fix: allow partition_email to process emails with no content (#797)

* version and changelog

* ingest-test-fixtures-update

CHANGELOG.md

@@ -1,4 +1,4 @@
-## 0.7.8-dev0
+## 0.7.8-dev1
 
 ### Enhancements
 
@@ -8,6 +8,7 @@
 
 ### Fixes
 
+* `partition_email` now works if there is no message content
 * Updates the `"fast"` strategy for `partition_pdf` so that it's able to recursively
 * Adds recursive functionality to all fsspec connectors
 * Adds generic --recursive ingest flag

test_unstructured/partition/test_email.py

@@ -322,7 +322,7 @@ def test_convert_to_iso_8601(time, expected):
     assert iso_time == expected
 
 
-def test_partition_email_raises_with_no_html_content():
+def test_partition_email_still_works_with_no_content():
     filename = os.path.join(EXAMPLE_DOCS_DIRECTORY, "email-no-html-content-1.eml")
-    with pytest.raises(ValueError):
-        partition_email(filename=filename)
+    elements = partition_email(filename=filename)
+    assert elements == []

unstructured/__version__.py

@@ -1 +1 @@
-__version__ = "0.7.8-dev0"  # pragma: no cover
+__version__ = "0.7.8-dev1"  # pragma: no cover

unstructured/partition/email.py

@@ -283,9 +283,9 @@ def partition_email(
 
     content = content_map.get(content_source, "")
     if not content:
-        raise ValueError(f"{content_source} content not found in email")
+        elements = []
 
-    if content_source == "text/html":
+    elif content_source == "text/html":
         # NOTE(robinson) - In the .eml files, the HTML content gets stored in a format that
         # looks like the following, resulting in extraneous "=" characters in the output if
         # you don't clean it up
@@ -316,6 +316,7 @@ def partition_email(
                             break
                         except (UnicodeDecodeError, UnicodeError):
                             continue
+
     elif content_source == "text/plain":
         list_content = split_by_paragraph(content)
         elements = partition_text(text=content, encoding=encoding)