2010/9/1 Itagaki Takahiro <itagaki.takah...@gmail.com>:
> On Wed, Sep 1, 2010 at 6:07 AM, Pavel Stehule <pavel.steh...@gmail.com> wrote:
>> I don't found a nice mix for placeholders and positional placeholders
>
> How about %pos$format, used in C-printf()? It might be
> only in Linux's libc.
>
> printf("<%2$s> <%1$d>\n", 123, "abc");
> => <abc> <123>
> http://linux.die.net/man/3/printf

same syntax I designed and didn't implement, because it isn't too
readable (my opinion). But I am not against - just thinking so
separate function can be better.
>
>> %i ... sql identifier
>> %v ... sql value
>> %s ... string --- the most used tag I expect
>> %l ... literal
>
> Looks good designed.  I have a couple of comments and questions:
>
> * There is no examples for %l. What's the difference from %v and %s?
>  If it always quotes, how does it work? Like as quote_literal()
>  or quote_nullable()?
>

%l is always quoted

it is designed for syntax: SELECT integer '10';

%s versus %v ... %s is never quoted X %v is quoted when it is
necessary, NULL is showed as empty string for %s and as "NULL" for %v.
%s is used for messages (the behave is same like "concat"), %v is used
for SQL statement building

> * %v quotes text values (and maybe all non-numeric values) with
>  single quotes, but doesn't numeric values. How do we determine
>  the difference? By type oid?
>

every datatype has typecategory attribute

    typname     │ typcategory
─────────────────┼─────────────
 int8            │ N
 int2            │ N
 int4            │ N
 regproc         │ N
 oid             │ N
 float4          │ N
 float8          │ N
 money           │ N
 numeric         │ N
 regprocedure    │ N
 regoper         │ N
 regoperator     │ N
 regclass        │ N
 regtype         │ N
 regconfig       │ N
 regdictionary   │ N
 cardinal_number │ N

so these types are unquoted


> * %v also doesn't quote boolean values, but t and f are not valid.
>  You should use true and false (or 't' and 'f') for the cases.
>  (So, your "INSERT INTO" example is broken.)
>

you have a true - it should be fixed

Regards

Pavel

> --
> Itagaki Takahiro
>

-- 
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

Reply via email to