Better exception handling when bad json (#40)

pyproject.toml

@@ -69,7 +69,7 @@ warn_unreachable = true
 disallow_untyped_decorators = false
 
 [tool.ruff]
-line-length = 88
+line-length = 99
 fix = true
 
 [tool.ruff.lint]

src/clean_notebook/clean.py

@@ -1,6 +1,7 @@
 from __future__ import annotations
 
 import json
+import sys
 import uuid
 from pathlib import Path
 from typing import Any, AnyStr, Iterator
@@ -36,7 +37,11 @@ def clean_single_notebook(
         raw = f.read()
 
     newline = _find_line_ending(raw)
-    nb = json.loads(raw)
+    try:
+        nb = json.loads(raw)
+    except json.JSONDecodeError as e:
+        print(f"{file}: Failed to json decode ({e})")
+        sys.exit(1)
     set_id = _check_set_id(nb)
 
     cleaned, sort_keys = False, False
@@ -114,7 +119,7 @@ def _find_line_ending(s: AnyStr) -> AnyStr:
         endings = [b"\n", b"\r", b"\r\n"]
     else:
         msg = "Not str or bytes"
-        raise ValueError(msg)
+        raise TypeError(msg)
 
     counter = {s.count(e): e for e in endings}
     return counter[max(counter)]

tests/data/dirty_bad_json.ipynb

@@ -0,0 +1,71 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "sys.version_info(major=3, minor=9, micro=13, releaselevel='final', serial=0)"
+      ]
+     },
+     "execution_count": 1,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "import sys\n",
+    "\n",
+    "sys.version_info"
+==== THIS IS A BAD EDIT ===
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "print(\"saved from vscode\")"
+   ]
+  },
+  {
+   "attachments": {},
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# Markdown"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3.9.13 ('clean_notebook')",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.9.13"
+  },
+  "orig_nbformat": 4,
+  "vscode": {
+   "interpreter": {
+    "hash": "b6343796ef4448f7e41a24425e4d6603c7973238ae27f15472679aaf07fca50e"
+   }
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}

tests/test_clean_notebook.py

@@ -21,7 +21,7 @@ def load_file(path: Path) -> bytes:
     return file_bytes.replace(le, b"\n")
 
 
-@pytest.fixture(scope="session")
+@pytest.fixture()
 def temp_path(tmp_path_factory: TempPathFactory) -> Iterator[Path]:
     src = Path("tests/data").resolve(strict=True)
     dst = tmp_path_factory.mktemp("data")
@@ -121,3 +121,12 @@ def test_empty_notebook(capsys: CaptureFixture[str], temp_path: Path) -> None:
 
     captured = capsys.readouterr()
     assert captured.out.strip() == f"Notebook '{dirty}' does not have any valid cells."
+
+
+def test_bad_json(capsys: CaptureFixture[str], temp_path: Path) -> None:
+    file = temp_path / "dirty_bad_json.ipynb"
+    with pytest.raises(SystemExit):
+        clean_single_notebook(file)
+    captured = capsys.readouterr()
+    msg = f"{file}: Failed to json decode (Expecting ',' delimiter: line 23 column 1 (char 412))"
+    assert captured.out.strip() == msg