Spacy EntityRuler not adding new patterns when built via add_pipe()

Question

I am having issues with the spaCy Entity Ruler component. I am trying to add new patterns to it in this way:

def add_rule_patterns(input_references, model_ruler):
    ruler = model_ruler
    ref_patterns = []
    print("Adding new patterns")
    for ref in input_references.iloc[:,0]:
        ref_patterns.append({"label":"SYS", "pattern": {"TEXT": {"FUZZY": ref}}})
        pattern = [{"label":"SYS", "pattern": {"TEXT": {"FUZZY": ref}}}]

        # print(f"Adding pattern:  {ref}. Stored as: {pattern}")

    ruler.add_patterns(ref_patterns)

    for pattern in ruler.patterns:
            print(f"Added pattern for:  {ref}. Stored as: {pattern}")
    return  ref_patterns    

Then when preparing the nlp model:

nlp = spacy.blank('fr')
nlp.add_pipe("ner")
ner = nlp.get_pipe("ner")

config = {
   "phrase_matcher_attr": None,
   "validate": True,
   "overwrite_ents": True,
   "ent_id_sep": "||",
}

nlp.add_pipe("entity_ruler", before="ner", config=config)
ruler = nlp.get_pipe("entity_ruler")

Finally, to add the patterns:

patterns = add_rule_patterns(df_sys, ruler)
ruler.add_patterns(patterns)
print(ruler.patterns)

The issue is that "ruler" as instantiated does not contain any "add_patterns()" method. If i add the ruler like this:

from spacy.pipeline import EntityRuler
ruler = EntityRuler(nlp, overwrite_ents=True)

The method is present, but the patterns are not added to the pipe. I've looked into some related questions, like this or this one, but I still can't seem to add the new patterns. Am I missing something or is it some known issue with spaCy?