extract gdf index and use it during serialization of .to_geojson() and .to_gdf()

[mattijn]

This PR fix issue #164.

If a source geodataframe has a unique index, it will be used during extraction and used as index again when using the .to_gdf() function call. This also means that the commonly used identifier is being used as a member for all features in the serialization process creating geosjon (.to_geojson()).

Additionally, this PR also tries to use the .crs of the source geodataframe.

[Zeroto521]

Would it be better to give a warning to the user, that the index is not unique?

[Zeroto521]

It's a little weird for me to use like GeoDataFrame().from_features and geopandas.pd.json_normalize.
from_features is a class method, we can directly call it from GeoDataFrame.
geopandas.pd is a shortcut to hook pandas, but it's a little bit confusing.

from geopandas import GeoDataFrame
from pandas import json_normalize

return (
    GeoDataFrame.from_features(features=fc["features"], crs=crs)
    .set_axis(json_normalize(fc["features"])["id"].values)
)

[mattijn]

Thanks @Zeroto521 for the valuable feedback!
If you've time, you can send a PR?
Otherwise I will come back to this in a few days.

[Zeroto521]

Please check #166.

I only lint serialize_as_geodataframe.
But about the warning information, I'm still not sure whether add it or not.

        if geom.index.is_unique:
            self._data = geom.to_dict(orient="index")
        else:
            # warn the user the index is not unique

[mattijn]

Thanks for your concern. I think it is best to work with a unique index internally and store the index of the source separately. Than the source index can contains duplicates without creating potential issues for this package.