From bd878e9bc06e9fa0cfb41a997d771465585e572b Mon Sep 17 00:00:00 2001
From: Claude Lab <claude-lab@localhost>
Date: Wed, 1 Apr 2026 17:44:46 +0200
Subject: [PATCH] fix: fall back to stdlib mimetypes when puremagic returns
 empty string

When puremagic cannot identify a file type it may return an empty string
instead of raising PureError. This empty string was passed through as the
MIME type, causing attachment validation to fail with:

    Error: This model does not support attachments of type ''

Fix: treat empty puremagic results the same as PureError. For
mimetype_from_path, additionally fall back to Python's stdlib
mimetypes.guess_type() which handles common extensions (.mp4, .jpg, etc.)
reliably.

Closes #1340

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
---
 llm/utils.py        | 15 +++++++++++----
 tests/test_utils.py | 24 ++++++++++++++++++++++++
 2 files changed, 35 insertions(+), 4 deletions(-)

diff --git a/llm/utils.py b/llm/utils.py
index 587f19284..02b17a7fa 100644
--- a/llm/utils.py
+++ b/llm/utils.py
@@ -3,6 +3,7 @@
 import httpx
 import itertools
 import json
+import mimetypes
 import pathlib
 import puremagic
 import re
@@ -37,17 +38,23 @@ def id(self):
 def mimetype_from_string(content) -> Optional[str]:
     try:
         type_ = puremagic.from_string(content, mime=True)
-        return MIME_TYPE_FIXES.get(type_, type_)
+        if type_:
+            return MIME_TYPE_FIXES.get(type_, type_)
     except puremagic.PureError:
-        return None
+        pass
+    return None
 
 
 def mimetype_from_path(path) -> Optional[str]:
     try:
         type_ = puremagic.from_file(path, mime=True)
-        return MIME_TYPE_FIXES.get(type_, type_)
+        if type_:
+            return MIME_TYPE_FIXES.get(type_, type_)
     except puremagic.PureError:
-        return None
+        pass
+    # Fall back to stdlib mimetypes when puremagic returns empty or raises
+    guessed, _ = mimetypes.guess_type(str(path))
+    return guessed
 
 
 def dicts_to_table_string(
diff --git a/tests/test_utils.py b/tests/test_utils.py
index 51fb8754f..bead8f848 100644
--- a/tests/test_utils.py
+++ b/tests/test_utils.py
@@ -4,6 +4,7 @@
     extract_fenced_code_block,
     instantiate_from_spec,
     maybe_fenced_code,
+    mimetype_from_path,
     schema_dsl,
     simplify_usage_dict,
     truncate_string,
@@ -516,3 +517,26 @@ class Tool6(Toolbox):
         pass
 
     assert Tool6()._config == {}
+
+
+def test_mimetype_from_path_falls_back_to_stdlib(tmp_path):
+    """When puremagic returns empty string, fall back to mimetypes.guess_type().
+
+    Regression test for https://github.com/simonw/llm/issues/1340
+    """
+    # Create a file with a known extension but content that puremagic may
+    # not recognise (e.g. an empty .mp4 file).
+    mp4_file = tmp_path / "video.mp4"
+    mp4_file.write_bytes(b"\x00" * 16)
+
+    result = mimetype_from_path(str(mp4_file))
+    assert result == "video/mp4"
+
+
+def test_mimetype_from_path_returns_none_for_unknown(tmp_path):
+    """Unknown extension and unrecognisable content returns None."""
+    unknown = tmp_path / "data.xyzzy123"
+    unknown.write_bytes(b"\x00" * 16)
+
+    result = mimetype_from_path(str(unknown))
+    assert result is None