Fetching contributors…
Cannot retrieve contributors at this time
89 lines (74 sloc) 2.9 KB

Release Notes

Version 1.7

  • Added null handler for logging
  • Upgrade to HTTPStream 1.2.1
  • Renamed GraphDatabaseService to Graph and created deprecated alias
  • Renamed all graph_db arguments and attributes to graph (including Resource and ServiceRoot)
  • Removed support for Python 2.6

Version 1.6.1

  • Cypher transactions
  • Implicit authentication when user info passed in URI
  • Improved reconnection handling on MacOS
  • Renamed schema.get_index to 'get_indexed_property_keys'
  • Miscellaneous bug fixes (see issue list for further details)

Version 1.6


  • HTTP/REST transport layer rewritten to use HTTPStream library
  • Minimum supported server version increased to 1.8
  • Changed some server calls to be "lazy", particularly on creation of GraphDatabaseService
  • Added feature detection methods (GraphDatabaseService.supports_*)
  • Changed Node.exists and Relationship.exists from function to property
  • Changed Node.is_abstract and Relationship.is_abstract from function to property
  • Changed and to '_id'
  • Changed GDB.order and size from function to property
  • Changed Path.order and size from function to property
  • Removed all previously deprecated features
  • DEFAULT_URI now points to service root, not graph database (i.e. without trailing /db/data/)


  • Deprecated cypher module in favour of CypherQuery class
  • Revised execution methods: run, execute, execute_one, stream
  • Results returns as CypherResults or IterableCypherResults


  • Batch methods return BatchRequest object for use in other requests
  • Added create_path and get_or_create_path to WriteBatch
  • Revised all WriteBatch methods for better consistency
  • Improved comments and documentation on batch classes
  • Batches no longer auto-clear on submission
  • Removed ReadBatch.get_properties, use GraphDatabaseService.get_properties instead
  • Cypher queries now supported in batches (append_cypher method)
  • Revised execution methods: run, stream, submit


  • Changed all match methods to return iterators
  • Removed bidirectional argument from Node.match (now always bidirectional)
  • Added Node.match_incoming
  • Added Node.match_outgoing
  • Removed Node.match_one

Labels & Schema Indexes

  • Label support in Node and WriteBatch classes
  • GraphDatabaseService.find for iterating through labelled nodes
  • Added schema resource to GraphDatabaseService


  • Changed GraphDatabaseService.delete_index to throw LookupError if not found
  • Index.query now returns iterator instead of list


  • Command line scripts for cypher/geoff removed in favour of unified neotool
  • Changed neotool text output to match minimal format used by PostgreSQL
  • Neotool now fully supports unicode


  • Removed Gremlin module


  • Monitor class moved from admin module into neo4j module, admin module removed