public inbox for [email protected]  

help / color / mirror / Atom feed

From: Tomas Vondra <[email protected]>
To: Zhihong Yu <[email protected]>
Cc: PostgreSQL Hackers <[email protected]>
Cc: Nikita Glukhov <[email protected]>
Cc: Mahendra Thalor <[email protected]>
Cc: Oleg Bartunov <[email protected]>
Subject: Re: Collecting statistics about contents of JSONB columns
Date: Thu, 6 Jan 2022 20:22:33 +0100
Message-ID: <[email protected]> (raw)
In-Reply-To: <CALNJ-vT8qwNd46Qf+1hSiR+9rz=Sv9w-PQabiwbwM-=CgixyZw@mail.gmail.com>
References: <[email protected]>
	<CALNJ-vQ7p=ZaZ=prwU=tV3FqXhF6CrW+u2Wve3iJVn2wGMBS8A@mail.gmail.com>
	<CALNJ-vT8qwNd46Qf+1hSiR+9rz=Sv9w-PQabiwbwM-=CgixyZw@mail.gmail.com>

On 1/1/22 22:16, Zhihong Yu wrote:
> Hi,
> 
> +static JsonPathStats
> +jsonStatsFindPathStats(JsonStats jsdata, char *path, int pathlen)
> 
> Stats appears twice in the method name. I think findJsonPathStats() 
> should suffice.
> It should check `if (jsdata->nullfrac >= 1.0)` 
> as jsonStatsGetPathStatsStr does.
> 
> +JsonPathStats
> +jsonStatsGetPathStatsStr(JsonStats jsdata, const char *subpath, int 
> subpathlen)
> 
> This func can be static, right ?
> I think findJsonPathStatsWithPrefix() would be a better name for the func.
> 
> + * XXX Doesn't this need ecape_json too?
> + */
> +static void
> +jsonPathAppendEntryWithLen(StringInfo path, const char *entry, int len)
> +{
> +   char *tmpentry = pnstrdup(entry, len);
> +   jsonPathAppendEntry(path, tmpentry);
> 
> ecape_json() is called within jsonPathAppendEntry(). The XXX comment can 
> be dropped.
> 
> +jsonPathStatsGetArrayIndexSelectivity(JsonPathStats pstats, int index)
> 
> It seems getJsonSelectivityWithArrayIndex() would be a better name.
> 

Thanks. I'll think about the naming changes.

> +       sel = scalarineqsel(NULL, operator,
> +                           operator == JsonbGtOperator ||
> +                           operator == JsonbGeOperator,
> +                           operator == JsonbLeOperator ||
> +                           operator == JsonbGeOperator,
> 
> Looking at the comment for scalarineqsel():
> 
>   *  scalarineqsel       - Selectivity of "<", "<=", ">", ">=" for scalars.
>   *
>   * This is the guts of scalarltsel/scalarlesel/scalargtsel/scalargesel.
>   * The isgt and iseq flags distinguish which of the four cases apply.
> 
> It seems JsonbLtOperator doesn't appear in the call, can I ask why ?
> 

Because the scalarineqsel signature is this

     scalarineqsel(PlannerInfo *root, Oid operator, bool isgt, bool iseq,
                   Oid collation,
                   VariableStatData *vardata, Datum constval,
                   Oid consttype)

so

     /* is it greater or greater-or-equal */
     isgt = operator == JsonbGtOperator ||
            operator == JsonbGeOperator

     /* is it equality? */
     iseq = operator == JsonbLeOperator ||
            operator == JsonbGeOperator,

So I think this is correct. A comment explaining this would be nice.


regards

-- 
Tomas Vondra
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company

view thread (3+ messages)

reply

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Reply to all the recipients using the --to and --cc options:
  reply via email

  To: [email protected]
  Cc: [email protected], [email protected], [email protected], [email protected], [email protected], [email protected]
  Subject: Re: Collecting statistics about contents of JSONB columns
  In-Reply-To: <[email protected]>

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

This inbox is served by agora; see mirroring instructions
for how to clone and mirror all data and code used for this inbox