fix: strip code fences, add MIME/size validation, add tests

scripts/convert-to-cooklang

@@ -16,9 +16,9 @@ Usage:
 
 import argparse
 import base64
-import json
 import mimetypes
 import os
+import re
 import sys
 
 import yaml
@@ -66,11 +66,30 @@ def build_file_content(filepath: str) -> list[dict]:
 
     Returns a list with one document/image content block.
     """
+    SUPPORTED_DOCUMENT_TYPES = {"application/pdf"}
+    SUPPORTED_IMAGE_TYPES = {"image/jpeg", "image/png", "image/gif", "image/webp"}
+
     mime_type, _ = mimetypes.guess_type(filepath)
     if mime_type is None:
-        # Default to octet-stream; let the API decide
         mime_type = "application/octet-stream"
 
+    if mime_type not in SUPPORTED_DOCUMENT_TYPES | SUPPORTED_IMAGE_TYPES:
+        print(
+            f"Error: unsupported file type '{mime_type}'. "
+            f"Supported types: PDF, JPEG, PNG, GIF, WebP",
+            file=sys.stderr,
+        )
+        sys.exit(1)
+
+    MAX_FILE_SIZE = 20 * 1024 * 1024  # 20MB
+    file_size = os.path.getsize(filepath)
+    if file_size > MAX_FILE_SIZE:
+        print(
+            f"Error: file too large ({file_size} bytes). Maximum file size is 20MB.",
+            file=sys.stderr,
+        )
+        sys.exit(1)
+
     with open(filepath, "rb") as f:
         data = base64.b64encode(f.read()).decode("utf-8")
 
@@ -114,6 +133,14 @@ def cooklang_from_anthropic(
         ],
     )
 
+    # Warn if response was truncated
+    if response.stop_reason == "max_tokens":
+        print(
+            "Warning: response may be truncated (reached max_tokens limit). "
+            "Consider splitting the recipe into smaller parts.",
+            file=sys.stderr,
+        )
+
     # Extract text from response content blocks
     parts: list[str] = []
     for block in response.content:
@@ -122,6 +149,12 @@ def cooklang_from_anthropic(
 
     cooklang = "\n".join(parts).strip()
 
+    # Strip markdown code fences if present (the prompts may instruct the
+    # model to wrap output in ```cooklang blocks)
+    cooklang = re.sub(r"^```\w*\n?", "", cooklang)
+    cooklang = re.sub(r"\n```\s*$", "", cooklang)
+    cooklang = cooklang.strip()
+
     # Basic validation: must start with front matter marker
     if not cooklang.startswith("---"):
         print(

tests/test_convert_to_cooklang.py

@@ -0,0 +1,91 @@
+"""Tests for scripts/convert-to-cooklang."""
+
+import importlib.machinery
+import importlib.util
+import os
+import tempfile
+
+# Load the script as a module (no .py extension, so use SourceFileLoader)
+SCRIPT_PATH = os.path.join(
+    os.path.dirname(__file__), "..", "scripts", "convert-to-cooklang"
+)
+_loader = importlib.machinery.SourceFileLoader("convert_to_cooklang", SCRIPT_PATH)
+_spec = importlib.util.spec_from_loader(
+    "convert_to_cooklang", _loader, origin=SCRIPT_PATH
+)
+mod = importlib.util.module_from_spec(_spec)
+_loader.exec_module(mod)
+
+
+def test_load_config_valid():
+    """Test loading a valid config file."""
+    with tempfile.NamedTemporaryFile(mode="w", suffix=".yaml", delete=False) as f:
+        f.write("anthropic:\n  api_key: sk-test-123\n")
+        config_path = f.name
+    try:
+        config = mod.load_config(config_path)
+        assert config["api_key"] == "sk-test-123"
+    finally:
+        os.unlink(config_path)
+
+
+def test_load_config_missing_api_key():
+    """Test that missing api_key exits with error."""
+    with tempfile.NamedTemporaryFile(mode="w", suffix=".yaml", delete=False) as f:
+        f.write("anthropic:\n  model: foo\n")
+        config_path = f.name
+    try:
+        try:
+            mod.load_config(config_path)
+            assert False, "Expected sys.exit"
+        except SystemExit as e:
+            assert e.code == 1
+    finally:
+        os.unlink(config_path)
+
+
+def test_build_text_content():
+    """Test building a text content block."""
+    result = mod.build_text_content("Hello")
+    assert result == [{"type": "text", "text": "Hello"}]
+
+
+def test_build_file_content_pdf():
+    """Test building file content for a PDF."""
+    with tempfile.NamedTemporaryFile(suffix=".pdf", delete=False) as f:
+        f.write(b"%PDF- test")
+        filepath = f.name
+    try:
+        result = mod.build_file_content(filepath)
+        assert result[0]["type"] == "document"
+        assert result[0]["source"]["media_type"] == "application/pdf"
+    finally:
+        os.unlink(filepath)
+
+
+def test_build_file_content_jpeg():
+    """Test building file content for a JPEG image."""
+    with tempfile.NamedTemporaryFile(suffix=".jpg", delete=False) as f:
+        f.write(b"\xff\xd8\xff\xe0 test jpeg")
+        filepath = f.name
+    try:
+        result = mod.build_file_content(filepath)
+        assert result[0]["type"] == "image"
+        assert result[0]["source"]["media_type"] == "image/jpeg"
+    finally:
+        os.unlink(filepath)
+
+
+def test_unsupported_file_type():
+    """Test that unsupported file types are rejected."""
+    with tempfile.NamedTemporaryFile(suffix=".tiff", delete=False) as f:
+        f.write(b"test")
+        filepath = f.name
+    try:
+        try:
+            mod.build_file_content(filepath)
+            assert False, "Expected sys.exit for unsupported type"
+        except SystemExit as e:
+            assert e.code == 1
+    finally:
+        os.unlink(filepath)