test fixes

Author: tomkralidis

pycsw/ogc/api/records.py

@@ -647,9 +647,9 @@ def items(self, headers_, json_post_data, args, collection='metadata:main'):
                         collections = ','.join(f'"{x}"' for x in v.split(','))
                     else:
                         collections = ','.join(f'"{x}"' for x in v)
-                    query_args.append(f"parentidentifier IN ({collections})")
+                    query_args.append(f"collection IN ({collections})")
                 elif k == 'anytext':
-                    query_args.append(build_anytext(k, v))
+                    query_args.append(build_anytext(k, v, self.repository))
                 elif k == 'bbox':
                     query_args.append(f'BBOX(geometry, {v})')
                 elif k == 'keywords':
@@ -660,20 +660,24 @@ def items(self, headers_, json_post_data, args, collection='metadata:main'):
                     else:
                         begin, end = v.split('/')
                         if begin != '..':
-                            query_args.append(f'time_begin >= "{begin}"')
+                            query_args.append(f'"properties.datetime" >= "{begin}"')
+                            #query_args.append(f'time_begin >= "{begin}"')
                         if end != '..':
                             query_args.append(f'time_end <= "{end}"')
                 elif k == 'q':
                     if v not in [None, '']:
-                        query_args.append(build_anytext('anytext', v))
+                        query_args.append(build_anytext('anytext', v, self.repository))
                 else:
                     query_args.append(f'{k} = "{v}"')
 
         facets_requested = str2bool(args.get('facets', False))
 
         if collection != 'metadata:main':
             LOGGER.debug('Adding virtual collection filter')
-            query_args.append(f'parentidentifier = "{collection}"')
+            query_args.append(f'collection = "{collection}"')
+
+        if self.repository.dbtype == 'Elasticsearch':
+            query_args = [qa.replace('"', "'") for qa in query_args]
 
         LOGGER.debug('Evaluating CQL and other specified filtering parameters')
         if cql_query is not None and query_args:
@@ -771,10 +775,12 @@ def items(self, headers_, json_post_data, args, collection='metadata:main'):
         offset = int(args.get('offset', 0))
 
         if query_parser is not None and cql_query != {}:
+            LOGGER.debug('Parsing CQL into AST')
             LOGGER.debug('Parsing CQL into AST')
             LOGGER.debug(json_post_data)
             LOGGER.debug(cql_query)
             try:
+                print("CQL QUERY", cql_query)
                 ast = query_parser(cql_query)
                 LOGGER.debug(f'Abstract syntax tree: {ast}')
             except Exception as err:
@@ -1451,13 +1457,13 @@ def record2json(record, url, collection, mode='ogcapi-records'):
 
     return record_dict
 
-
-def build_anytext(name, value):
+def build_anytext(name, value, repository):
     """
     deconstructs free-text search into CQL predicate(s)
 
     :param name: property name
-    :param name: property value
+    :param value: property value
+    :param repository: repository object
 
     :returns: string of CQL predicate(s)
     """
@@ -1470,17 +1476,30 @@ def build_anytext(name, value):
 
     if len(tokens) == 1 and ' ' not in value:  # single term
         LOGGER.debug('Single term with no spaces')
-        return f"{name} ILIKE '%{value}%'"
+        predicates = [f"{name} ILIKE '%{value}%'"]
 
-    for token in tokens:
-        if ' ' in token:
-            tokens2 = token.split()
-            predicates2 = []
-            for token2 in tokens2:
-                predicates2.append(f"{name} ILIKE '%{token2}%'")
+    else:
+        for token in tokens:
+            if ' ' in token:
+                tokens2 = token.split()
+                predicates2 = []
+                for token2 in tokens2:
+                    predicates2.append(f"{name} ILIKE '%{token2}%'")
 
-            predicates.append('(' + ' AND '.join(predicates2) + ')')
-        else:
-            predicates.append(f"{name} ILIKE '%{token}%'")
+                predicates.append('(' + ' AND '.join(predicates2) + ')')
+            else:
+                predicates.append(f"{name} ILIKE '%{token}%'")
+
+    if name == 'anytext' and repository.dbtype == 'Elasticsearch':
+        predicates2 = []
+        for p in predicates:
+            p2 = p
+            predicates2.append(p2.replace(name, '"properties.title"'))
+            predicates2.append(p2.replace(name, '"properties.description"'))
+
+        predicates = predicates2
 
-    return f"({' OR '.join(predicates)})"
+    if len(predicates) == 1:
+        return predicates[0]
+    else:
+        return f"({' OR '.join(predicates)})"

pycsw/plugins/repository/elastic.py

@@ -44,6 +44,8 @@
 class Record(Document):
     identifier = Text()
     parentidentifier = Text()
+#    anytext = Text()
+#    datetime = Text()
 
 
 class ElasticsearchRepository:
@@ -69,17 +71,20 @@ def __init__(self, repo_object: dict, context):
 
         self.query_mappings = {
             # OGC API - Records mappings
+            'geometry': 'geometry',
             'identifier': 'identifier',
             'type': 'type',
             'typename': 'typename',
             'parentidentifier': 'parentidentifier',
+            'collection': 'parentidentifier',
             'collections': 'parentidentifier',
             'updated': 'insert_date',
             'title': 'title',
+            'abstract': 'abstract',
             'description': 'abstract',
             'keywords': 'keywords',
             'edition': 'edition',
-            'anytext': 'anytext',
+            'anytext': 'title',
             'bbox': 'wkt_geometry',
             'date': 'date',
             'datetime': 'date',
@@ -184,7 +189,7 @@ def query_collections(self, filters=None, limit=10) -> list:
             raise RuntimeError(msg)
 
         for hit in response['hits']['hits']:
-            if hit['_source']['type'] in ['Collection', 'stac:Collection']:
+            if hit['_source']['type'] == 'Collection':
                 dataset = type('dataset', (object,), hit['_source'])
                 dataset.identifier = dataset.id
                 dataset.title = hit['_source'].get('title')
@@ -288,8 +293,9 @@ def query(self, constraint=None, sortby=None, typenames=None,
 
         if constraint.get('ast') is not None:
             LOGGER.debug('Applying filter')
+
             es_query = es_query.query(
-                to_filter(constraint['ast'], {'ows:BoundingBox': 'geometry'},
+                to_filter(constraint['ast'], self.query_mappings,
                           version=ES_VERSION)).extra(track_total_hits=True)
 
         es_response = es_query.execute()