./build.sh

./run.sh <new store dir> <nodes csv> <relationships csv>

Plain CSV with a leading header line. The header line supports special annotations for property types and indexing.

Note: values with commas and/or quoted values are NOT supported right not!

The node format:

id[,[indexname|]propertyname[@type]]*

The relationship format:

from,to,type[,[indexname|]propertyname[@type]]*

Here’s a nodes input file for nodes with two properties, the first of which is indexed in the “users” index, and the
second of which is a long number:

id,users|username,age@long

1,Dave,27

2,Mike,21

And here’s a file connecting these two users, with a timestamp on the relationship:

from,to,type,since@long

1,2,KNOWS,13242352355

Sparse property values are also supported:

id,firstname,make,year

1,Dave,,

2,,Honda,2004

from,to,type
1,2,OWNS