from config import FunctionScoreConfig, IndexConfig, QueryConfig, RerankConfig, SPUConfig, SearchConfig
  from query.query_parser import QueryParser
  

  import pytest
  
  pytestmark = [pytest.mark.query, pytest.mark.regression]
  

  
  class _DummyTranslator:
      def translate(self, text, target_lang, source_lang, scene, model_name):
          return f"{text}-{target_lang}"
  
  

  def _tokenizer(text):
      return str(text).split()
  
  

  def _build_config() -> SearchConfig:
      return SearchConfig(
          es_index_name="test_products",
          field_boosts={"title.en": 3.0, "title.zh": 3.0},
          indexes=[IndexConfig(name="default", label="default", fields=["title.en", "title.zh"])],
          query_config=QueryConfig(
              enable_text_embedding=False,
              enable_query_rewrite=False,
              supported_languages=["en", "zh"],
              default_language="zh",
          ),
          function_score=FunctionScoreConfig(),
          rerank=RerankConfig(),
          spu_config=SPUConfig(enabled=False),
      )
  
  

  def test_parse_mixed_zh_query_translates_to_en(monkeypatch):

      parser = QueryParser(_build_config(), translator=_DummyTranslator(), tokenizer=_tokenizer)

      monkeypatch.setattr(parser.language_detector, "detect", lambda text: "zh")

      result = parser.parse(
          "法式 dress 连衣裙",
          tenant_id="162",
          generate_vector=False,
          target_languages=["zh", "en"],
      )

  
      assert result.detected_language == "zh"

      assert result.translations == {"en": "法式 dress 连衣裙-en"}
      assert result.query_tokens == ["法式", "dress", "连衣裙"]
      assert not hasattr(result, "query_text_by_lang")
      assert not hasattr(result, "search_langs")

  def test_parse_mixed_en_query_translates_to_zh(monkeypatch):

      parser = QueryParser(_build_config(), translator=_DummyTranslator(), tokenizer=_tokenizer)

      monkeypatch.setattr(parser.language_detector, "detect", lambda text: "en")

      result = parser.parse(
          "red 连衣裙",
          tenant_id="0",
          generate_vector=False,
          target_languages=["en", "zh"],
      )

  
      assert result.detected_language == "en"

      assert result.translations == {"zh": "red 连衣裙-zh"}
      assert result.query_tokens == ["red", "连衣裙"]

  
  
  def test_parse_waits_for_translation_when_source_in_index_languages(monkeypatch):
      """en 在 index_languages 内时仍应等待并采纳 en->zh 翻译结果（与向量共用预算）。"""

      parser = QueryParser(_build_config(), translator=_DummyTranslator(), tokenizer=_tokenizer)

      monkeypatch.setattr(parser.language_detector, "detect", lambda text: "en")

      result = parser.parse(
          "off shoulder top",
          tenant_id="0",
          generate_vector=False,
          target_languages=["en", "zh"],
      )

  
      assert result.detected_language == "en"

      assert result.translations.get("zh") == "off shoulder top-zh"

      assert not hasattr(result, "source_in_index_languages")

  
  
  def test_parse_reuses_tokenization_across_tail_stages(monkeypatch):
      tokenize_calls = []
  
      def counting_tokenizer(text):
          tokenize_calls.append(str(text))
          return str(text).split()
  
      config = SearchConfig(
          es_index_name="test_products",
          field_boosts={"title.en": 3.0, "title.zh": 3.0},
          indexes=[IndexConfig(name="default", label="default", fields=["title.en", "title.zh"])],
          query_config=QueryConfig(
              enable_text_embedding=False,
              enable_query_rewrite=False,
              supported_languages=["en", "zh"],
              default_language="en",
              style_intent_terms={
                  "color": [
                      {"en_terms": ["black"], "zh_terms": ["黑色"], "attribute_terms": ["black"]}
                  ],
              },
              style_intent_dimension_aliases={"color": ["color", "颜色"]},
              product_title_exclusion_rules=[
                  {
                      "zh_trigger_terms": ["修身"],
                      "en_trigger_terms": ["fitted"],
                      "zh_title_exclusions": ["宽松"],
                      "en_title_exclusions": ["loose"],
                  }
              ],
          ),
          function_score=FunctionScoreConfig(),
          rerank=RerankConfig(),
          spu_config=SPUConfig(enabled=False),
      )
      parser = QueryParser(
          config,
          translator=_DummyTranslator(),
          tokenizer=counting_tokenizer,
      )
      monkeypatch.setattr(parser.language_detector, "detect", lambda text: "en")
  
      result = parser.parse(
          "black fitted dress",
          tenant_id="0",
          generate_vector=False,
          target_languages=["en", "zh"],
      )
  
      assert result.translations == {"zh": "black fitted dress-zh"}
      assert result.style_intent_profile is not None
      assert result.style_intent_profile.is_active is True
      assert result.product_title_exclusion_profile is not None
      assert result.product_title_exclusion_profile.is_active is True
      assert tokenize_calls == []
  
  

  def test_parse_ascii_latin_query_uses_language_detector(monkeypatch):

      parser = QueryParser(_build_config(), translator=_DummyTranslator(), tokenizer=_tokenizer)

      detector_calls = []
  

      monkeypatch.setattr(
          parser.language_detector,
          "detect",

          lambda text: detector_calls.append(text) or "es",

      )
  
      result = parser.parse(

          "falda negra oficina",

          tenant_id="0",
          generate_vector=False,
          target_languages=["en", "zh"],
      )
  

      assert detector_calls == ["falda negra oficina"]
      assert result.detected_language == "es"
      assert result.translations == {"en": "falda negra oficina-en", "zh": "falda negra oficina-zh"}
      assert result.query_tokens == ["falda", "negra", "oficina"]