Just in time for GraphConnect 2081 in NYC, we are happy to release APOC 3.4.0.3.

Thanks so much to all the contributors that have been really busy.

This release has some neat new features.

We added support for defining custom procedures and functions implemented in plain Cypher and then call them like regular ones.

CALL apoc.custom.asProcedure('neighbours',
  'MATCH (n:Person {name:$name})-->(nb) RETURN nb as neighbour','read',
  [['neighbour','NODE']],[['name','STRING']]);

CALL custom.neighbours('Keanu Reeves') YIELD neighbour;

You can now load data from inside of zip archives by appending the !path/to/file.csv to your file url.

To scrape data from websites you can use apoc.load.html, which is really useful.

The warmup.run() procedure was rewritten to also allow warm-up of index files and others.

You can now stream the results of apoc.export.csv.* directly to the client within your Cypher result instead writing a file on the server.

Explicit index procedures can now return the number of hits instead of the data.

Date formatting can now use named formats, something we got inspiration for from ElasticSearch.

Couchbase (4.x and 5.x) new authentication methods are now supported.

With apoc.text.index(es)Of you can get one or more occurrences of substrings.

There were a lot of docs updates, for JDBC, text, map, XML, schema and much more.

Please try out the new features and report back how well they work for you.

You can also ask questions around APOC in the new Neo4j Community Forum