feat: add mime_type to ByteStream in LibreOfficeFileConverter (#3057)

* feat: add mime_type to ByteStream in LibreOfficeFileConverter

* fix: address copilot docstring and test comments

* check returned mime_type in integration tests

---------

Co-authored-by: Julian Risch <julian.risch@deepset.ai>

Author: StealthTensor

integrations/libreoffice/src/haystack_integrations/components/converters/libreoffice/converter.py

@@ -2,6 +2,7 @@
 #
 # SPDX-License-Identifier: Apache-2.0
 
+import mimetypes
 import os
 import shutil
 import subprocess
@@ -109,6 +110,26 @@ class LibreOfficeFileConverter:
     """A non-exhaustive mapping of supported conversion types by this component.
     See https://help.libreoffice.org/latest/en-GB/text/shared/guide/convertfilters.html for more information."""
 
+    MIME_TYPE_FALLBACKS: ClassVar[dict[str, str]] = {
+        "pdf": "application/pdf",
+        "doc": "application/msword",
+        "docx": "application/vnd.openxmlformats-officedocument.wordprocessingml.document",
+        "odt": "application/vnd.oasis.opendocument.text",
+        "rtf": "application/rtf",
+        "txt": "text/plain",
+        "html": "text/html",
+        "xlsx": "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet",
+        "xls": "application/vnd.ms-excel",
+        "ods": "application/vnd.oasis.opendocument.spreadsheet",
+        "csv": "text/csv",
+        "pptx": "application/vnd.openxmlformats-officedocument.presentationml.presentation",
+        "ppt": "application/vnd.ms-powerpoint",
+        "odp": "application/vnd.oasis.opendocument.presentation",
+        "epub": "application/epub+zip",
+        "png": "image/png",
+        "jpg": "image/jpeg",
+    }
+
     def __init__(
         self,
         output_file_type: OUTPUT_FILE_TYPE | None = None,
@@ -224,6 +245,12 @@ def _validate_args(self, output_file_type: str, input_file_type: str | None = No
             )
             raise ValueError(msg)
 
+    def _resolve_mime_type(self, output_path: Path, output_file_type: str) -> str | None:
+        mime_type, _ = mimetypes.guess_type(str(output_path))
+        if mime_type is None:
+            return self.MIME_TYPE_FALLBACKS.get(output_file_type)
+        return mime_type
+
     @component.output_types(output=list[ByteStream])
     def run(
         self,
@@ -269,14 +296,24 @@ def run(
                     output_path, args = self._get_conversion_args(tmp_path, tmpdir, resolved_output_file_type)
 
                     subprocess.run(args, check=True)  # noqa: S603 - ruff doesn't know the arguments have been validated
-                    outputs.append(ByteStream(data=output_path.read_bytes()))
+                    outputs.append(
+                        ByteStream(
+                            data=output_path.read_bytes(),
+                            mime_type=self._resolve_mime_type(output_path, resolved_output_file_type),
+                        )
+                    )
                     continue
 
                 self._validate_args(resolved_output_file_type, str(source).split(".")[-1])
                 output_path, args = self._get_conversion_args(source, tmpdir, resolved_output_file_type)
 
                 subprocess.run(args, check=True)  # noqa: S603
-                outputs.append(ByteStream(data=output_path.read_bytes()))
+                outputs.append(
+                    ByteStream(
+                        data=output_path.read_bytes(),
+                        mime_type=self._resolve_mime_type(output_path, resolved_output_file_type),
+                    )
+                )
 
         return {"output": outputs}
 
@@ -329,7 +366,12 @@ async def run_async(
                     process = await create_subprocess_exec(*args)
                     # Wait for process to complete as only one instance of soffice can occur at once
                     await process.wait()
-                    outputs.append(ByteStream(data=output_path.read_bytes()))
+                    outputs.append(
+                        ByteStream(
+                            data=output_path.read_bytes(),
+                            mime_type=self._resolve_mime_type(output_path, resolved_output_file_type),
+                        )
+                    )
                     continue
 
                 self._validate_args(resolved_output_file_type, str(source).split(".")[-1])
@@ -339,6 +381,11 @@ async def run_async(
                 # Wait for process to complete as only one instance of soffice can occur at once
                 await process.wait()
 
-                outputs.append(ByteStream(data=output_path.read_bytes()))
+                outputs.append(
+                    ByteStream(
+                        data=output_path.read_bytes(),
+                        mime_type=self._resolve_mime_type(output_path, resolved_output_file_type),
+                    )
+                )
 
         return {"output": outputs}

integrations/libreoffice/tests/test_converter.py

@@ -83,6 +83,7 @@ def test_run(self, converter: LibreOfficeFileConverter, test_files_path: Path) -
         for stream in output:
             assert isinstance(stream, ByteStream)
             assert len(stream.data) > 0
+            assert stream.mime_type == "application/pdf"
 
     @pytest.mark.integration
     def test_run_bytestream_source(self, converter: LibreOfficeFileConverter, test_files_path: Path) -> None:
@@ -95,6 +96,7 @@ def test_run_bytestream_source(self, converter: LibreOfficeFileConverter, test_f
         assert len(output) == 1
         assert isinstance(output[0], ByteStream)
         assert len(output[0].data) > 0
+        assert output[0].mime_type == "application/pdf"
 
     @pytest.mark.asyncio
     @pytest.mark.integration
@@ -112,6 +114,7 @@ async def test_run_async(self, converter: LibreOfficeFileConverter, test_files_p
         for stream in output:
             assert isinstance(stream, ByteStream)
             assert len(stream.data) > 0
+            assert stream.mime_type == "application/pdf"
 
     @pytest.mark.asyncio
     @pytest.mark.integration
@@ -127,3 +130,17 @@ async def test_run_async_bytestream_source(
         assert len(output) == 1
         assert isinstance(output[0], ByteStream)
         assert len(output[0].data) > 0
+        assert output[0].mime_type == "application/pdf"
+
+    def test_resolve_mime_type(self, mock_converter: LibreOfficeFileConverter) -> None:
+        with patch("mimetypes.guess_type", return_value=("application/pdf", None)):
+            mime = mock_converter._resolve_mime_type(Path("test.pdf"), "pdf")
+            assert mime == "application/pdf"
+
+        with patch("mimetypes.guess_type", return_value=(None, None)):
+            mime = mock_converter._resolve_mime_type(Path("test.docx"), "docx")
+            assert mime == "application/vnd.openxmlformats-officedocument.wordprocessingml.document"
+
+        with patch("mimetypes.guess_type", return_value=(None, None)):
+            mime = mock_converter._resolve_mime_type(Path("test.unknown"), "unknown")
+            assert mime is None