💥 ♻️ drop requests in favor of httpx (#414)

Author: sebastianMindee

.pre-commit-config.yaml

@@ -35,8 +35,7 @@ repos:
         args: []
         exclude: "tests/|examples/|docs/"
         additional_dependencies:
-          - toml
-          - types-requests
+          - httpx
           - types-setuptools
           - importlib-metadata
           - types-Pillow

mindee/image/extracted_image.py

@@ -64,11 +64,11 @@ def save_to_file(self, output_path: Path | str, file_format: str | None = None):
             else:
                 image.save(resolved_path)
             logger.info("File saved successfully to '%s'.", resolved_path)
-        except TypeError as exc:
-            raise MindeeError("Invalid path/filename provided.") from exc
-        except Exception as exc:
-            print(exc)
-            raise MindeeError(f"Could not save file {Path(output_path).name}.") from exc
+        except TypeError as e:
+            raise MindeeError("Invalid path/filename provided.") from e
+        except Exception as e:
+            print(e)
+            raise MindeeError(f"Could not save file {Path(output_path).name}.") from e
 
     def as_input_source(self) -> FileInput:
         """

mindee/input/local_input_source.py

@@ -48,9 +48,9 @@ def __init__(self) -> None:
             try:
                 pdf = pdfium.PdfDocument(self.file_object)
                 self.page_count = len(pdf)
-            except pdfium.PdfiumError as exc:
+            except pdfium.PdfiumError as e:
                 logger.warning(
-                    "Could not open PDF file: %s due to %s", self.filename, exc
+                    "Could not open PDF file: %s due to %s", self.filename, e
                 )
                 self.page_count = 0
             self.file_object.seek(0)
@@ -106,11 +106,11 @@ def fix_pdf(self, maximum_offset: int = 500) -> None:
                     f"PDF couldn't be fixed. PDF tag was found at position {pos}."
                 )
             self.file_mimetype = "application/pdf"
-        except MimeTypeError as exc:
-            raise exc
-        except Exception as exc:
-            print(f"Attempt to fix pdf raised exception {exc}.")
-            raise exc
+        except MimeTypeError as e:
+            raise e
+        except Exception as e:
+            logger.error("Attempt to fix pdf raised exception %s.", e)
+            raise e
 
     def is_pdf(self) -> bool:
         """:return: True if the file is a PDF."""

mindee/input/local_response.py

@@ -53,8 +53,8 @@ def as_dict(self) -> dict[str, Any]:
         try:
             self._file.seek(0)
             out_json = json.loads(self._file.read())
-        except json.decoder.JSONDecodeError as exc:
-            raise MindeeError("File is not a valid dictionary.") from exc
+        except json.decoder.JSONDecodeError as e:
+            raise MindeeError("File is not a valid dictionary.") from e
         return out_json
 
     @staticmethod
@@ -87,8 +87,8 @@ def get_hmac_signature(self, secret_key: str | bytes | bytearray):
                 self._file.read(),
                 algorithm,
             )
-        except (TypeError, ValueError) as exc:
-            raise MindeeError("Could not get HMAC signature from payload.") from exc
+        except (TypeError, ValueError) as e:
+            raise MindeeError("Could not get HMAC signature from payload.") from e
 
         return mac.hexdigest()
 
@@ -114,5 +114,5 @@ def deserialize_response(self, response_class: type[ResponseT]) -> ResponseT:
         """
         try:
             return response_class(self.as_dict)
-        except KeyError as exc:
-            raise MindeeError("Invalid class specified for deserialization.") from exc
+        except KeyError as e:
+            raise MindeeError("Invalid class specified for deserialization.") from e

mindee/input/url_input_source.py

@@ -5,7 +5,7 @@
 from pathlib import Path
 from urllib.parse import urlparse
 
-import requests
+import httpx
 
 from mindee.error.mindee_error import MindeeSourceError
 from mindee.input.bytes_input import BytesInput
@@ -153,7 +153,7 @@ def __get_file_extension(filename) -> str | None:
         :return: The lowercase file extension or None if not found.
         """
         ext = os.path.splitext(filename)[1]
-        return ext.lower() if ext else None
+        return str(ext).lower() if ext else None
 
