handle blank strings in link target field when converting from json to csv

[micahstubbs]

hit this issue when I run: node convert-json-to-csv-for-neo4j-readme-links.js

➜  blockbuilder-graph-search-index git:(master) ✗ sh load-csv-graph-into-neo4j.sh
WARNING: neo4j-import is deprecated and support for it will be removed in a future
version of Neo4j; please use neo4j-admin import instead.

Neo4j version: 3.2.2
Importing the contents of these files into data/databases/blockbuilder-graph-search.db:
Nodes:
  /Users/m/workspace/blockbuilder-graph-search-index/data/csv-graphs-for-neo4j/readme-links-blocks.csv
Relationships:
  /Users/m/workspace/blockbuilder-graph-search-index/data/csv-graphs-for-neo4j/readme-links-relationships.csv

Available resources:
  Total machine memory: 16.00 GB
  Free machine memory: 315.75 MB
  Max heap memory : 3.56 GB
  Processors: 8
  Configured max memory: -3138095923.00 B

Nodes, started 2017-08-14 20:43:10.809+0000
[>:??------------------------------|NODE:7.6|*PROPERTIES-----------------------|v:??----------]10.9K ∆10.9K
Done in 216ms
Prepare node index, started 2017-08-14 20:43:11.075+0000
[*DETECT:7.72 MB------------------------------------------------------------------------------]10.0K ∆10.0K
Done in 43ms
Relationships, started 2017-08-14 20:43:11.127+0000
[*>:??----------------------------------------------------------------------------------------]    0 ∆    0
Done in 23ms
Relationship missing mandatory field 'END_ID', read more about relationship format in the manual:  https://neo4j.com/docs/operations-manual/3.2/tools/import/file-header-format/#import-tool-header-format-rels
Caused by:InputRelationship:
   source: /Users/m/workspace/blockbuilder-graph-search-index/data/csv-graphs-for-neo4j/readme-links-relationships.csv:287
   startNode: 9262169 (global id space)
   endNode: null (global id space)
   type: LINKS_TO
 is missing END_ID field