[pgrouting-users] ultimate solution to assign_vertex_id() bug :-)

[Daniel Kastl]

On Wed, Oct 5, 2011 at 12:22 PM, Daniel Kastl <daniel at georepublic.de> wrote:

>
>
> On Wed, Oct 5, 2011 at 12:07 PM, Stephen Woodbridge <
> woodbri at swoodbridge.com> wrote:
>
>> YuLongzhen,
>>
>> Thank you for your report. I think your analysis and solution are correct.
>>
>> Daniel or Anton,
>>
>> quote_ident() should only be used when quoting a column or table when it
>> is actually being used as a column or table object in the SQL and not when
>> it is being used as a string. But you guys know that and I assume this is
>> just a brain slip when editing the script. Anyway, please update git when
>> you have a chance.
>>
>> Thanks,
>>  -Steve
>
>
>
> Hi YuLongzhen and Steve,
>
> As far as I remember assign_vertex_id has almost not been modified since it
> was part of pgDijkstra many years ago.
> So thank you for looking into this, YuLongzhen! Nobody has ever notified
> this (and reported).
>
> YuLongzhen, if you want to clone the Github repository, make the
> modifications and submit a "pull request", then the credits for the patch
> will be yours.
>
> Daniel
>


Or quote_literal() might be even better:
http://www.postgresql.org/docs/8.4/static/plpgsql-statements.html#PLPGSQL-QUOTE-LITERAL-EXAMPLE

...
FOR _r IN EXECUTE 'SELECT srid FROM geometry_columns WHERE f_table_name='||
quote_literal(geom_table)||';' LOOP
 srid := _r.srid;
    END LOOP;
...

Or am I wrong?
I have never really taken note much at these string functions, though there
seem to be many useful ones:
http://www.postgresql.org/docs/8.4/static/functions-string.html

Daniel


-- 
Georepublic UG & Georepublic Japan
eMail: daniel.kastl at georepublic.de
Web: http://georepublic.de
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.osgeo.org/pipermail/pgrouting-users/attachments/20111005/95505dc8/attachment.html