Search API Shortcomings
I am using the API to find specific files within a know folder. The files follow the naming convention 1-1.tif, 1-2.tif, 1-3.tif, and so on, for numbers between 1 to 2500. Sometimes there may be 3 files per number, other times more or less. The API seems severely limited in this type of application. I have filter thousands of search results to find matches for a file name which starts with 1 or 2. Primary, Box appears to completely ignore the dash. Searching for 1- is the same as searching for 1. Secondarily, There does not seem a way to filter the search pattern to indicate that the file name starts a 1, as opposed to contains a 1 somewhere in the name. Is there any way to limit the number of potential matches in this application for single or double digit searches?
-
Hi ,
Thanks for the feedback on search, and sorry for the experiences that you've had thus far from it. I'll try to be as transparent as possible with where we're at and what you can do now, and where we're going.
The current search endpoint does have some limitations, as you've outlined. Basically you'll be able to use the filtering mechanisms that are outlined here, and I might suggest taking a loot into the possibilities of using metadata filtering as an option for getting back better result sets, or confining the search to a time range for smaller result sets. With that said, it might be valuable to add a request in Box Pulse for better search features, as these requests are managed by the PMs who help scope the roadmap.
Now as far as where we're going, we know there are deficiencies in the capabilities of search, and are actively working towards better options for providing granular level searching in different ways, so there is work being invested towards improving the feature.
I know these answers are not ideal, but I hope they help provide some context.
Thanks,
Jon
Please sign in to leave a comment.
Comments
1 comment