feat(extractor): add Google News headlines extractor with Foxcape headless and URL resolution

- Add standalone CLI script scripts/extract_google_news.py for Google News RSS scraping - Integrate foxcape in headless mode as primary stealth anti-bot engine - Implement parallel article URL resolution using googlenewsdecoder and ThreadPoolExecutor - Support language and regional locale mapping (-l, --lang, --locale) - Implement real-time progress logging in stderr and --silent flag - Add unit, integration, and live E2E tests in tests/test_extract_google_news.py - Add full SpecKit documentation (specs/002-google-news-extractor/) - Create comprehensive README.md covering both NLP Classifier and Google News Extractor
2026-08-20 11:50:16 -03:00
parent 67cc40f91a
commit 6e3d57619b
59 changed files with 16118 additions and 2160 deletions
@@ -0,0 +1,355 @@
+"""
+Testes unitários e de integração para o Extrator de Manchetes do Google News.
+
+Cobre validação de entrada, mapeamento de idiomas/locales, parsing e
+higienização de XML/HTML, orquestração, decodificação de URLs e contrato de execução CLI.
+"""
+
+from __future__ import annotations
+
+import json
+from pathlib import Path
+from unittest.mock import patch
+
+import pytest
+
+from scripts.extract_google_news import (
+    ExtractionResult,
+    NewsArticle,
+    SearchQuery,
+    extract_google_news,
+    get_hl_gl_ceid,
+    main,
+    parse_google_news_rss,
+    resolve_article_url,
+    resolve_articles_urls,
+)
+
+FIXTURE_PATH = Path(__file__).parent / "fixtures" / "google_news_sample.xml"
+
+
+@pytest.fixture
+def sample_rss_xml() -> str:
+    """Fixture que fornece o conteúdo do XML de exemplo para testes offline."""
+    return FIXTURE_PATH.read_text(encoding="utf-8")
+
+
+def test_get_hl_gl_ceid_default_mappings():
+    """Valida o mapeamento padrão de idiomas para pares (hl, gl, ceid)."""
+    hl, gl, ceid = get_hl_gl_ceid("pt")
+    assert hl == "pt-BR"
+    assert gl == "BR"
+    assert ceid == "BR:pt-BR"
+
+    hl, gl, ceid = get_hl_gl_ceid("en")
+    assert hl == "en-US"
+    assert gl == "US"
+    assert ceid == "US:en-US"
+
+    hl, gl, ceid = get_hl_gl_ceid("es")
+    assert hl == "es-419"
+    assert gl == "AR"
+    assert ceid == "AR:es-419"
+
+    hl, gl, ceid = get_hl_gl_ceid("de")
+    assert hl == "de"
+    assert gl == "DE"
+    assert ceid == "DE:de"
+
+
+def test_get_hl_gl_ceid_with_custom_locale():
+    """Valida a sobrescrita geográfica quando o argumento locale é especificado."""
+    hl, gl, ceid = get_hl_gl_ceid("es", locale="MX")
+    assert hl == "es-419"
+    assert gl == "MX"
+    assert ceid == "MX:es-419"
+
+    hl, gl, ceid = get_hl_gl_ceid("en", locale="GB")
+    assert hl == "en-GB"
+    assert gl == "GB"
+    assert ceid == "GB:en-GB"
+
+    hl, gl, ceid = get_hl_gl_ceid("es", locale="ES")
+    assert hl == "es"
+    assert gl == "ES"
+    assert ceid == "ES:es"
+
+
+def test_get_hl_gl_ceid_dynamic_fallback():
+    """Valida fallback dinâmico para idiomas regionais não listados explicitamente."""
+    hl, gl, ceid = get_hl_gl_ceid("ja_jp")
+    assert hl == "ja-JP"
+    assert gl == "JP"
+    assert ceid == "JP:ja-JP"
+
+
+def test_search_query_validation():
+    """Valida as regras de negócio e limites de SearchQuery."""
+    # Instanciação válida
+    q = SearchQuery(keyword="inteligencia artificial", language="pt", max_pages=1)
+    assert q.clean_keyword == "inteligencia artificial"
+    assert q.clean_language == "pt"
+    assert q.clean_locale is None
+    assert q.max_pages == 1
+
+    # Palavra-chave vazia ou apenas espaços deve lançar ValueError
+    with pytest.raises(ValueError, match="palavra-chave"):
+        SearchQuery(keyword="   ", language="pt")
+
+    # Idioma com menos de 2 caracteres deve lançar ValueError
+    with pytest.raises(ValueError, match="idioma"):
+        SearchQuery(keyword="test", language="p")
+
+    # Intervalo de páginas fora de 1..10 deve lançar ValueError
+    with pytest.raises(ValueError, match="páginas"):
+        SearchQuery(keyword="test", language="pt", max_pages=0)
+
+    with pytest.raises(ValueError, match="páginas"):
+        SearchQuery(keyword="test", language="pt", max_pages=11)
+
+
+def test_parse_google_news_rss_with_fixture(sample_rss_xml: str):
+    """Valida o parsing do feed RSS, higienização de tags HTML e deduplicação."""
