#1981

Author: saleweaver

sp_api/api/reports/reports.py

@@ -378,7 +378,9 @@ def get_report_document(
                     with client.stream("GET", res.payload.get("url")) as document_response:
                         if not character_code:
                             character_code = resolve_character_code(
-                                document_response.encoding, fallback="iso-8859-1"
+                                document_response.encoding,
+                                fallback="iso-8859-1",
+                                response_charset_encoding=getattr(document_response, "charset_encoding", None),
                             )
                         stream_to_file_sync(
                             document_response,
@@ -390,7 +392,9 @@ def get_report_document(
                     document_response = client.get(res.payload.get("url"))
                     if not character_code:
                         character_code = resolve_character_code(
-                            document_response.encoding, fallback="iso-8859-1"
+                            document_response.encoding,
+                            fallback="iso-8859-1",
+                            response_charset_encoding=getattr(document_response, "charset_encoding", None),
                         )
                     document = decompress_bytes(
                         document_response.content, compression_algorithm

sp_api/asyncio/api/reports/reports.py

@@ -373,7 +373,9 @@ async def get_report_document(
                 ) as document_response:
                     if not character_code:
                         character_code = resolve_character_code(
-                            document_response.encoding, fallback="iso-8859-1"
+                            document_response.encoding,
+                            fallback="iso-8859-1",
+                            response_charset_encoding=getattr(document_response, "charset_encoding", None),
                         )
                     await stream_to_file_async(
                         document_response,
@@ -389,7 +391,9 @@ async def get_report_document(
                 )
                 if not character_code:
                     character_code = resolve_character_code(
-                        document_response.encoding, fallback="iso-8859-1"
+                        document_response.encoding,
+                        fallback="iso-8859-1",
+                        response_charset_encoding=getattr(document_response, "charset_encoding", None),
                     )
                 document = decompress_bytes(
                     document_response.content, compression_algorithm

sp_api/util/report_document.py

@@ -3,8 +3,24 @@
 from io import BytesIO, StringIO
 
 
-def resolve_character_code(response_encoding, fallback="iso-8859-1"):
-    character_code = response_encoding or fallback
+def resolve_character_code(
+    response_encoding=None,
+    fallback="iso-8859-1",
+    response_charset_encoding=None,
+):
+    """Resolve report text encoding while preserving pre-httpx behavior.
+
+    requests exposed ``None`` when a response had no charset, and the reports
+    client then fell back to iso-8859-1. httpx exposes ``utf-8`` by default in
+    the same scenario, so we treat implicit utf-8 as "missing" and keep the
+    historical fallback.
+    """
+    if response_charset_encoding:
+        character_code = response_charset_encoding
+    elif response_encoding and response_encoding.lower() != "utf-8":
+        character_code = response_encoding
+    else:
+        character_code = fallback
     if character_code and character_code.lower() == "windows-31j":
         character_code = "cp932"
     return character_code

tests/api/test_report_document_util.py

@@ -0,0 +1,56 @@
+from sp_api.util.report_document import resolve_character_code
+
+
+def test_resolve_character_code_uses_charset_encoding_when_available():
+    assert (
+        resolve_character_code(
+            response_encoding="utf-8",
+            response_charset_encoding="iso-8859-1",
+            fallback="iso-8859-1",
+        )
+        == "iso-8859-1"
+    )
+
+
+def test_resolve_character_code_falls_back_from_default_utf8_without_charset():
+    assert (
+        resolve_character_code(
+            response_encoding="utf-8",
+            response_charset_encoding=None,
+            fallback="iso-8859-1",
+        )
+        == "iso-8859-1"
+    )
+
+
+def test_resolve_character_code_falls_back_when_encoding_is_missing():
+    assert (
+        resolve_character_code(
+            response_encoding=None,
+            response_charset_encoding=None,
+            fallback="iso-8859-1",
+        )
+        == "iso-8859-1"
+    )
+
+
+def test_resolve_character_code_keeps_non_utf8_encoding():
+    assert (
+        resolve_character_code(
+            response_encoding="windows-31j",
+            response_charset_encoding=None,
+            fallback="iso-8859-1",
+        )
+        == "cp932"
+    )
+
+
+def test_resolve_character_code_keeps_explicit_utf8_charset():
+    assert (
+        resolve_character_code(
+            response_encoding="utf-8",
+            response_charset_encoding="utf-8",
+            fallback="iso-8859-1",
+        )
+        == "utf-8"
+    )