Add more granular exceptions / error pages, especially when handling HUC-8s and protected areas

[cstephen]

For many of the endpoints, we have made a first pass at handling exceptions by overgeneralizing. In the boundary.py endpoint, for example:

try:
    poly = akpa_gdf.loc[[akpa_id]].to_crs(4326)
except:
    return render_template("422/invalid_protected_area.html"), 422

Failing to support the provided protected area ID is just one of the things that can go wrong here.

We need to take an inventory of the try/except blocks that are overgeneralizing and deal with more specific things that can go wrong, and possible make more HTTP error templates with more precise language to handle each case.

[brucecrevensten]

Another comment to sharpen this. Here's an example from another code path:

# Before (code as it exists now):

# boundary.py ~l77
poly_type = validate_var_id(var_id)
if type(poly_type) is tuple:
     return poly_type

 try:
     poly = type_di[poly_type].loc[[var_id]].to_crs(4326)
 except:
     return render_template("422/invalid_area.html"), 422

# validate_request.py ~l63
def validate_var_id(var_id):
    if re.search("[^A-Za-z0-9]", var_id):
        return render_template("400/bad_request.html"), 400
    for filename in os.listdir("data/jsons/"):
        with open(os.path.join("data/jsons/", filename), "r") as f:
            for jsonline in json.loads(f.read()):
                if var_id == jsonline["id"]:
                    return jsonline["type"]
    return render_template("422/invalid_area.html"), 400

What could be done differently here would be to use exceptions so that the calling code can be written like this:

# boundary.py
try:
    poly_type = validate_var_id(var_id)
    poly = type_di[poly_type].loc[[var_id]].to_crs(4326)
except InvalidRequestError: ...
except UnknownAreaIdError: ...