feat: complete translator stack

backend/tests/test_health.py

@@ -0,0 +1,9 @@
+from fastapi.testclient import TestClient
+from backend.app.main import app
+
+client = TestClient(app)
+
+def test_health_check():
+    resp = client.get("/health")
+    assert resp.status_code == 200
+    assert resp.json().get("status") == "ok"

backend/tests/test_processors.py

@@ -0,0 +1,69 @@
+import json
+from pathlib import Path
+from typing import Callable
+
+import fitz
+from ebooklib import epub
+from docx import Document
+
+from backend.app.processors.registry import select_processor
+from backend.app.services.chunking import chunk_document
+
+
+def _create_docx(path: Path) -> Path:
+    doc = Document()
+    doc.add_heading("Titolo principale", level=1)
+    doc.add_paragraph("Primo paragrafo del documento.")
+    doc.add_paragraph("Secondo paragrafo con alcune frasi.")
+    doc.save(path)
+    return path
+
+
+def _create_pdf(path: Path) -> Path:
+    doc = fitz.open()
+    page = doc.new_page()
+    page.insert_text((72, 72), "Titolo PDF\n\nParagrafo PDF di prova.")
+    doc.save(path)
+    doc.close()
+    return path
+
+
+def _create_epub(path: Path) -> Path:
+    book = epub.EpubBook()
+    book.set_identifier("test-123")
+    book.set_title("EPUB di prova")
+    book.set_language("it")
+
+    chapter = epub.EpubHtml(title="Capitolo", file_name="chap.xhtml")
+    chapter.content = "<h1>Capitolo</h1><p>Paragrafo EPUB.</p>"
+    book.add_item(chapter)
+    book.toc = (epub.Link("chap.xhtml", "Capitolo", "chap"),)
+    book.add_item(epub.EpubNcx())
+    book.add_item(epub.EpubNav())
+    epub.write_epub(str(path), book)
+    return path
+
+
+def _create_txt(path: Path) -> Path:
+    path.write_text("Titolo TXT\n\nParagrafo 1\n\nParagrafo 2", encoding="utf-8")
+    return path
+
+
+generator_map: dict[str, Callable[[Path], Path]] = {
+    ".docx": _create_docx,
+    ".pdf": _create_pdf,
+    ".epub": _create_epub,
+    ".txt": _create_txt,
+}
+
+
+def test_processor_chain(tmp_path: Path):
+    for suffix, generator in generator_map.items():
+        target = tmp_path / f"sample{suffix}"
+        path = generator(target)
+        processor = select_processor(path)
+        assert processor is not None, f"Processor missing for {suffix}"
+        extraction = processor(path)
+        chunks = chunk_document(f"session-{suffix}", extraction)
+        assert chunks, f"Nessun chunk per {suffix}"
+        assert all(chunk.text.strip() for chunk in chunks)