+    articles = parse_google_news_rss(sample_rss_xml, max_pages=1)
+
+    # 4 itens no fixture, mas 1 não possui link -> exatamente 3 válidos
+    assert len(articles) == 3
+
+    # Artigo 1: InfoMoney com HTML no description que deve ser limpo
+    art1 = articles[0]
+    assert "InfoMoney" in art1.titulo
+    assert art1.url.startswith("https://news.google.com/rss/articles/")
+    assert art1.quando_publicado == "Thu, 20 Aug 2026 10:30:00 GMT"
+    assert art1.pagina == 1
+    assert "<" not in (art1.subtitulo or "")
+    assert ">" not in (art1.subtitulo or "")
+    assert "crescimento expressivo" in (art1.subtitulo or "")
+
+    # Artigo 2: G1 com parágrafos limpos
+    art2 = articles[1]
+    assert "G1" in art2.titulo
+    assert "<p>" not in (art2.subtitulo or "")
+
+    # Artigo 3: Folha com descrição redundante/igual ao título -> subtitulo deve ser None
+    art3 = articles[2]
+    assert art3.subtitulo is None
+
+
+def test_resolve_article_url_fallback():
+    """Valida fallback gracioso de URL quando não é link do Google News ou em erro."""
+    direct_url = "https://www.globo.com/noticia/123"
+    assert resolve_article_url(direct_url) == direct_url
+
+    with patch(
+        "scripts.extract_google_news.gnewsdecoder", return_value={"status": False}
+    ):
+        gn_url = "https://news.google.com/rss/articles/fake_token"
+        assert resolve_article_url(gn_url) == gn_url
+
+
+def test_resolve_article_url_success():
+    """Valida resolução bem-sucedida de URL do Google News para o portal destino."""
+    gn_url = "https://news.google.com/rss/articles/valid_token"
+    dest_url = "https://infomoney.com.br/mercados/artigo-ia"
+
+    with patch(
+        "scripts.extract_google_news.gnewsdecoder",
+        return_value={"status": True, "decoded_url": dest_url},
+    ):
+        resolved = resolve_article_url(gn_url)
+        assert resolved == dest_url
+
+
+def test_resolve_articles_urls_batch():
+    """Valida a resolução concorrente em lote de uma lista de NewsArticle."""
+    articles = [
+        NewsArticle(
+            titulo="Notícia 1",
+            url="https://news.google.com/rss/articles/1",
+            pagina=1,
+        ),
+        NewsArticle(
+            titulo="Notícia 2",
+            url="https://news.google.com/rss/articles/2",
+            pagina=1,
+        ),
+    ]
+
+    with patch(
+        "scripts.extract_google_news.resolve_article_url",
+        side_effect=lambda u: f"https://destinofinal.com/{u.split('/')[-1]}",
+    ):
+        resolved = resolve_articles_urls(articles)
+        assert len(resolved) == 2
+        assert resolved[0].url == "https://destinofinal.com/1"
+        assert resolved[1].url == "https://destinofinal.com/2"
+
+
+def test_extract_google_news_orchestration_mocked(sample_rss_xml: str):
+    """Valida a consolidação do ExtractionResult a partir da busca mockada com URLs resolvidas."""
+    query = SearchQuery(
+        keyword="inteligência artificial", language="pt", locale="BR", max_pages=1
+    )
+
+    with (
+        patch(
+            "scripts.extract_google_news._fetch_rss_content",
+            return_value=sample_rss_xml,
+        ),
+        patch(
+            "scripts.extract_google_news.resolve_article_url",
+            side_effect=lambda u: f"https://resolved.com/{u[-5:]}",
+        ),
+    ):
+        result = extract_google_news(query, resolve_urls=True)
+
+        assert isinstance(result, ExtractionResult)
+        assert result.query == "inteligência artificial"
+        assert result.language == "pt"
+        assert result.locale == "BR"
+        assert result.total_paginas == 1
+        assert result.total_itens == 3
+        assert len(result.items) == 3
+        assert result.scraped_at is not None
+        assert result.items[0].url.startswith("https://resolved.com/")
+
+
+def test_cli_execution_stdout(sample_rss_xml: str, capsys: pytest.CaptureFixture[str]):
+    """Valida execução padrão do CLI com saída JSON no stdout."""
+    with (
+        patch(
+            "scripts.extract_google_news._fetch_rss_content",
+            return_value=sample_rss_xml,
+        ),
+        patch(
+            "scripts.extract_google_news.resolve_article_url", side_effect=lambda u: u
+        ),
+    ):
+        exit_code = main(
+            ["--query", "inteligencia artificial", "--lang", "pt", "--pretty"]
+        )
+        assert exit_code == 0
+
+        captured = capsys.readouterr()
+        data = json.loads(captured.out)
+        assert data["query"] == "inteligencia artificial"
+        assert data["total_itens"] == 3
+        assert len(data["items"]) == 3
+        # Validar indentação presente por causa de --pretty
+        assert "\n  " in captured.out
+
+
+def test_cli_execution_file_output(sample_rss_xml: str, tmp_path: Path):
+    """Valida gravação em arquivo com criação automática de diretórios pais."""
