Updated geoff.py to support relationship index entries

examples/example1.py

@@ -6,7 +6,7 @@
 
 # Allow the import path to access neo4j modules
 import sys
-sys.path.append("../src") 
+sys.path.append("../src")
 
 from py2neo import neo4j
 

examples/geoff/load.py

@@ -0,0 +1,18 @@
+#!/usr/bin/env python
+
+"""
+Script to load sample GEOFF file into database
+"""
+
+from py2neo import neo4j
+
+# Attach to the graph db instance
+gdb = neo4j.GraphDatabaseService("http://localhost:7474/db/data")
+
+# Obtain a link to the reference node of the db
+ref_node = gdb.get_reference_node()
+
+# Load the royal family into the database
+handle = geoff.load(file("royals.geoff"), gdb)
+print "New data available from node %s" % (handle)
+

examples/royals.geoff → examples/geoff/royals.geoff

@@ -1,4 +1,6 @@
 # This data represents a portion of the British Royal Family
+
+# First, we define the nodes (in this case, people)
 (Liz)	{"name": "Elizabeth", "title": "Queen of the Commonwealth Realms", "birth.date": "1926-04-21"}
 (Phil)	{"name": "Philip", "title": "Duke of Edinburgh", "birth.date": "1921-06-21"}
 (Chaz)	{"name": "Charles", "title": "Prince of Wales", "birth.date": "1948-11-14"}
@@ -7,15 +9,19 @@
 (Wills)	{"name": "William", "title": "Duke of Cambridge", "birth.date": "1982-06-21"}
 (Harry)	{"name": "Henry", "birth.date": "1984-09-15"}
 (Kate)	{"name": "Catherine", "title": "Duchess of Cambridge", "birth.date": "1982-01-09"}
-(Liz)-[:MARRIED]->(Phil)    {"marriage.place": "Westminster Abbey", "marriage.date": "1947-11-20"}
+
+# Next come the realtionships between the people
+(Liz)-[1947:MARRIED]->(Phil)    {"marriage.place": "Westminster Abbey", "marriage.date": "1947-11-20"}
 (Liz)-[:MOTHER_OF]->(Chaz)
 (Phil)-[:FATHER_OF]->(Chaz)
-(Di)-[:MARRIED]->(Chaz)     {"marriage.place": "St Paul's Cathedral", "marriage.date": "1981-07-29", "divorce.date": "1996-08-28"}
+(Di)-[1981:MARRIED]->(Chaz)     {"marriage.place": "St Paul's Cathedral", "marriage.date": "1981-07-29", "divorce.date": "1996-08-28"}
 (Di)-[:MOTHER_OF]->(Wills)
 (Chaz)-[:FATHER_OF]->(Wills)
 (Di)-[:MOTHER_OF]->(Harry)
-(Cam)-[:MARRIED]->(Chaz)    {"marriage.place": "Windsor Guildhall", "marriage.date": "2005-04-09"}
-(Kate)-[:MARRIED]->(Wills)  {"marriage.place": "Westminster Abbey", "marriage.date": "2011-04-29"}
+(Cam)-[2005:MARRIED]->(Chaz)    {"marriage.place": "Windsor Guildhall", "marriage.date": "2005-04-09"}
+(Kate)-[2011:MARRIED]->(Wills)  {"marriage.place": "Westminster Abbey", "marriage.date": "2011-04-29"}
+
+# And finally the index entries for the nodes and relationships
 {People}->(Liz)     {"name": "Elizabeth"}
 {People}->(Phil)    {"name": "Philip"}
 {People}->(Chaz)    {"name": "Charles"}
@@ -24,4 +30,8 @@
 {People}->(Wills)   {"name": "William"}
 {People}->(Harry)   {"name": "Henry"}
 {People}->(Kate)    {"name": "Catherine"}
+{Weddings}->[1947]  {"year": 1947}
+{Weddings}->[1981]  {"year": 1981}
+{Weddings}->[2005]  {"year": 2005}
+{Weddings}->[2011]  {"year": 2011}
 

