skip to navigation
skip to content

AjguDB 0.8.1

Explore you connected data

AjguDB wants to be a fast enough python graph database for exploring connected data.

  • graphdb
  • pythonic
  • schemaless
  • networkless
  • with gremlin querying
  • GPLv2 or GPLv3


Check out the presentation @ [speakerdeck](



  • improve documentation
  • fix scatter step
  • improve test coverage (89%)
  • more experience with conceptnet


  • move to tuple space implementation
  • work with wiredtiger develop branch


  • small typofix in README…


  • storage: only wiredtiger 2.6.1 backend storage is supported
  • ajgudb: you can threat AjguDB as simple key/value store via its methods AjguDB.get(key), AjguDB.set(key, value) and AjguDB.remove(key)
  • rewrite gremlin querying
  • gremlin: select is renamed where because it match the SQL terminology. SQL’s FROM is vertices and edges steps.
  • storage: rework the backend to use less write and similar read operations count.
  • storage: Now edges and vertices are stored in different tables this might also lead to performance improvement during querying.
  • storage: elements identifiers are now computed by the backend storage, wiredtiger.
  • add fuzzy search


  • ajgudb
    • add bsddb backend
    • add wiredtiger backend
    • leveldb: increase block size to 1GB
  • gremlin:
    • add keys to retrieve several keys at the same time
    • use lazy itertools.imap instead of the gready python2’s map


  • ajgudb:
    • add a shortcut method**kwargs) to query for one element.
  • gremlin:
    • fix group_count, now it’s a step and not a final step
    • fix each to return GremlinResult so that history is not lost and back can be used
    • add scatter, it’s only useful after group_count so far.

Coverage: 62%


Say hi!

File Type Py Version Uploaded on Size
AjguDB-0.8.1.tar.gz (md5) Source 2016-08-28 14KB