+    out_file = tmp_path / "sub_dir" / "news_out.json"
+
+    with (
+        patch(
+            "scripts.extract_google_news._fetch_rss_content",
+            return_value=sample_rss_xml,
+        ),
+        patch(
+            "scripts.extract_google_news.resolve_article_url", side_effect=lambda u: u
+        ),
+    ):
+        exit_code = main(["-q", "IA", "-p", "1", "-o", str(out_file)])
+        assert exit_code == 0
+
+        assert out_file.exists()
+        data = json.loads(out_file.read_text(encoding="utf-8"))
+        assert data["total_itens"] == 3
+
+
+def test_cli_empty_query_error(capsys: pytest.CaptureFixture[str]):
+    """Valida tratamento de erro e código de saída 1 para parâmetro vazio."""
+    exit_code = main(["--query", "   "])
+    assert exit_code == 1
+
+    captured = capsys.readouterr()
+    assert "Erro de validação" in captured.err
+
+
+def test_cli_network_error_handling(capsys: pytest.CaptureFixture[str]):
+    """Valida tratamento de erro e código de saída 2 para falhas de rede."""
+    with patch(
+        "scripts.extract_google_news._fetch_rss_content",
+        side_effect=RuntimeError("Connection refused"),
+    ):
+        exit_code = main(["--query", "IA"])
+        assert exit_code == 2
+
+        captured = capsys.readouterr()
+        assert "Erro na extração" in captured.err
+        assert "Connection refused" in captured.err
+
+
+# ---------------------------------------------------------------------------
+# Testes E2E (End-to-End) com resolução real de rede e validação de URLs finais
+# ---------------------------------------------------------------------------
+
+
+def test_e2e_resolve_real_google_news_url():
+    """Valida E2E que o decodificador resolve uma URL real do Google News para o veículo de imprensa."""
+    # URL real de artigo extraída do Google News RSS
+    sample_gn_url = (
+        "https://news.google.com/rss/articles/"
+        "CBMi6wFBVV95cUxQUS0tMGxacUJycDlCWXpWSGR3T0hfR1E0Q0txWjdqek9LWjNZTUo1WEx3"
+        "UEw3Skt1eW1Wa2VYUUE2ajNYMmpEckhsdGFXUGtmQktYX2JWa1hmclhEZEVIa3hNODhpVXNO"
+        "UGN4cDhmWmJMczBEUVFDNS1aX3EzRHh6VmQ3cVY0ZnZmaW9YWDZJSE9UNFJ2dXpyNFlHaVVs"
+        "VlY0V1FiZ0tzZ3FpRVhUYnhPbmdnOFRveW5oOVB3WDAzS3c0eWFjMDBZSERwNmRkRk1MRHZF"
+        "UE92UE9GMmpRcFZ5cUU2Ym1NeDdQU2tn?oc=5"
+    )
+
+    resolved_url = resolve_article_url(sample_gn_url)
+
+    # Não deve mais ser URL do Google News
+    assert "news.google.com" not in resolved_url
+    # Deve ser uma URL absoluta http/https apontando para o portal real (TyC Sports)
+    assert resolved_url.startswith("http")
+    assert "tycsports.com" in resolved_url
+
+
+def test_e2e_extract_google_news_live_pipeline():
+    """Valida E2E o fluxo completo de busca, parsing e resolução de URLs reais ao vivo."""
+    query = SearchQuery(keyword="tecnologia", language="pt", locale="BR", max_pages=1)
+    result = extract_google_news(query, resolve_urls=True)
+
+    assert isinstance(result, ExtractionResult)
+    assert result.total_itens > 0
+    assert len(result.items) == result.total_itens
+
+    for article in result.items:
+        assert article.titulo
+        assert article.url.startswith("http")
+        # Garante que as URLs foram decodificadas e não permanecem no formato intermediário
+        assert "news.google.com/rss/articles/" not in article.url
+
+
+def test_e2e_cli_live_file_output(tmp_path: Path):
+    """Valida E2E a execução do CLI com saída real em arquivo e URLs decodificadas."""
+    out_file = tmp_path / "e2e_result.json"
+    exit_code = main(
+        [
+            "--query",
+            "economia",
+            "--lang",
+            "pt",
+            "--max-pages",
+            "1",
+            "--output",
+            str(out_file),
+        ]
+    )
+
+    assert exit_code == 0
+    assert out_file.exists()
+
+    data = json.loads(out_file.read_text(encoding="utf-8"))
+    assert data["query"] == "economia"
+    assert data["total_itens"] > 0
+    assert len(data["items"]) > 0
+
+    first_item = data["items"][0]
+    assert first_item["titulo"]
+    assert first_item["url"].startswith("http")
+    assert "news.google.com/rss/articles/" not in first_item["url"]