Welcome Michele!
In ArangoDB, document keys are strings. They must not be numeric, although
it's perfectly fine to turn a number into a string and use that instead
(note the quote marks):
{ "_key": "123" }
In CSV, entries like 123 or 0.1 are actually interpreted as numbers. You
must store handles as strings however. So instead of
Tracks/0001,Tracks/0002,0.1
you should store it like this:
"Tracks/0001","Tracks/0002","0.1"
Here's a sample CSV I successfully imported:
"_from","_to","_key"
"vert/001","vert/002","0.1"
"vert/002","vert/003","0.2"
"vert/003","vert/004","0.3"
Note the blank line at the end. In my test, arangoimp ignored the last row,
so the 3rd edge was not imported. The extra line fixed this.
Also note that this will not create any vertex documents with keys "0.1"
etc. but just the edges (they are dangling edges). As this shows, it is not
checked whether the documents you reference in _from and _to exist. It is
checked whether the referenced vertex collections exist however, which
means there has to be document collection "vert" or you will see a
"collection does not exist" error during import. The command-line option
--create-collection
true will create the target edge collection if necessary, but not
implicitly create those vertex collections.
Hope this helped.
Best, Simran
--
You received this message because you are subscribed to the Google Groups
"ArangoDB" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
For more options, visit https://groups.google.com/d/optout.
