Skip to content

Python tool for converting files and office documents to Markdown.

fix: support extended content-disposition filenames (#2045)

Rāna(Bass Ver.)76b965ad520d
diff --git a/packages/markitdown/src/markitdown/_markitdown.py b/packages/markitdown/src/markitdown/_markitdown.py
index ca435d5f..50901133 100644
--- a/packages/markitdown/src/markitdown/_markitdown.py
+++ b/packages/markitdown/src/markitdown/_markitdown.py
@@ -6,6 +6,8 @@ import shutil
import traceback
import io
from dataclasses import dataclass
+from email.message import Message
+from email.utils import collapse_rfc2231_value
from importlib.metadata import entry_points
from typing import Any, List, Dict, Optional, Union, BinaryIO
from pathlib import Path
@@ -51,6 +53,23 @@ from ._exceptions import (
)
+def _get_content_disposition_filename(content_disposition: str) -> Optional[str]:
+ message = Message()
+ message["content-disposition"] = content_disposition
+
+ fallback_filename: Optional[str] = None
+ extended_filename: Optional[str] = None
+ for key, value in message.get_params(header="content-disposition", unquote=True):
+ if key != "filename":
+ continue
+ if isinstance(value, tuple):
+ extended_filename = collapse_rfc2231_value(value)
+ elif fallback_filename is None:
+ fallback_filename = value
+
+ return extended_filename or fallback_filename
+
+
# Lower priority values are tried first.
PRIORITY_SPECIFIC_FILE_FORMAT = (
0.0 # e.g., .docx, .pdf, .xlsx, Or specific pages, e.g., wikipedia
@@ -512,9 +531,10 @@ class MarkItDown:
filename: Optional[str] = None
extension: Optional[str] = None
if "content-disposition" in response.headers:
- m = re.search(r"filename=([^;]+)", response.headers["content-disposition"])
- if m:
- filename = m.group(1).strip("\"'")
+ filename = _get_content_disposition_filename(
+ response.headers["content-disposition"]
+ )
+ if filename is not None:
_, _extension = os.path.splitext(filename)
if len(_extension) > 0:
extension = _extension
diff --git a/packages/markitdown/tests/test_module_misc.py b/packages/markitdown/tests/test_module_misc.py
index 6b46f612..dd04e142 100644
--- a/packages/markitdown/tests/test_module_misc.py
+++ b/packages/markitdown/tests/test_module_misc.py
@@ -8,6 +8,7 @@ import pytest
from unittest.mock import MagicMock
from markitdown._uri_utils import parse_data_uri, file_uri_to_path
+from markitdown._markitdown import _get_content_disposition_filename
from markitdown import (
MarkItDown,
@@ -289,6 +290,55 @@ def test_input_as_strings() -> None:
assert "# Test" in result.text_content
+def _mock_response(content_disposition: str) -> MagicMock:
+ response = MagicMock()
+ response.headers = {"content-disposition": content_disposition}
+ response.url = "https://example.com/download"
+ response.iter_content.return_value = [b"name,value\nalpha,beta\n"]
+ response.raise_for_status.return_value = None
+ return response
+
+
+def test_convert_response_uses_rfc5987_content_disposition_filename() -> None:
+ markitdown = MarkItDown()
+ result = markitdown.convert_response(
+ _mock_response("attachment; filename*=UTF-8''data.csv")
+ )
+
+ assert result.markdown == "\n".join(
+ [
+ "| name | value |",
+ "| --- | --- |",
+ "| alpha | beta |",
+ ]
+ )
+
+
+def test_convert_response_prefers_extended_content_disposition_filename() -> None:
+ markitdown = MarkItDown()
+ result = markitdown.convert_response(
+ _mock_response("attachment; filename=fallback.txt; filename*=UTF-8''data.csv")
+ )
+
+ assert result.markdown == "\n".join(
+ [
+ "| name | value |",
+ "| --- | --- |",
+ "| alpha | beta |",
+ ]
+ )
+
+
+def test_get_content_disposition_filename_decodes_rfc5987() -> None:
+ assert (
+ _get_content_disposition_filename(
+ "attachment; filename=fallback.txt; "
+ "filename*=UTF-8''d%C3%A1t%C3%A1%2Ecsv"
+ )
+ == "d\u00e1t\u00e1.csv"
+ )
+
+
def test_pptx_chart_multi_series_conversion() -> None:
"""Charts with multiple series and many categories must convert correctly.