Re: Proposed p.tch for error locations

Tom Lane wrote:
> Alvaro Herrera <alvherre@xxxxxxxxxxxxxxxxx> writes:
> > This doesn't work on terminal using a variable-width font, does it?
> > Sure, you can have all the interfaces use a monospace font, but it seems
> > a weird thing to do.  I think the line/character position should be
> > returned in a separate error attribute in ereport.  So for example
> > pgAdmin could count characters and mark it in bold or use a different
> > color.
> 
> That information is already available to pgAdmin, and has been since
> 7.4; if they are failing to exploit it that's their problem not libpq's.

Aye, that's fine.  I don't really know if pgAdmin uses the info or not.
I thought the proposal to remove the "at character N" in the error
message meant that the only way to get that information would be from
the "LINE Y: ...\n[some whitespace]^" message, which would be pretty
cumbersome.  But if there already is a field in the error message for
the position, then nothing is lost.

-- 
Alvaro Herrera                                http://www.CommandPrompt.com/
The PostgreSQL Company - Command Prompt, Inc.

---------------------------(end of broadcast)---------------------------
TIP 9: In versions below 8.0, the planner will ignore your desire to
       choose an index scan if your joining column's datatypes do not
       match