # return a list of package ids
     data_dict['fl'] = 'id'
 
@@ -672,7 +676,7 @@ def package_search(context, data_dict):
         pkg_query = session.query(model.PackageRevision)\
             .filter(model.PackageRevision.id == package)\
             .filter(and_(
-                model.PackageRevision.state == u'active', 
+                model.PackageRevision.state == u'active',
                 model.PackageRevision.current == True
             ))
         pkg = pkg_query.first()
@@ -686,12 +690,19 @@ def package_search(context, data_dict):
         result_dict = package_dictize(pkg,context)
         results.append(result_dict)
 
-    return {
+
+    search_results = {
         'count': query.count,
         'facets': query.facets,
         'results': results
     }
 
+    # check if some extension needs to modify the search results
+    for item in PluginImplementations(IPackageController):
+        search_results = item.after_search(search_results,data_dict)
+
+    return search_results
+
 def _extend_package_dict(package_dict,context):
     model = context['model']
 
@@ -735,7 +746,7 @@ def resource_search(context, data_dict):
             raise SearchError('Field "%s" not recognised in Resource search.' % field)
         for term in terms:
             model_attr = getattr(model.Resource, field)
-            if field == 'hash':                
+            if field == 'hash':
                 q = q.filter(model_attr.ilike(unicode(term) + '%'))
             elif field in model.Resource.get_extra_columns():
                 model_attr = getattr(model.Resource, 'extras')
@@ -747,15 +758,15 @@ def resource_search(context, data_dict):
                 q = q.filter(like)
             else:
                 q = q.filter(model_attr.ilike('%' + unicode(term) + '%'))
-    
+
     if order_by is not None:
         if hasattr(model.Resource, order_by):
             q = q.order_by(getattr(model.Resource, order_by))
 
     count = q.count()
     q = q.offset(offset)
     q = q.limit(limit)
-    
+
     results = []
     for result in q:
         if isinstance(result, tuple) and isinstance(result[0], model.DomainObject):

ckan/plugins/interfaces.py

@@ -231,6 +231,37 @@ def authz_remove_role(self, object_role):
     def delete(self, entity):
         pass
 
+    def before_search(self, search_params):
+        '''
+            Extensions will receive a dictionary with the query parameters,
+            and should return a modified (or not) version of it.
+
+            search_params will include an 'extras' dictionary with all values
+            from fields starting with 'ext_', so extensions can receive user
+            input from specific fields.
+
+        '''
+        return search_params
+
+    def after_search(self, search_results, search_params):
+        '''
+            Extensions will receive the search results, as well as the search
+            parameters, and should return a modified (or not) object with the
+            same structure:
+
+                {'count': '', 'results': '', 'facets': ''}
+
+            Note that count and facets may need to be adjusted if the extension
+            changed the results for some reason.
+
+            search_params will include an 'extras' dictionary with all values
+            from fields starting with 'ext_', so extensions can receive user
+            input from specific fields.
+
+        '''
+
+        return search_results
+
 class IPluginObserver(Interface):
     """
     Plugin to the plugin loading mechanism

ckan/templates/package/search_form.html

@@ -12,6 +12,7 @@
     <input type="hidden" name="${k}" value="${v}" />  
   </py:for>
   </span>
+  <div id="dataset-search-ext"></div>
   <input type="submit" value="${_('Search')}" class="pretty-button primary button" />
 </form>