Re: Bug in ILIKE? - Mailing list pgsql-hackers

From Tom Lane
Subject Re: Bug in ILIKE?
Date
Msg-id 7006.1222399050@sss.pgh.pa.us
Whole thread Raw
In response to Re: Bug in ILIKE?  (Andrew Dunstan <andrew@dunslane.net>)
Responses Re: Bug in ILIKE?  (Andrew Dunstan <andrew@dunslane.net>)
List pgsql-hackers
Andrew Dunstan <andrew@dunslane.net> writes:
> The docs actually don't state what are the semantics of escape followed 
> by something that is not escape or a metachar. Does the spec say 
> anything about that?

The spec says it's an error, per the SQL92 excerpt I quoted in the
previous thread.  (SQL99 says about the same with more notation;
I didn't bother looking in the later specs.)

I find that position too restrictive, mainly because of this
consideration: suppose some future version of the spec invents
additional metacharacters.  To be concrete, suppose ? means
something special in SQL2010.  Now how do you make a pattern that
works in both older and newer servers?  \? means literal ? to the
newer server, but if it throws an error on the older, you're stuck.

So I'm for the definition that escape-anything means exactly anything,
without any special treatment that it would otherwise have.  And in
the case of ILIKE it seems like "no special treatment" should mean
"case insensitive match".
        regards, tom lane


pgsql-hackers by date:

Previous
From: "Robert Haas"
Date:
Subject: Re: Updates of SE-PostgreSQL 8.4devel patches
Next
From: KaiGai Kohei
Date:
Subject: Re: Updates of SE-PostgreSQL 8.4devel patches