Thanks for letting me know.
- With the movie database:
MATCH (a:Person)-[:ACTED_IN]-(b:Movie) RETURN DISTINCT a.name LIMIT 10 returns:
MATCH (a:Person)-[:ACTED_IN]-(b:Movie) RETURN DISTINCT a.name SKIP 2 LIMIT 5, starts on Laurence Fishburne and stops at Charlize Theron. But this
MATCH (a:Person)-[:ACTED_IN]-(b:Movie) RETURN DISTINCT a.name LIMIT 5 SKIP 2 starts at Emil Eifrem and stops at Keanu Reeves.
And I just noticed that the parser has basically cut the "SKIP 2" part out completely because in the result set window (in the browser), the query that is depicted on the top does not include the "SKIP" clause.
From a parsing point of view, I understand why the clause is ignored but I feel that it should complain because otherwise you still get back results but they are not as expected.
This may have been the byproduct of another error that was not reported as such. The original query that was giving me this error was using
apoc.text.join in conjuction with
split. This was a quick test to "group by" a specific subset of terms contained in a string. In reality, these terms have already been expanded to the graph they represent and the query will be adapted accordingly. The bottom line is that I came across this error trying to split a string and then re-compose a subset of it.
Here is what I am faced with at the moment trying to replicate the error on the Movie db, which is a bit worrying:
CALL apoc.text.phonetic("Alpha") returns
MATCH (a:Person)-[:ACTED_IN]-(b:Movie) RETURN apoc.text.phonetic(a.name) fails with "Unknown function apoc.text.phonetic...."
CALL dbms.procedures() clearly lists
apoc.text.phonetic with two prototypes, one with value::ANY and one with value:STRING and
What I suspect might have happened originally is the
apoc function failing silently and reporting a different error. I will try the query on my data later on again and provide an update.
All the best