BUG: append new df to old GDB file failed

[liuhao0826]

i generate source openFileGDB with a geopandas dataframe used：
gpd_linestring_df.to_file(filename2, driver="OpenFileGDB")
and dtyps：
uid object
agmach_id object
calc_date object
work_start_time object
work_end_time object
work_area float64
geometry geometry
agmach_type_name object
pro_ent_name object
mdl_name object
feed_rate float64
machine_width float64
area_name object
parent_area_name object

then i with the same type df，used：gpd_linestring_df.to_file(filename2, driver="OpenFileGDB",mode='a')
i am sure the diff df is the same dtyps，why failed with:
File "fiona\ogrext.pyx", line 1397, in fiona.ogrext.WritingSession.writerecs
ValueError: Record does not match collection schema: ['uid', 'agmach_id', 'calc_date', 'work_start_time', 'work_end_time', 'work_area', 'agmach_type_name', 'pro_ent_name', 'mdl_name', 'feed_rate', 'machine_width', 'area_name', 'parent_area_name'] != []

[liuhao0826]

generate old file path is ./linestring/北京市-202403010505.dgb
and it is generate sucess with

and append the same path failed；

[liuhao0826]

I'm not very familiar with the GDB file format, and I also want to ask why the directory I generated is a directory that contains sub files. shouldn't it have only one main file?

[brendan-ward]

ESRI File GDB is structured as a directory containing a lot of files. That is entirely expected here.

I'm not sure about why Fiona is failing here (probably an error to report there instead), as you are passing the same dataset each time, but this should be possible using engine='pyogrio' if you have that installed; it worked for me with the latest version of pyogrio / GDAL 3.8.3.

[liuhao0826]

ESRI File GDB is structured as a directory containing a lot of files. That is entirely expected here.

I'm not sure about why Fiona is failing here (probably an error to report there instead), as you are passing the same dataset each time, but this should be possible using engine='pyogrio' if you have that installed; it worked for me with the latest version of pyogrio / GDAL 3.8.3.

Although the geometry in my geopandas df is a linestring and polygon, when it is written to gdb and read out again, it shows that the geometry is a multilining and multipolygon, which causes the prompt format to not match and results in failure. Therefore, my current solution is to convert the linestring and polygon of the geometry in my df to multi (xx)
with the code：

_{gpd_convex_Hull_df = gpd.GeoDataFrame(pd_polygon_df, geometry="geometry", crs="EPSG:4326")
gpd_convex_Hull_df['geometry'] = gpd_convex_Hull_df['geometry'].apply(
lambda x: MultiPolygon([x]) if isinstance(x, Polygon) else x)
gpd_linestring_df = gpd.GeoDataFrame(pd_linestring_df, geometry="geometry", crs="EPSG:4326")
gpd_linestring_df['geometry'] = gpd_linestring_df['geometry'].apply(
lambda x: MultiLineString([x]) if isinstance(x, LineString) else x)}