[postgis-users] PostGIS manager: importing multilinestring

[Paolo Cavallini]

Hi all.
While importing a shapefile into postgis, I see that the command contains 2 "u" that
I do not understand:

['shp2pgsql', '-c', '-s', '3003',
u'/home/paolo/Desktop/Dati_corso_QGIS/strade_regionali.shp', u'public.strade']

Furthermore, importing a linear vector throws an error:

Something's wrong:
['Shapefile type: Arc\n', 'Postgis type: MULTILINESTRING[2]\n']

If I select "Create single-part geometries instead of multi-part", I get instead:

syntax error at or near "''"
RIGA 2: ...','0','0','0','0.00000','0','0',NULL,NULL,NULL,NULL,NULL'')

I suspect this is due to some special character in the dbf; the same command crashes
shp2pgsql from the command line:

...
INSERT INTO "public"."strade"
("fnode_","tnode_","lpoly_","rpoly_","length","statali_","statali_id","codice","nome","tipo","codifica",the_geom)
VALUES ('0','0','0','0','0.00000','0','0',NULL,NULL,NULL,NULL,NULL�9      ��9     ');
Segmentation fault

On another machine, I get a more explicit error:

Something's wrong:
['Shapefile type: Arc\n', 'Postgis type: MULTILINESTRING[2]\n', '*** glibc detected
*** shp2pgsql: free(): invalid pointer: 0x00007fce65fc7e60 ***\n', '=======
Backtrace: =========\n', '/lib/libc.so.6[0x7fce65ce8d16]\n',
'/lib/libc.so.6(cfree+0x6c)[0x7fce65ced9bc]\n', 'shp2pgsql[0x40ab82]\n',
'shp2pgsql[0x40b77e]\n', '/lib/libc.so.6(__libc_start_main+0xfd)[0x7fce65c96abd]\n',
'shp2pgsql[0x402969]\n', '======= Memory map: ========\n', '00400000-00436000 r-xp
00000000 08:01 1190782                            /usr/bin/shp2pgsql\n',
'00635000-00636000 rw-p 00035000 08:01 1190782
/usr/bin/shp2pgsql\n', '01767000-01788000 rw-p 00000000 00:00 0
            [heap]\n', '7fce60000000-7fce60021000 rw-p 00000000 00:00 0 \n',
'7fce60021000-7fce64000000 ---p 00000000 00:00 0 \n', '7fce655dd000-7fce655f3000 r-xp
00000000 08:01 1042654                    /lib/libgcc_s.so.1\n',
'7fce655f3000-7fce657f2000 ---p 00016000 08:01 1042654
/lib/libgcc_s.so.1\n', '7fce657f2000-7fce657f3000 rw-p 00015000 08:01 1042654
            /lib/libgcc_s.so.1\n', '7fce657f3000-7fce657f5000 r-xp 00000000 08:01
100246                     /usr/lib/gconv/CP1252.so\n', '7fce657f5000-7fce659f4000
---p 00002000 08:01 100246                     /usr/lib/gconv/CP1252.so\n',
'7fce659f4000-7fce659f5000 r--p 00001000 08:01 100246
/usr/lib/gconv/CP1252.so\n', '7fce659f5000-7fce659f6000 rw-p 00002000 08:01 100246
                  /usr/lib/gconv/CP1252.so\n', '7fce659f6000-7fce65a77000 r-xp
00000000 08:01 1043745                    /lib/libm-2.10.2.so\n',
'7fce65a77000-7fce65c76000 ---p 00081000 08:01 1043745
/lib/libm-2.10.2.so\n', '7fce65c76000-7fce65c77000 r--p 00080000 08:01 1043745
             /lib/libm-2.10.2.so\n', '7fce65c77000-7fce65c78000 rw-p 00081000 08:01
1043745                    /lib/libm-2.10.2.so\n', '7fce65c78000-7fce65dc2000 r-xp
00000000 08:01 1043738                    /lib/libc-2.10.2.so\n',
'7fce65dc2000-7fce65fc2000 ---p 0014a000 08:01 1043738
/lib/libc-2.10.2.so\n', '7fce65fc2000-7fce65fc6000 r--p 0014a000 08:01 1043738
             /lib/libc-2.10.2.so\n', '7fce65fc6000-7fce65fc7000 rw-p 0014e000 08:01
1043738                    /lib/libc-2.10.2.so\n', '7fce65fc7000-7fce65fcc000 rw-p
00000000 00:00 0 \n', '7fce65fcc000-7fce65fe9000 r-xp 00000000 08:01 1043115
           /lib/ld-2.10.2.so\n', '7fce661ab000-7fce661ac000 rw-p 00000000 00:00 0
\n', '7fce661ac000-7fce661b3000 r--s 00000000 08:01 98890
/usr/lib/gconv/gconv-modules.cache\n', '7fce661b3000-7fce661d5000 rw-p 00000000 00:00
0 \n', '7fce661d5000-7fce661e8000 rw-p 00000000 00:00 0 \n',
'7fce661e8000-7fce661e9000 r--p 0001c000 08:01 1043115
/lib/ld-2.10.2.so\n', '7fce661e9000-7fce661ea000 rw-p 0001d000 08:01 1043115
           /lib/ld-2.10.2.so\n', '7fffe2c34000-7fffe2c49000 rw-p 00000000 00:00 0
                      [stack]\n', '7fffe2dd7000-7fffe2dd8000 r-xp 00000000 00:00 0
                       [vdso]\n', 'ffffffffff600000-ffffffffff601000 r-xp 00000000
00:00 0                  [vsyscall]\n']

So it must be a PostGIS problem.
I can make the shp available for testing if necessary. It has a wrong record.

All the best.
-- 
Paolo Cavallini: http://www.faunalia.it/pc