any-file to PDF (Phase 1): converters/ dispatcher, images/text/md/csv/json/xml/svg/pdf-passthrough, 112 tests green

converters/pdf_passthrough.py

@@ -0,0 +1,66 @@
+"""PDF -> PDF passthrough (.pdf uploads).
+
+A valid, unencrypted PDF is validated with pypdf and passed through to the
+output with the standard naming/collision contract, re-stamping /Title from
+the uploaded filename stem. It reuses common.finalize_pdf so the atomic
+write path (.tmp + os.replace) and metadata handling are identical to every
+other converter.
+
+Rejections (all ValueError -> surfaced verbatim as a job error / CLI error):
+  - corrupt / unparseable   -> "Invalid PDF: ... could not be parsed"
+  - encrypted / password    -> "PDF rejected: ... encrypted or password-protected"
+  - more than MAX_PAGES     -> "PDF too large: ... pages (max N)"
+"""
+from __future__ import annotations
+
+import io
+from pathlib import Path
+
+from pypdf import PdfReader
+
+from . import common
+
+# req 1 (pdf passthrough): upper bound on page count for a pass-through PDF.
+MAX_PAGES = 500
+
+
+def convert(path: Path, job_id: str, progress_cb=None, output_dir=None,
+            source_name=None) -> str:
+    """Validate a PDF and pass it through to a same-named output PDF.
+
+    Satisfies: req 1 (pdf passthrough — validate, stamp /Title, standard
+    naming/collision, atomic write).
+    """
+    name = source_name or path.name
+    stem = Path(name).stem or "book"
+    raw = path.read_bytes()
+
+    # req 1: validate with pypdf. Construction raises on header/body corruption.
+    try:
+        reader = PdfReader(io.BytesIO(raw))
+    except Exception as exc:  # noqa: BLE001 - pypdf raises many corrupt-file types
+        raise ValueError(f"Invalid PDF: {name} could not be parsed ({exc})") from exc
+
+    # req 1: reject encrypted / password-protected BEFORE touching .pages
+    # (accessing pages on an encrypted reader raises, not a clean error).
+    if reader.is_encrypted:
+        raise ValueError(f"PDF rejected: {name} is encrypted or password-protected")
+
+    # req 1: count pages; a corrupt body beyond a valid header fails here.
+    try:
+        page_count = len(reader.pages)
+    except Exception as exc:  # noqa: BLE001
+        raise ValueError(f"Invalid PDF: {name} could not be parsed ({exc})") from exc
+
+    # req 1: reject oversized documents (clear error, distinct from corrupt).
+    if page_count > MAX_PAGES:
+        raise ValueError(
+            f"PDF too large: {name} has {page_count} pages (max {MAX_PAGES})"
+        )
+
+    # req 1: stamp /Title from the uploaded filename stem; standard naming/collision.
+    common.report(progress_cb, 30.0, f"Passing through {name} (1/1)")
+    out_dir = common.default_output_dir(output_dir)
+    filename = common.derive_export_name(name, out_dir)
+    return common.finalize_pdf(raw, out_dir, filename, job_id,
+                               progress_cb=progress_cb, title=stem)