Details
-
Bug
-
Status: Closed
-
Critical
-
Resolution: Invalid
-
0.7.0
-
None
Description
I'm hoping that this is the bug that is buggering up the testCLI. From what I can see in the above class
61 ToolRunner.main( 62 String.format( 63 "crawler -f nquads --maxpages 50 --maxdepth 1 --politenessdelay 500 -o %s " + 64 "http://eventiesagre.it/", 65 outFile.getAbsolutePath() 66 ).split(" ")
There is no such option/switch on the crawler CLI for a -f parameter, therefore based on the hardcoded nature of the test, at no stage does the crawler become aware of the .nq logic. We could overcome this by changing the logic in line 91 of the same class to use RDFUtils.getRDFParser() instead.
This however does not address the point that we have either a bogus or else missing CLI parameter which should define the type of document we are trying to fetch.
On the other hand, please excuse me if this is a wrong assumption altogether as it's been a long day and I'm calling it to a close just now. However I hope this is not the case...