Changes between Initial Version and Version 1 of Ticket #5392, comment 5
- Timestamp:
- 06/03/23 12:13:03 (19 months ago)
Legend:
- Unmodified
- Added
- Removed
- Modified
-
Ticket #5392, comment 5
initial v1 27 27 28 28 select st_union(a.geom, b.geom) 29 from polys a30 cross join polys b31 where a.name = 'kpc' and b.name = 'tonga';29 from polys a 30 cross join polys b 31 where a.name = 'kpc' and b.name = 'tonga'; 32 32 }}} 33 33 … … 38 38 {{{ 39 39 insert into polys 40 select name || ' valid', st_makevalid(geom) as geom41 from polys42 where name in ('kpc', 'tonga');40 select name || ' valid', st_makevalid(geom) as geom 41 from polys 42 where name in ('kpc', 'tonga'); 43 43 44 44 select st_area(st_union(geom)) from polys where name ~ 'valid'; 45 45 46 46 select st_area(st_union(a.geom, b.geom)) 47 from polys a48 cross join polys b49 where a.name = 'kpc valid'50 and b.name = 'tonga valid'47 from polys a 48 cross join polys b 49 where a.name = 'kpc valid' 50 and b.name = 'tonga valid' 51 51 }}} 52 52 … … 55 55 {{{ 56 56 select st_area(st_union(a.geom, b.geom)) 57 from polys a58 cross join polys b59 where a.name = 'kpc'60 and b.name = 'tonga'57 from polys a 58 cross join polys b 59 where a.name = 'kpc' 60 and b.name = 'tonga' 61 61 }}} 62 62 … … 64 64 65 65 I sometimes wonder if we should do automagical makevalid as part of the try/catch block on the overlay code, so the case of the TopologyException would end up actually working, automagically (catch exception, make valid, retry). But seeing the silent failure of the pairwise code makes me question that. Automagical makevalid would make those cases of silent failure even more likely to just slide by people's notice. 66 67 68