fix: add support for base64 URLs to the image filter (#60)

Author: masci

src/banks/filters/image.py

@@ -1,15 +1,25 @@
 # SPDX-FileCopyrightText: 2023-present Massimiliano Pippi <mpippi@gmail.com>
 #
 # SPDX-License-Identifier: MIT
+import re
 from pathlib import Path
 from urllib.parse import urlparse
 
 from banks.types import ContentBlock, ImageUrl
 
+BASE64_PATH_REGEX = re.compile(r"image\/.*;base64,.*")
+
 
 def _is_url(string: str) -> bool:
     result = urlparse(string)
-    return all([result.scheme, result.netloc])
+    if not result.scheme:
+        return False
+
+    if not result.netloc:
+        # The only valid format when netloc is empty is base64 data urls
+        return all([result.scheme == "data", BASE64_PATH_REGEX.match(result.path)])
+
+    return True
 
 
 def image(value: str) -> str:

tests/test_image.py

@@ -57,6 +57,23 @@ def test_image_with_file_path(tmp_path):
     assert content_block["image_url"]["url"].startswith("data:image/jpeg;base64,")
 
 
+def test_image_base64(tmp_path):
+    """Test image filter with a binary input"""
+    test_image = "data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAABMgA"
+    result = image(test_image)
+
+    # Verify the content block wrapper
+    assert result.startswith("<content_block>")
+    assert result.endswith("</content_block>")
+
+    # Parse the JSON content
+    json_content = result[15:-16]  # Remove wrapper tags
+    content_block = json.loads(json_content)
+
+    assert content_block["type"] == "image_url"
+    assert content_block["image_url"]["url"].startswith("data:image/png;base64,")
+
+
 def test_image_with_nonexistent_file():
     """Test image filter with a nonexistent file path"""
     with pytest.raises(FileNotFoundError):