parquet: add ogr support for parquet

[missinglink]

example parquet generation (using the 'shapefile' view):

wof sqlite export --sql="@exportable" whosonfirst-data-admin-nz-latest.db \
  | wof feature map view://shapefile.js \
  | wof feature ogr --format='Parquet' \
    whosonfirst-data-admin-nz-latest.parquet

DuckDB:

DESCRIBE SELECT * FROM read_parquet('whosonfirst-data-admin-nz-latest.parquet');

column_name|column_type|null|key|default|extra
id|INTEGER|YES|||
parent_id|INTEGER|YES|||
name|VARCHAR|YES|||
placetype|VARCHAR|YES|||
placelocal|VARCHAR|YES|||
country|VARCHAR|YES|||
repo|VARCHAR|YES|||
lat|DOUBLE|YES|||
lon|DOUBLE|YES|||
min_lat|DOUBLE|YES|||
min_lon|DOUBLE|YES|||
max_lat|DOUBLE|YES|||
max_lon|DOUBLE|YES|||
min_zoom|VARCHAR|YES|||
max_zoom|VARCHAR|YES|||
min_label|VARCHAR|YES|||
max_label|VARCHAR|YES|||
modified|DATE|YES|||
is_funky|VARCHAR|YES|||
population|VARCHAR|YES|||
country_id|INTEGER|YES|||
region_id|VARCHAR|YES|||
county_id|VARCHAR|YES|||
concord_ke|VARCHAR|YES|||
concord_id|VARCHAR|YES|||
iso_code|VARCHAR|YES|||
hasc_id|VARCHAR|YES|||
gn_id|VARCHAR|YES|||
wd_id|VARCHAR|YES|||
name_ara|VARCHAR|YES|||
name_ben|VARCHAR|YES|||
name_deu|VARCHAR|YES|||
name_eng|VARCHAR|YES|||
name_ell|VARCHAR|YES|||
name_fas|VARCHAR|YES|||
name_fra|VARCHAR|YES|||
name_heb|VARCHAR|YES|||
name_hin|VARCHAR|YES|||
name_hun|VARCHAR|YES|||
name_ind|VARCHAR|YES|||
name_ita|VARCHAR|YES|||
name_jpn|VARCHAR|YES|||
name_kor|VARCHAR|YES|||
name_nld|VARCHAR|YES|||
name_pol|VARCHAR|YES|||
name_por|VARCHAR|YES|||
name_rus|VARCHAR|YES|||
name_spa|VARCHAR|YES|||
name_swe|VARCHAR|YES|||
name_tur|VARCHAR|YES|||
name_ukr|VARCHAR|YES|||
name_urd|VARCHAR|YES|||
name_vie|VARCHAR|YES|||
name_zho|VARCHAR|YES|||
geom_src|VARCHAR|YES|||
geometry|GEOMETRY|YES|||
geometry_bbox|STRUCT(xmin FLOAT, ymin FLOAT, xmax FLOAT, ymax FLOAT)|YES|||