Query for empty string or Null using SQLAlchemy

Question

I'm pretty new to database design and web design, so please bear with me.

I have a Flask app where I need to search in a database depending on arguments passed in the url, using flask-sqlalschemy, such as:

.../index?part_number=123*321$description1=&description2=TestText

The asterisk will be interpreted as one or more char and ** as a single char. So I'll replace them with % and _, and terminate any special chars.

My question is: In the database some fields will be empty strings and some will be Null. But I need to have them both be interpreted as an empty string, such that they will be returned when the arg is empty or *.

I know I can do something along the lines of .like(part_number == None | part_number == '%'). But I only want to do this in the case where the search string is *, but how do I do that in a clever way when I have 10 different args, all and'ed together?

Here's a snippet and boiled down version of the relevant code. I'm not quite sure how to make a small program that can run in itself for testing.

filter_args = ['part_number', 'description1', 'description2']  # actual code has ~10 args
filters = dict()
for arg in filter_args:
  filter_str = request.args.get(arg, type=str, default='*')  # get filter arg from url

  filter_str = '*' if filter_str == '' else filter_str  # if the filter string is empty, search for all

  #  Replace * by % and ** by _ and terminate special chars
  filter_str = filter_str.replace('\\', '\\\\')
  filter_str = filter_str.replace('%', '\%')
  filter_str = filter_str.replace('_', '\_')
  filter_str = filter_str.replace('**', '_')
  filter_str = filter_str.replace('*', '%')
  filters[filter_name] = filter_str

parts = Part.query.filter(
  Part.part_number.ilike(filters['part_number']),  # and
  Part.description1.ilike(filters['description1']),  # and
  Part.description2.ilike(filters['description2'])
).order_by(Part.part_number)

Answer 1

I was searching for a way to filter for a null or empty string when I came across this question. While Ilja's answer is correct for the specifics of OP's question I thought I'd add that the coalesce function work for my situation too. If OP was looking for rows where part_number was None or empty:

Pat.query.filter(func.coalesce(Part.part_number, '') == '')