Open russellseymour opened 11 years ago
Hey, sorry for the slow reply. Been travelling a lot for work.
1) Unfortunately, there isn't a way to define an entire mapping with an array right now. You can do per-type arrays:
$mapping = \Sherlock\Sherlock::mappingBuilder('settings') -> String($myArray);
But cannot supply an array for the entire mapping. I agree that this would be a useful feature, but it doesn't exist yet :(
2) Theoretically, you should be able to do something like this:
$index = $sherlock->index('test_mappings');
$index->type("new_type")->mappings($mapping);
$response = $index->updateMapping();
If you're using syntax like that and its throwing errors, let me know - it's probably a bug.
3) No list of breaking changes for 0.2, but it will be considerable. Much of the API is being reorganized to be more consistent. "Leaf API" calls (like each individual query type) will remain unchanged, but the query-builder part to generate complex queries is changing a lot. That said, I don't see 0.2 being released for a few more months. It's temporarily on hold due to waiting for an third-party library (which is going to be released soon). Once that library is released, I can resume work on the 0.2 branch.
Good morning,
Last night I stumbled across the SherlockPHP library and I think it will be very good for my application. I am looking to move from using a combination of Mongo and Solr to just using ElasticSearch and I have been looking for a library to help me with this and I think I may have found it :-).
I wonder if you could help me with a couple of things:
But this came back with an 'Undefined offset' error, which I suspect is to do with the way I applied the mapping to the index.
Finally it is stated on the GitHub page that version 0.2 will be released soon and it has some breaking changes, are these documented anywhere? I want to use this library but may hold off for a while of there are lots of breaking changes.
Lastly I apologise for cross posting this to Google Groups and GutHub issues, I was not sure which one would be more active.
Thanks very much, Russell