fix: suppress file conversion warnings in convert_office_doc (#703)

* test that output is suppressed

* add test for error output

* changelog and version

Author: MthwRobinson

CHANGELOG.md

@@ -1,4 +1,4 @@
-## 0.7.3-dev0
+## 0.7.3-dev1
 
 ### Enhancements
 
@@ -8,6 +8,7 @@
 
 ### Fixes
 
+* `convert_office_doc` no longers prints file conversion info messages to stdout.
 * `partition_via_api` reflects the actual filetype for the file processed in the API.
 
 ## 0.7.2

test_unstructured/partition/test_common.py

@@ -147,3 +147,19 @@ def test_normalize_layout_element_bulleted_list():
         ListItem(text="You're cool too.", coordinates=((1, 2), (1, 4), (3, 4), (3, 2))),
         ListItem(text="We're all cool!", coordinates=((1, 2), (1, 4), (3, 4), (3, 2))),
     ]
+
+
+class MockPopenWithError:
+    def __init__(self, *args, **kwargs):
+        pass
+
+    def communicate(self):
+        return b"", b"an error occurred"
+
+
+def test_convert_office_doc_captures_errors(monkeypatch, caplog):
+    import subprocess
+
+    monkeypatch.setattr(subprocess, "Popen", MockPopenWithError)
+    common.convert_office_doc("no-real.docx", "fake-directory", target_format="docx")
+    assert "an error occurred" in caplog.text

test_unstructured/partition/test_doc.py

@@ -56,7 +56,7 @@ def expected_elements():
     ]
 
 
-def test_partition_doc_with_filename(mock_document, expected_elements, tmpdir):
+def test_partition_doc_with_filename(mock_document, expected_elements, tmpdir, capsys):
     docx_filename = os.path.join(tmpdir.dirname, "mock_document.docx")
     doc_filename = os.path.join(tmpdir.dirname, "mock_document.doc")
     mock_document.save(docx_filename)
@@ -67,6 +67,9 @@ def test_partition_doc_with_filename(mock_document, expected_elements, tmpdir):
     assert elements[0].metadata.filename == "mock_document.doc"
     assert elements[0].metadata.file_directory == tmpdir.dirname
 
+    assert capsys.readouterr().out == ""
+    assert capsys.readouterr().err == ""
+
 
 def test_partition_doc_matches_partition_docx(mock_document, expected_elements, tmpdir):
     docx_filename = os.path.join(tmpdir.dirname, "mock_document.docx")
@@ -84,7 +87,7 @@ def test_partition_raises_with_missing_doc(mock_document, expected_elements, tmp
         partition_doc(filename=doc_filename)
 
 
-def test_partition_doc_with_file(mock_document, expected_elements, tmpdir):
+def test_partition_doc_with_file(mock_document, expected_elements, tmpdir, capsys):
     docx_filename = os.path.join(tmpdir.dirname, "mock_document.docx")
     doc_filename = os.path.join(tmpdir.dirname, "mock_document.doc")
     mock_document.save(docx_filename)
@@ -94,6 +97,9 @@ def test_partition_doc_with_file(mock_document, expected_elements, tmpdir):
         elements = partition_doc(file=f)
     assert elements == expected_elements
 
+    assert capsys.readouterr().out == ""
+    assert capsys.readouterr().err == ""
+
 
 def test_partition_doc_raises_with_both_specified(mock_document, tmpdir):
     docx_filename = os.path.join(tmpdir.dirname, "mock_document.docx")

unstructured/__version__.py

@@ -1 +1 @@
-__version__ = "0.7.3-dev0"  # pragma: no cover
+__version__ = "0.7.3-dev1"  # pragma: no cover

unstructured/partition/common.py

@@ -15,6 +15,7 @@
     PageBreak,
     Text,
 )
+from unstructured.logger import logger
 from unstructured.nlp.patterns import ENUMERATED_BULLETS_RE, UNICODE_BULLETS_RE
 
 
@@ -139,18 +140,22 @@ def convert_office_doc(input_filename: str, output_directory: str, target_format
     # users who do not have LibreOffice installed
     # ref: https://stackoverflow.com/questions/38468442/
     #       multiple-doc-to-docx-file-conversion-using-python
+    command = [
+        "soffice",
+        "--headless",
+        "--convert-to",
+        target_format,
+        "--outdir",
+        output_directory,
+        input_filename,
+    ]
     try:
-        subprocess.call(
-            [
-                "soffice",
-                "--headless",
-                "--convert-to",
-                target_format,
-                "--outdir",
-                output_directory,
-                input_filename,
-            ],
+        process = subprocess.Popen(
+            command,
+            stdout=subprocess.PIPE,
+            stderr=subprocess.PIPE,
         )
+        output, error = process.communicate()
     except FileNotFoundError:
         raise FileNotFoundError(
             """soffice command was not found. Please install libreoffice
@@ -161,6 +166,10 @@ def convert_office_doc(input_filename: str, output_directory: str, target_format
 - Debian: https://wiki.debian.org/LibreOffice""",
         )
 
+    logger.info(output.decode().strip())
+    if error:
+        logger.error(error.decode().strip())
+
 
 def exactly_one(**kwargs) -> None:
     """