ogr2ogr not handling null column name values in CSV

[andrewxhill]

Data exported from R (and others I'm sure), will often contain a null column name over an ID column. Something like,

"", "Name","Species", "Geom"
1, "Boxelder", "Acer negundo", "some geom"
2, "Juneberry", "Amelanchier arborea", "some geom"

These files are imported to PostgreSQL (via -f "PostgreSQL") incorrectly. The data are shifted over one column, and one column is dropped. Alternatively, in the quoted example above, even a space will make the file work correctly,

" ", "Name","Species", "Geom"
1, "Boxelder", "Acer negundo", "some geom"
2, "Juneberry", "Amelanchier arborea", "some geom"

[Even Rouault]

Fixed in trunk (r24381) and branches/1.9 (r24382)