fix: minor fixes for unit tests

Author: pawel-kmiecik

src/unstructured_client/_hooks/custom/pdf_utils.py

@@ -2,7 +2,7 @@
 
 import io
 import logging
-from typing import cast, Optional, BinaryIO
+from typing import cast, Optional, BinaryIO, Union
 
 from pypdf import PdfReader
 from pypdf.errors import PdfReadError
@@ -17,7 +17,7 @@
 pdf_logger = logging.getLogger("pypdf")
 pdf_logger.setLevel(logging.ERROR)
 
-def read_pdf(pdf_file: BinaryIO) -> Optional[PdfReader]:
+def read_pdf(pdf_file: Union[BinaryIO, bytes]) -> Optional[PdfReader]:
     """Reads the given PDF file.
 
     Args:

src/unstructured_client/_hooks/custom/request_utils.py

@@ -32,6 +32,9 @@ def get_multipart_stream_fields(request: httpx.Request) -> dict[str, Any]:
 
     Returns:
         The multipart fields.
+
+    Raises:
+        Exception: If the filename is not set
     """
     content_type = request.headers.get("Content-Type", "")
     if "multipart" not in content_type:
@@ -50,6 +53,8 @@ def get_multipart_stream_fields(request: httpx.Request) -> dict[str, Any]:
                 mapped_fields[name].append(field.value)
             mapped_fields[field.name] = field.value
         elif isinstance(field, FileField):
+            if field.filename is None or not field.filename.strip():
+                raise ValueError("Filename can't be an empty string.")
             mapped_fields[field.name] = {
                 "filename": field.filename,
                 "content_type": field.headers.get("Content-Type", ""),
@@ -75,6 +80,7 @@ def create_pdf_chunk_request_params(
         PARTITION_FORM_SPLIT_PDF_ALLOW_FAILED_KEY,
         PARTITION_FORM_FILES_KEY,
         PARTITION_FORM_PAGE_RANGE_KEY,
+        PARTITION_FORM_PAGE_RANGE_KEY.replace("[]", ""),
         PARTITION_FORM_STARTING_PAGE_NUMBER_KEY,
     ]
     chunk_payload = {key: form_data[key] for key in form_data if key not in fields_to_drop}

src/unstructured_client/_hooks/custom/split_pdf_hook.py

@@ -312,7 +312,7 @@ def before_request(
 
         page_range_start, page_range_end = form_utils.get_page_range(
             form_data,
-            key=PARTITION_FORM_PAGE_RANGE_KEY,
+            key=PARTITION_FORM_PAGE_RANGE_KEY.replace("[]", ""),
             max_pages=pdf.get_num_pages(),
         )