mdify-cli 2.6.0__tar.gz → 2.7.0__tar.gz

{mdify_cli-2.6.0/mdify_cli.egg-info → mdify_cli-2.7.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: mdify-cli
-Version: 2.6.0
+Version: 2.7.0
 Summary: Convert PDFs and document images into structured Markdown for LLM workflows
 Author: tiroq
 License-Expression: MIT

{mdify_cli-2.6.0 → mdify_cli-2.7.0}/mdify/__init__.py

@@ -1,3 +1,3 @@
 """mdify - Convert documents to Markdown via Docling container."""
 
-__version__ = "2.6.0"
+__version__ = "2.7.0"

{mdify_cli-2.6.0 → mdify_cli-2.7.0}/mdify/docling_client.py

@@ -4,6 +4,8 @@ from dataclasses import dataclass
 from pathlib import Path
 from typing import Optional
 
+import mimetypes
+
 import requests
 
 
@@ -40,6 +42,12 @@ class DoclingHTTPError(DoclingClientError):
         super().__init__(f"HTTP {status_code}: {message}")
 
 
+def _get_mime_type(file_path: Path) -> str:
+    """Get MIME type for file, with fallback for unknown types."""
+    mime_type, _ = mimetypes.guess_type(str(file_path))
+    return mime_type or "application/octet-stream"
+
+
 def check_health(base_url: str) -> bool:
     """Check if docling-serve is healthy.
 
@@ -77,7 +85,7 @@ def convert_file(
         with open(file_path, "rb") as f:
             response = requests.post(
                 f"{base_url}/v1/convert/file",
-                files={"files": (file_path.name, f, "application/pdf")},
+                files={"files": (file_path.name, f, _get_mime_type(file_path))},
                 data={"to_formats": to_format, "do_ocr": str(do_ocr).lower()},
             )
 
@@ -126,7 +134,7 @@ def convert_file_async(
         with open(file_path, "rb") as f:
             response = requests.post(
                 f"{base_url}/v1/convert/file/async",
-                files={"files": (file_path.name, f, "application/pdf")},
+                files={"files": (file_path.name, f, _get_mime_type(file_path))},
                 data={"to_formats": to_format, "do_ocr": str(do_ocr).lower()},
             )
 

{mdify_cli-2.6.0 → mdify_cli-2.7.0/mdify_cli.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: mdify-cli
-Version: 2.6.0
+Version: 2.7.0
 Summary: Convert PDFs and document images into structured Markdown for LLM workflows
 Author: tiroq
 License-Expression: MIT

{mdify_cli-2.6.0 → mdify_cli-2.7.0}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "mdify-cli"
-version = "2.6.0"
+version = "2.7.0"
 description = "Convert PDFs and document images into structured Markdown for LLM workflows"
 readme = "README.md"
 requires-python = ">=3.8"

{mdify_cli-2.6.0 → mdify_cli-2.7.0}/tests/test_docling_client.py

@@ -356,3 +356,97 @@ class TestDoclingHTTPError:
         error = DoclingHTTPError(400, "Bad Request")
 
         assert isinstance(error, Exception)
+
+
+class TestMimeTypeDetection:
+    """Test MIME type detection in file conversion."""
+
+    def test_convert_file_sends_correct_mime_for_xlsx(self, tmp_path):
+        """Test that .xlsx files are sent with correct MIME type."""
+        test_file = tmp_path / "test.xlsx"
+        test_file.write_bytes(b"fake xlsx content")
+
+        with patch("mdify.docling_client.requests.post") as mock_post:
+            mock_response = Mock()
+            mock_response.status_code = 200
+            mock_response.json.return_value = [
+                {"content": "# Test Spreadsheet\n\nContent here."}
+            ]
+            mock_post.return_value = mock_response
+
+            convert_file("http://localhost:5001", test_file)
+
+            mock_post.assert_called_once()
+            call_args = mock_post.call_args
+            files_param = call_args[1]["files"]
+            filename, file_obj, mime_type = files_param["files"]
+            assert (
+                mime_type
+                == "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet"
+            )
+
+    def test_convert_file_sends_correct_mime_for_pdf(self, tmp_path):
+        """Test that .pdf files are sent with correct MIME type (regression test)."""
+        test_file = tmp_path / "test.pdf"
+        test_file.write_bytes(b"fake pdf content")
+
+        with patch("mdify.docling_client.requests.post") as mock_post:
+            mock_response = Mock()
+            mock_response.status_code = 200
+            mock_response.json.return_value = [
+                {"content": "# Test Document\n\nContent here."}
+            ]
+            mock_post.return_value = mock_response
+
+            convert_file("http://localhost:5001", test_file)
+
+            mock_post.assert_called_once()
+            call_args = mock_post.call_args
+            files_param = call_args[1]["files"]
+            filename, file_obj, mime_type = files_param["files"]
+            assert mime_type == "application/pdf"
+
+    def test_convert_file_sends_correct_mime_for_docx(self, tmp_path):
+        """Test that .docx files are sent with correct MIME type."""
+        test_file = tmp_path / "test.docx"
+        test_file.write_bytes(b"fake docx content")
+
+        with patch("mdify.docling_client.requests.post") as mock_post:
+            mock_response = Mock()
+            mock_response.status_code = 200
+            mock_response.json.return_value = [
+                {"content": "# Test Document\n\nContent here."}
+            ]
+            mock_post.return_value = mock_response
+
+            convert_file("http://localhost:5001", test_file)
+
+            mock_post.assert_called_once()
+            call_args = mock_post.call_args
+            files_param = call_args[1]["files"]
+            filename, file_obj, mime_type = files_param["files"]
+            assert (
+                mime_type
+                == "application/vnd.openxmlformats-officedocument.wordprocessingml.document"
+            )
+
+    def test_convert_file_fallback_for_unknown_extension(self, tmp_path):
+        """Test that unknown file extensions fall back to application/octet-stream."""
+        test_file = tmp_path / "test.unknownext123"
+        test_file.write_bytes(b"fake unknown content")
+
+        with patch("mdify.docling_client.requests.post") as mock_post:
+            mock_response = Mock()
+            mock_response.status_code = 200
+            mock_response.json.return_value = [
+                {"content": "# Test Content\n\nContent here."}
+            ]
+            mock_post.return_value = mock_response
+
+            convert_file("http://localhost:5001", test_file)
+
+            mock_post.assert_called_once()
+            call_args = mock_post.call_args
+            files_param = call_args[1]["files"]
+            filename, file_obj, mime_type = files_param["files"]
+            assert mime_type == "application/octet-stream"