Turn any technical book PDF into a Claude Code skill — ready to study, reference, and use while you work.
Python agent-skillsai-agentsbook-to-skillcontext-engineeringdocument-processingedtechknowledge-baseknowledge-managementllmpdf-to-markdownragself-studystudy-tools
fix(extractor): skip an unreadable source instead of aborting the batch (#120)
Zhengzhongjie535828ed063f
diff --git a/book_to_skill/utils.py b/book_to_skill/utils.pyindex cfee623..4021101 100644--- a/book_to_skill/utils.py+++ b/book_to_skill/utils.py@@ -477,10 +477,21 @@ def extract_single_file(input_path: Path, extraction_mode: str, install_mode: stext = input_path.suffix.lower()document_format = ext.lstrip(".")- # Sniff magic bytes if suffix is not supported+ # Sniff magic bytes if suffix is not supported.+ #+ # Every failure in this function has to surface as ExtractionError: the+ # batch loop in main() catches only that, and anything else aborts the whole+ # run — including the sources that would have extracted fine. An unreadable+ # or unopenable file is a per-source problem, so translate it here. (The+ # ZipFile branch below already does this for OSError.)if ext not in SUPPORTED_EXTENSIONS:- with open(input_str, "rb") as f:- header = f.read(8)+ try:+ with open(input_str, "rb") as f:+ header = f.read(8)+ except OSError as exc:+ raise ExtractionError(+ f"Could not read {input_path.name}: {exc.strerror or exc}"+ ) from excif header[:4] == b"%PDF":ext = ".pdf"document_format = "pdf"diff --git a/tests/test_batch_resilience_unreadable.py b/tests/test_batch_resilience_unreadable.pynew file mode 100644index 0000000..dbe21d4--- /dev/null+++ b/tests/test_batch_resilience_unreadable.py@@ -0,0 +1,103 @@+"""Regression tests: an unreadable source must be skipped, not abort the batch.++`main()` catches only `ExtractionError`, so every failure inside+`extract_single_file` has to arrive as one. The magic-byte sniff — reached when+a file's suffix is not recognised — opened the file without translating+`OSError`, so a single unreadable file aborted the entire run with a traceback+and the remaining sources were never processed.++The pre-existing batch tests do not cover this: they use recognised suffixes,+which take the `read_text_file` path and never reach the sniff.+"""++import os+import stat+import sys+from pathlib import Path++import pytest++ROOT_DIR = Path(__file__).resolve().parent.parent+sys.path.insert(0, str(ROOT_DIR))++from book_to_skill.exceptions import ExtractionError # noqa: E402+from book_to_skill.utils import extract_single_file, main # noqa: E402+++def _make_unreadable(path: Path) -> Path:+ """A file that exists, has an unrecognised suffix, and cannot be opened."""+ path.write_bytes(b"junk")+ path.chmod(0o000)+ return path+++skip_if_readable_anyway = pytest.mark.skipif(+ os.geteuid() == 0 if hasattr(os, "geteuid") else True,+ reason="root (or a platform without POSIX permissions) can read mode-000 files",+)+++@skip_if_readable_anyway+def test_unreadable_unknown_suffix_raises_extraction_error(tmp_path):+ """The sniff translates OSError instead of letting it escape."""+ bad = _make_unreadable(tmp_path / "mystery.dat")+ try:+ with pytest.raises(ExtractionError) as excinfo:+ extract_single_file(bad, "text", "no")+ assert "mystery.dat" in str(excinfo.value)+ finally:+ bad.chmod(stat.S_IRUSR | stat.S_IWUSR)+++@skip_if_readable_anyway+def test_batch_survives_unreadable_source(tmp_path, monkeypatch, capsys):+ """The good source still extracts when an unreadable one comes first."""+ bad = _make_unreadable(tmp_path / "mystery.dat")+ good = tmp_path / "ok.md"+ good.write_text("Chapter 1\nReal content.\n", encoding="utf-8")++ workdir = tmp_path / "work"+ monkeypatch.setenv("BOOK_SKILL_WORKDIR", str(workdir))+ # config caches OUTPUT_* at import time; point the module constants at the+ # temp workdir so the run does not touch the shared default.+ import book_to_skill.config as config+ import book_to_skill.utils as utils++ for module in (config, utils):+ monkeypatch.setattr(module, "OUTPUT_DIR", workdir, raising=False)+ monkeypatch.setattr(module, "OUTPUT_TEXT", workdir / "full_text.txt", raising=False)+ monkeypatch.setattr(module, "OUTPUT_META", workdir / "metadata.json", raising=False)++ monkeypatch.setattr(+ sys, "argv", ["extract.py", str(bad), str(good), "--mode", "text", "--install-missing", "no"]+ )++ try:+ main()+ finally:+ bad.chmod(stat.S_IRUSR | stat.S_IWUSR)++ text = (workdir / "full_text.txt").read_text(encoding="utf-8")+ assert "Real content." in text++ out = capsys.readouterr()+ combined = out.out + out.err+ assert "mystery.dat" in combined+ assert "Skipping" in combined or "skipped" in combined+++def test_missing_file_still_reports_not_found(tmp_path):+ """The pre-existing not-found path is unchanged."""+ with pytest.raises(ExtractionError) as excinfo:+ extract_single_file(tmp_path / "nope.dat", "text", "no")+ assert "File not found" in str(excinfo.value)+++def test_readable_unknown_suffix_still_rejected_by_format(tmp_path):+ """A readable but unrecognised file still fails on format, not on IO."""+ odd = tmp_path / "mystery.dat"+ odd.write_bytes(b"not a pdf or a zip")++ with pytest.raises(ExtractionError) as excinfo:+ extract_single_file(odd, "text", "no")+ assert "Unsupported format" in str(excinfo.value)