multi-threaded curl would seem the answer to this.
the view definition itself would enable a binary tree for the searches to
happen quickly, wouldn't it? <asking everyone here>

On Sat, Sep 3, 2022 at 11:29 AM Paul Milner <[email protected]> wrote:

> Hello
>
> I'm looking to execute a view multiple times at once for different keys
> with different limits and skips (associated to each query). So I need the
> following:
>
> POST /recipes/_design/recipes/_view/by_title/queries
> HTTP/1.1Content-Type: application/jsonAccept: application/jsonHost:
> localhost:5984
> {    "queries": [        {            "keys": [
> "meatballs",                "spaghetti"            ]        },
> {            "limit": 3,            "skip": 2        }    ]}
>
> And I get the results object for each query. But can someone tell me
> how I form the queries object for multiple queries please? Something
> like:
>
> {    "queries": [
>
>         {            "keys": [                "meatballs",
>    "spaghetti"            ]        },        {            "limit": 3,
>           "skip": 2        },
>
>         {            "keys": [                "meatballs3",
>     "spaghetti3"            ]        },        {            "limit":
> 4,            "skip": 5        }
>
> ] }
>
>  or should this be done a different way? I'm only doing one key per
> view. So it's a way to execute a query multiple times in bulk with
> different limits and skips and get the array of results for each key.
> I'm looking to do the following:
>
> 1) Get the count of each key and use this to get a random item within
> the duplicates of the key.
>
> 2) Delete a random item with the duplicates of the key.
>
> I see it as a way to prevent concurrence when deleting items with
> multiple concurrent jobs.
>
> I hope that's clear.
>
> Thanks a lot
>
> Best regards
>
> Paul
>

Reply via email to