Fix issues: #274, #280 (#282)

* Fix issue with reading directories in `iterate_files()` (#280)
* Add directory checking logic in `iterate_files()` (#274)
* Added tests for #282, #274, #280

---------

Co-authored-by: Simon Willison <swillison@gmail.com>

llm/cli.py

@@ -1268,7 +1268,13 @@ def embed_multi(
 
         def iterate_files():
             for directory, pattern in files:
+                p = pathlib.Path(directory)
+                if not p.exists() or not p.is_dir():
+                    # fixes issue/274 - raise error if directory does not exist
+                    raise click.UsageError(f"Invalid directory: {directory}")
                 for path in pathlib.Path(directory).glob(pattern):
+                    if path.is_dir():
+                        continue  # fixed issue/280 - skip directories
                     relative = path.relative_to(directory)
                     content = None
                     if binary:

tests/test_embed_cli.py

@@ -432,7 +432,7 @@ def test_embed_multi_files(multi_files, scenario):
         ("nested/two.txt", b"two"),
         ("nested/more/three.txt", b"three"),
         # This tests the fallback to latin-1 encoding:
-        ("nested/more/ignored.ini", b"Has weird \x96 character"),
+        ("nested/more.txt/ignored.ini", b"Has weird \x96 character"),
     ):
         path = pathlib.Path(files / filename)
         path.parent.mkdir(parents=True, exist_ok=True)
@@ -483,6 +483,20 @@ def test_embed_multi_files(multi_files, scenario):
         ]
 
 
+@pytest.mark.parametrize(
+    "args,expected_error",
+    ((["not-a-dir", "*.txt"], "Invalid directory: not-a-dir"),),
+)
+def test_embed_multi_files_errors(multi_files, args, expected_error):
+    runner = CliRunner()
+    result = runner.invoke(
+        cli,
+        ["embed-multi", "files", "-m", "embed-demo", "--files"] + args,
+    )
+    assert result.exit_code == 2
+    assert expected_error in result.output
+
+
 @pytest.mark.parametrize(
     "extra_args,expected_error",
     (