     def __fill_filename(self, filename=None) -> str:
         """
@@ -167,7 +167,7 @@ def __fill_filename(self, filename=None) -> str:
 
         if not filename or not os.path.splitext(filename)[1]:
             filename = self.__generate_file_name(
-                extension=URLInputSource.__get_file_extension(filename)
+                extension=URLInputSource.__get_file_extension(filename) or ""
             )
 
         return filename
@@ -185,15 +185,15 @@ def __make_request(url, auth, headers, redirects, max_redirects) -> bytes:
         :return: The content of the response.
         :raises MindeeSourceError: If max redirects are exceeded or the request fails.
         """
-        result = requests.get(url, headers=headers, timeout=120, auth=auth)
+        result = httpx.get(url, headers=headers, timeout=120, auth=auth)
         if 299 < result.status_code < 400:
             if redirects == max_redirects:
                 raise MindeeSourceError(
                     f"Can't reach URL after {redirects} out of {max_redirects} redirects, "
                     f"aborting operation."
                 )
             return URLInputSource.__make_request(
-                redirects.location, auth, headers, redirects + 1, max_redirects
+                result.headers["Location"], auth, headers, redirects + 1, max_redirects
             )
 
         if result.status_code >= 400 or result.status_code < 200:

mindee/mindee_http/response_validation.py

@@ -1,31 +1,34 @@
 import json
 
-import requests
+import httpx
 
 from mindee.parsing.common import StringDict
 
 
-def is_valid_sync_response(response: requests.Response) -> bool:
+def is_valid_sync_response(response: httpx.Response) -> bool:
     """
     Checks if the synchronous response is valid. Returns True if the response is valid.
 
     :param response: a requests response object.
     :return: bool
     """
-    if not response or not response.ok:
+    if not response or response.is_error:
+        return False
+    try:
+        response_json = response.json()
+    except httpx.DecodingError:
         return False
-    response_json = json.loads(response.content)
     # EXTREMELY rare edge case where raw html is sent instead of json.
     return isinstance(response_json, dict)
 
 
-def is_valid_async_response(response: requests.Response) -> bool:
+def is_valid_async_response(response: httpx.Response) -> bool:
     """
     Checks if the asynchronous response is valid. Also checks if it is a valid synchronous response.
 
     Returns True if the response is valid.
 
-    :param response: a requests response object.
+    :param response: an httpx response object.
     :return: bool
     """
     if not is_valid_sync_response(response):
@@ -43,15 +46,15 @@ def is_valid_async_response(response: requests.Response) -> bool:
     return False
 
 
-def clean_request_json(response: requests.Response) -> StringDict:
+def clean_request_json(response: httpx.Response) -> StringDict:
     """
     Checks and correct the response error format depending on the two possible kind of returns.
 
     :param response: Raw request response.
     :return: Returns the job error if the error is due to parsing, returns the http error otherwise.
     """
     response_json = response.json()
-    if response.status_code < 200 or response.status_code > 302:
+    if response.is_error:
         response_json["status_code"] = response.status_code
         return response_json
     corrected_json = response_json

mindee/pdf/extracted_pdf.py

@@ -22,10 +22,10 @@ def get_page_count(self) -> int:
         try:
             pdf = pdfium.PdfDocument(self.pdf_bytes)
             return len(pdf)
-        except Exception as exc:
+        except Exception as e:
             raise MindeeError(
                 "Could not retrieve page count from Extracted PDF object."
-            ) from exc
+            ) from e
 
     def save_to_file(self, output_path: Path | str):
         """

mindee/v1/client.py

@@ -225,8 +225,8 @@ def load_prediction(
             if local_response.as_dict.get("job"):
                 return AsyncPredictResponse(product_class, local_response.as_dict)
             return PredictResponse(product_class, local_response.as_dict)
-        except KeyError as exc:
-            raise MindeeError("No prediction found in local response.") from exc
+        except KeyError as e:
+            raise MindeeError("No prediction found in local response.") from e
 
     def parse_queued(
         self,