[postgis-tickets] [PostGIS] #2260: Benchmarking speed between built-in tiger normalizer and pagc_address_parser

[PostGIS]

#2260: Benchmarking speed between built-in tiger normalizer and
pagc_address_parser
---------------------------------+------------------------------------------
 Reporter:  robe                 |       Owner:  robe         
     Type:  task                 |      Status:  new          
 Priority:  medium               |   Milestone:  PostGIS 2.1.0
Component:  pagc_address_parser  |     Version:  trunk        
 Keywords:                       |  
---------------------------------+------------------------------------------

Comment(by robe):

 Well I guess in theory I can wrap my (SELECT .. FROM somelex) As lextab
 and that will still work with your above signatures.  Only issue I see
 with taking just the name is a user may add other side line columns that
 just add extra baggage to your query e.g. my is_custom or whatever column.

 On other note and can't assume you can use the same set.

 The reason I say you need to check is that it is quite possible (though
 rare except for someone as perverted as me :) ) that someone might do
 something like this:


 {{{SELECT a.id, standardize_address(c.country_lex, c.country_gaz,
 c.country_rule, a.micro, a.macro)
 FROM  myaddresses AS a LEFT JOIN country_config As c ON (a.country =
 c.country);

 }}}

 And will be very puzzled when the wrong parse tables are used.  It might
 be slow except if I order by country so all records of same country use
 the same parser sets.

-- 
Ticket URL: <http://trac.osgeo.org/postgis/ticket/2260#comment:25>
PostGIS <http://trac.osgeo.org/postgis/>
The PostGIS Trac is used for bug, enhancement & task tracking, a user and developer wiki, and a view into the subversion code repository of PostGIS project.