import json

from nltk.corpus import wordnet as wn

from tests import EVENTS_FOLDER
from my_ghost_writer.jsonpath_extractor import JSONPathStructureAnalyzer
from my_ghost_writer.text_parsers2 import get_wordnet_synonyms


def analyze_detailed_report_lists(cls, detailed_report: dict, expected_detailed_report: dict):
    for (row_k, row_v), (expected_row_k, expected_row_v) in zip(detailed_report.items(), expected_detailed_report.items()):
        cls.assertEqual(row_k, expected_row_k)
        del row_v["samples"]
        del expected_row_v["samples"]
        del row_v["sample_count"]
        del expected_row_v["sample_count"]
        cls.assertDictEqual(row_v, expected_row_v)


def assert__json_structure__get_wordnet_synonyms(cls, word):
    with open(EVENTS_FOLDER / f"get_wordnet_synonyms_{word}_ok1.json", "r") as src:
        expected_detailed_report = json.load(src)
    related_words = get_wordnet_synonyms(word)
    first_related_words = related_words[0]
    analyzer = JSONPathStructureAnalyzer()
    analyzer.extract_all_paths(first_related_words)
    detailed_report = analyzer.get_detailed_type_report()
    analyze_detailed_report_lists(cls, detailed_report, expected_detailed_report)
    # with open(EVENTS_FOLDER / f"get_wordnet_synonyms_{word}_ok1.json", "w") as src:
    #     json.dump(detailed_report, src)


def get_relationships(synset):
    relationships = {
        'synonyms': len(synset.lemma_names()),
        'antonyms': sum(len(lemma.antonyms()) for lemma in synset.lemmas()),
        'hypernyms': len(synset.hypernyms()),
        'hyponyms': len(synset.hyponyms()),
        'holonyms': len(synset.member_holonyms()) + len(synset.part_holonyms()) + len(synset.substance_holonyms()),
        'meronyms': len(synset.member_meronyms()) + len(synset.part_meronyms()) + len(synset.substance_meronyms()),
        'similar_tos': len(synset.similar_tos()),
        'also_sees': len(synset.also_sees()),
        'causes': len(synset.causes())
    }
    return relationships


def extract_word_relationships():
    results = []
    holonym_results = []
    cause_results = []

    for pos in ['n', 'v']:
        for synset in wn.all_synsets(pos):
            rels = get_relationships(synset)
            total = sum(1 for v in rels.values() if v > 0)
            results.append((synset, total, rels))
            if rels['holonyms'] > 0:
                holonym_results.append((synset, rels['holonyms'], rels))
            if rels['causes'] > 0:
                cause_results.append((synset, rels['causes'], rels))

    # Sort and get top 5
    results.sort(key=lambda x: x[1], reverse=True)
    holonym_results.sort(key=lambda x: x[1], reverse=True)
    cause_results.sort(key=lambda x: x[1], reverse=True)

    print("Top 5 synsets with most relationships:")
    for synset, total, rels in results[:5]:
        print(f"{synset.name()} ({synset.definition()}): {rels}")

    print("\nTop 5 synsets with holonym relationships:")
    for synset, count, rels in holonym_results[:5]:
        print(f"{synset.name()} ({synset.definition()}): {rels}")

    print("\nTop 5 synsets with cause relationships:")
    for synset, count, rels in cause_results[:5]:
        print(f"{synset.name()} ({synset.definition()}): {rels}")


if __name__ == "__main__":
    extract_word_relationships()