I was thinking about that. Depends on what you mean by par. If you are not indexing array values, then you will have to touch every array element each time you query. It's back to one pass sql conversion then, and you would get all the benefits of sql in the bargain. The only time an array query api comes in handy is if you are only trying to filter an array for a key:value a few times. If you have to do this many times, conversion to a memory based sqlite database could really pay dividends for a large array, especially if you took the time to index the columns that mattered most.
I'm going to put together a benchmark soon I think. Shouldn't take too much time to do so. I have some good SQL data I can punch into an array then create a memory based SQL database from. Bob S > On Apr 18, 2018, at 10:08 , Richard Gaskin via use-livecode > <use-livecode@lists.runrev.com> wrote: > > Alex Tweedly wrote: > > > I expect the searches etc. would be faster with SQL / memory database > > Both use hashed lookups to find data. What would be needed to bring LC array > queries on par with SQLite? > > -- > Richard Gaskin _______________________________________________ use-livecode mailing list use-livecode@lists.runrev.com Please visit this url to subscribe, unsubscribe and manage your subscription preferences: http://lists.runrev.com/mailman/listinfo/use-livecode
