Thanks for the update Dylan - good to know that those cases I was concerned about are covered by the drivers it seems.
On Fri, Mar 4, 2016 at 8:14 PM, Dylan Millikin <[email protected]> wrote: > Hey, > I'm a little late to the party (still going through all my emails trying to > catch up on what I missed). > Just wanted to say that I do have these tests in the driver's test suit. > Literally all of the cases you mentioned, and they pass for 3.1.1 last I > checked. > > I never check wether or not the type on the server is correct though. I > just save the param to a vertex property and retrieve it, then I check it's > type in PHP. > > There's some lossiness inherent to PHP that I can't test for however. Such > as making sure that map keys typed as a numerical strings are actually > typed as string. This is because php converts {"1":36} to {1:36} > automatically so my tests have no way of knowing if my param {"1":36} is > saved as {string:int} or {int:int}, all I will ever get in native format is > {int:int}. Am I making sense? > > Hopefully that covers the cases you're concerned about. > > On Mon, Feb 15, 2016 at 5:40 PM, David Brown <[email protected]> wrote: > > > This seems to be working well, for example I can send a message like > this: > > > > 'args': {'gremlin': 'x.b'', 'bindings': {'x': {'f': {'foo': 'bar'}, 'b': > > ['bar', None, 1.5, {'b': 1}]}}, aliases': {}, 'language': > > 'gremlin-groovy'}, 'processor': '', 'op': 'eval', 'requestId': > > '7a19dc71-0fec-4468-abb4-587ee8301c44'} > > > > And I get the expected response: > > > > {'status': {'code': 200, 'attributes': {}, 'message': ''}, 'result': > > {'data': ['bar', None, 1.5, {'b': 1}], 'meta': {}}, 'requestId': > > '7a19dc71-0fec-4468-abb4-587ee8301c44'} > > > > > > > > On Sat, Feb 13, 2016 at 8:22 AM, Stephen Mallette <[email protected]> > > wrote: > > > > > Anyone get a chance to actually try this out yet? > > > > > > On Mon, Feb 1, 2016 at 6:45 AM, Stephen Mallette <[email protected] > > > > > wrote: > > > > > > > It would be nice to know if that kind of stuff works. My tests for > that > > > > kind of stuff span various parts of the repo so I can't point you to > > one > > > > test case that does it, but a bunch of them are here: > > > > > > > > > > > > > > > > > > https://github.com/apache/incubator-tinkerpop/tree/b50bb3f7449993bab08f0c0e1014baea8c3afa63/gremlin-driver/src/test/java/org/apache/tinkerpop/gremlin/driver/ser > > > > > > > > On Sat, Jan 30, 2016 at 10:49 AM, Jean-Baptiste Musso < > > [email protected] > > > > > > > > wrote: > > > > > > > >> I didn't add these specific tests yet for gremlin-javascript but > I'll > > > >> do so and let you know how it goes. > > > >> > > > >> Current test suite is located at > > > >> > > > >> > > > > > > https://github.com/jbmusso/gremlin-javascript/tree/78f7d548b83f2469a2fca720587143c9c753197b/test > > > >> > > > >> Jean-Baptiste > > > >> > > > >> On Fri, Jan 29, 2016 at 10:25 PM, Stephen Mallette < > > > [email protected]> > > > >> wrote: > > > >> > If you don't build a driver for Gremlin Server, no need to read > any > > > >> > further.... > > > >> > > > > >> > If you are still reading it means you develop drivers for Gremlin > > > >> Server - > > > >> > can you please comment on whether or not you have tested (unit > > tests, > > > >> > tested manually, whatever....) the sending the following as > > parameter > > > >> > values: > > > >> > > > > >> > 1. List/Map > > > >> > 2. Map of arbitrary depth (i.e. Maps with a Map) > > > >> > 3. List containing Maps/Lists > > > >> > 4. List with mixed types (including Map/List) > > > >> > > > > >> > It would be nice to hear if all that kind of stuff works nicely on > > > your > > > >> > ends. > > > >> > > > > >> > Thanks, > > > >> > > > > >> > Stephen > > > >> > > > > > > > > > > > > > > > > > > > -- > > David M. Brown > > R.A. CulturePlex Lab, Western University > > >
