fix: require HTTPS for remote file reads

docs/core/workflows.md

@@ -16,6 +16,8 @@ print(loaded["service"]["name"].upper_first())
 ```
 
 `DataFile` keeps source labels and metadata promoted and redacted before they enter workflow step names or result metadata.
+Remote reads accept HTTPS URLs only; clear-text and unsupported URL schemes are
+rejected before a request is created.
 
 ## DataWorkflow
 

packages/extended-data/src/extended_data/io/files.py

@@ -12,6 +12,7 @@
 from dataclasses import dataclass, field
 from pathlib import Path
 from typing import TYPE_CHECKING, Any, TypeAlias, cast
+from urllib.parse import urlsplit
 
 import validators
 
@@ -411,22 +412,21 @@ def is_url(path: str) -> bool:
     """Check if a string is a valid and safe URL.
 
     Uses the validators library for robust URL validation,
-    restricted to HTTP/HTTPS schemes only.
+    restricted to HTTPS URLs only so remote file reads are encrypted in transit.
 
     Args:
         path (str): The string to check.
 
     Returns:
-        bool: True if the string is a valid HTTP/HTTPS URL.
+        bool: True if the string is a valid HTTPS URL.
     """
     if not path:
         return False
     # validators.url returns True for valid URLs, ValidationError otherwise
     result = validators.url(path)
     if result is not True:
         return False
-    # Additional check: only allow http/https schemes
-    return path.startswith(("http://", "https://"))
+    return urlsplit(path).scheme == "https"
 
 
 def read_file(
@@ -455,11 +455,14 @@ def read_file(
 
     Raises:
         urllib.error.URLError: If the URL cannot be accessed.
-        ValueError: If the URL scheme is not allowed (only http/https permitted).
+        ValueError: If the URL scheme is not allowed (only HTTPS is permitted).
     """
     path_str = str(file_path)
 
-    # Handle URLs (is_url already validates HTTP/HTTPS only)
+    if "://" in path_str and not is_url(path_str):
+        raise ValueError("Remote file URLs must use HTTPS")
+
+    # Handle URLs (is_url already validates HTTPS only).
     if is_url(path_str):
         headers = headers or {}
         request = urllib.request.Request(path_str, headers=dict(headers))

packages/extended-data/tests/core/test_file_data_type.py

@@ -302,7 +302,7 @@ def test_file_path_rel_to_root(file_path: FilePath, expected_rel_to_root: str) -
 @pytest.mark.parametrize(
     ("path", "expected"),
     [
-        ("http://example.com/file.txt", True),
+        ("http://example.com/file.txt", False),
         ("https://example.com/file.txt", True),
         ("/path/to/file.txt", False),
         ("relative/path.txt", False),
@@ -323,6 +323,13 @@ def test_is_url(path: str, expected: bool) -> None:
     assert is_url(path) == expected
 
 
+@pytest.mark.parametrize("url", ["http://example.com/data.txt", "ftp://example.com/data.txt"])
+def test_read_file_rejects_unencrypted_or_unsupported_urls(url: str) -> None:
+    """Fail closed before a non-HTTPS URL can reach a network request."""
+    with pytest.raises(ValueError, match="must use HTTPS"):
+        read_file(url)
+
+
 def test_resolve_local_path_absolute() -> None:
     """Tests resolving an absolute path.