src/py2neo/geoff.py

@@ -81,7 +81,9 @@ def load(self):
 		first_node_id = None
 		nodes = {}
 		rels = []
-		index_entries = {}
+		named_rels = {}
+		node_index_entries = {}
+		rel_index_entries = {}
 		line_no = 0
 		for line in self.file:
 			# increment line no and trim whitespace from current line
@@ -101,12 +103,15 @@ def load(self):
 				)
 				if start_node not in nodes or end_node not in nodes:
 					raise ValueError("Invalid node reference on line %d: %s" % (line_no, repr(m.group(1))))
-				rels.append({
+				rel = {
 					'start_node': start_node,
 					'end_node': end_node,
 					'type': type,
 					'data': json.loads(m.group(8) or 'null')
-				})
+				}
+				if len(name or "") > 0:
+					named_rels[name] = len(rels)
+				rels.append(rel)
 				continue
 			# secondly, try as a node descriptor
 			if m:
@@ -116,20 +121,35 @@ def load(self):
 				nodes[node_id] = json.loads(m.group(8) or 'null')
 				first_node_id = first_node_id or node_id
 				continue
-			# neither of those, so try as an index entry descriptor
+			# neither of those, so try as a node index entry descriptor
 			m = self.NODE_INDEX_ENTRY_PATTERN.match(line)
 			if m:
 				(index, node) = (
 					unicode(m.group(2)),
 					unicode(m.group(3))
 				)
 				data = json.loads(m.group(5) or 'null')
-				if index not in index_entries:
-					index_entries[index] = {}
-				if node not in index_entries[index]:
-					index_entries[index][node] = {}
+				if index not in node_index_entries:
+					node_index_entries[index] = {}
+				if node not in node_index_entries[index]:
+					node_index_entries[index][node] = {}
 				if data:
-					index_entries[index][node].update(data)
+					node_index_entries[index][node].update(data)
+				continue
+			# or as a relationship index entry descriptor
+			m = self.RELATIONSHIP_INDEX_ENTRY_PATTERN.match(line)
+			if m:
+				(index, rel) = (
+					unicode(m.group(2)),
+					unicode(m.group(3))
+				)
+				data = json.loads(m.group(5) or 'null')
+				if index not in rel_index_entries:
+					rel_index_entries[index] = {}
+				if rel not in rel_index_entries[index]:
+					rel_index_entries[index][rel] = {}
+				if data:
+					rel_index_entries[index][rel].update(data)
 				continue
 			# no idea then... this line is invalid
 			raise ValueError("Cannot parse line %d: %s" % (line_no, repr(line)))
@@ -153,16 +173,26 @@ def load(self):
 			}
 			for rel in rels
 		])
-		# create index entries
-		if len(index_entries) > 0:
-			for index_key in index_entries.keys():
+		# create node index entries
+		if len(node_index_entries) > 0:
+			for index_key in node_index_entries.keys():
 				index = self.gdb.get_node_index(index_key)
 				index.start_batch()
-				for node_key in index_entries[index_key].keys():
+				for node_key in node_index_entries[index_key].keys():
 					node = nodes[node_key]
-					for (key, value) in index_entries[index_key][node_key].items():
+					for (key, value) in node_index_entries[index_key][node_key].items():
 						index.add(node, key, value)
 				index.submit_batch()
+		# create relationship index entries
+		if len(rel_index_entries) > 0:
+			for index_key in rel_index_entries.keys():
+				index = self.gdb.get_relationship_index(index_key)
+				index.start_batch()
+				for rel_key in rel_index_entries[index_key].keys():
+					rel = rels[named_rels[rel_key]]
+					for (key, value) in rel_index_entries[index_key][rel_key].items():
+						index.add(rel, key, value)
+				index.submit_batch()
 		return nodes[first_node_id]
 
 

src/py2neo/neo4j.py

@@ -915,14 +915,14 @@ def add(self, entity, key, value):
 		if self._batch is None:
 			self._post(self._template_uri.format(
 				key=key,
-				value=quote(value, "")
+				value=quote(value, "") if isinstance(value, basestring) else value
 			), entity._uri)
 		else:
 			self._batch.append({
 				"method": "POST",
 				"to": self._relative_template_uri.format(
 					key=key,
-					value=quote(value, "")
+					value=quote(value, "") if isinstance(value, basestring) else value
 				),
 				"body": entity._uri,
 				"id": len(self._batch)
@@ -974,7 +974,7 @@ def search(self, key, value):
 						'method': 'GET',
 						'to': self._relative_template_uri.format(
 							key=key,
-							value=quote(value[i], "")
+							value=quote(value[i], "") if isinstance(value[i], basestring) else value[i]
 						),
 						'id': i
 					}
@@ -991,7 +991,7 @@ def search(self, key, value):
 				)
 				for item in self._get(self._template_uri.format(
 					key=key,
-					value=quote(value, "")
+					value=quote(value, "") if isinstance(value, basestring) else value
 				))
 			]