fix: 🐛 fixed API error when using base64

README.md

@@ -84,5 +84,5 @@ with open('/path/to/file', 'rb') as fp:
 
 From a base64
 ```python
-receipt_data = mindee_client.parse_receipt(base64_string, input_type="base64")
+receipt_data = mindee_client.parse_receipt(base64_string, input_type="base64", filename="receipt.jpg")
 ```

mindee/__init__.py

@@ -50,6 +50,7 @@ def parse_receipt(
         cut_pdf=True,
         include_words=False,
         cut_pdf_mode=3,
+        filename=None,
     ):
         """
         :param cut_pdf_mode: Number (between 1 and 3 incl.) of pages to reconstruct a pdf with.
@@ -60,6 +61,7 @@ def parse_receipt(
         :param cut_pdf: Automatically reconstruct pdf with more than 4 pages
         :param input_type: String in {'path', 'stream', 'base64'}
         :param file: Receipt filepath (allowed jpg, png, tiff, pdf)
+        :param filename: the name of the file (without the path)
         :param version: expense_receipt api version
         :return: Wrapped response with Receipts objects parsed
         """
@@ -68,7 +70,13 @@ def parse_receipt(
                 "Missing 'expense_receipt_token' arg in parse_receipt() function."
             )
 
-        input_file = Inputs(file, input_type, cut_pdf=cut_pdf, n_pdf_pages=cut_pdf_mode)
+        input_file = Inputs(
+            file,
+            input_type,
+            filename=filename,
+            cut_pdf=cut_pdf,
+            n_pdf_pages=cut_pdf_mode,
+        )
 
         response = Receipt.request(
             input_file,
@@ -109,6 +117,7 @@ def parse_passport(
         version="1",
         cut_pdf=True,
         cut_pdf_mode=3,
+        filename=None,
     ):
         """
         :param cut_pdf_mode: Number (between 1 and 3 incl.) of pages to reconstruct a pdf with.
@@ -118,6 +127,7 @@ def parse_passport(
         :param cut_pdf: Automatically reconstruct pdf with more than 4 pages
         :param input_type: String in {'path', 'stream', 'base64'}
         :param file: Passport filepath (allowed jpg, png, pdf)
+        :param filename: the name of the file (without the path)
         :param version: passport api version
         :return: Wrapped response with passports objects parsed
         """
@@ -126,7 +136,13 @@ def parse_passport(
                 "Missing 'passport_token' arg in parse_passport() function."
             )
 
-        input_file = Inputs(file, input_type, cut_pdf=cut_pdf, n_pdf_pages=cut_pdf_mode)
+        input_file = Inputs(
+            file,
+            input_type,
+            filename=filename,
+            cut_pdf=cut_pdf,
+            n_pdf_pages=cut_pdf_mode,
+        )
 
         response = Passport.request(input_file, self.passport_token, version)
 
@@ -139,6 +155,7 @@ def parse_license_plate(
         version="1",
         cut_pdf=True,
         cut_pdf_mode=3,
+        filename=None,
     ):
         """
         :param cut_pdf_mode: Number (between 1 and 3 incl.) of pages to reconstruct a pdf with.
@@ -148,6 +165,7 @@ def parse_license_plate(
         :param cut_pdf: Automatically reconstruct pdf with more than 4 pages
         :param input_type: String in {'path', 'stream', 'base64'}
         :param file: CarPlate filepath (allowed jpg, png, pdf)
+        :param filename: the name of the file (without the path)
         :param version: license_plates api version
         :return: Wrapped response with CarPlates objects parsed
         """
@@ -156,7 +174,13 @@ def parse_license_plate(
                 "Missing 'license_plate_token' arg in license_plate_token() function."
             )
 
-        input_file = Inputs(file, input_type, cut_pdf=cut_pdf, n_pdf_pages=cut_pdf_mode)
+        input_file = Inputs(
+            file,
+            input_type,
+            filename=filename,
+            cut_pdf=cut_pdf,
+            n_pdf_pages=cut_pdf_mode,
+        )
 
         response = CarPlate.request(input_file, self.license_plate_token, version)
 
@@ -170,6 +194,7 @@ def parse_invoice(
         cut_pdf=True,
         include_words=False,
         cut_pdf_mode=3,
+        filename=None,
     ):
         """
         :param cut_pdf_mode: Number (between 1 and 3 incl.) of pages to reconstruct a pdf with.
@@ -179,14 +204,21 @@ def parse_invoice(
         :param include_words: Bool, extract all words into http_response
         :param cut_pdf: Automatically reconstruct pdf with more than 4 pages
         :param input_type: String in {'path', 'stream', 'base64'}
-        :param file: Invoice filepath (allowed jpg, png, pdf)
+        :param file: Invoice full path (allowed jpg, png, pdf)
+        :param filename: the name of the file (without the path)
         :param version: invoices api version
         :return: Wrapped response with Invoices objects parsed
         """
         if not self.invoice_token:
             raise Exception("Missing 'invoice_token' arg in parse_invoice() function.")
 
-        input_file = Inputs(file, input_type, cut_pdf=cut_pdf, n_pdf_pages=cut_pdf_mode)
+        input_file = Inputs(
+            file,
+            input_type,
+            filename=filename,
+            cut_pdf=cut_pdf,
+            n_pdf_pages=cut_pdf_mode,
+        )
 
         response = Invoice.request(
             input_file, self.invoice_token, version, include_words
@@ -201,6 +233,7 @@ def parse_financial_document(
         cut_pdf=True,
         include_words=False,
         cut_pdf_mode=3,
+        filename=None,
     ):
         """
         :param cut_pdf_mode: Number (between 1 and 3 incl.) of pages to reconstruct a pdf with.
@@ -211,14 +244,21 @@ def parse_financial_document(
         :param cut_pdf: Automatically reconstruct pdf with more than 4 pages
         :param input_type: String in {'path', 'stream', 'base64'}
         :param file: Invoice or Receipt filepath (allowed jpg, png, pdf)
+        :param filename: the name of the file (without the path)
         :return: Wrapped response with FinancialDocument objects parsed
         """
         if not self.invoice_token or not self.expense_receipt_token:
             raise Exception(
                 "parse_invoice() function must include 'invoice_token' and 'expense_receipt_token' args."
             )
 
-        input_file = Inputs(file, input_type, cut_pdf=cut_pdf, n_pdf_pages=cut_pdf_mode)
+        input_file = Inputs(
+            file,
+            input_type,
+            filename=filename,
+            cut_pdf=cut_pdf,
+            n_pdf_pages=cut_pdf_mode,
+        )
 
         response = FinancialDocument.request(
             input_file,

mindee/inputs.py

@@ -1,9 +1,17 @@
 import io
 import os
-from base64 import decodebytes
+import base64
 from mimetypes import guess_type
 import pikepdf
 
+ALLOWED_EXTENSIONS = [
+    "image/png",
+    "image/jpg",
+    "image/jpeg",
+    "image/webp",
+    "application/pdf",
+]
+
 
 class Inputs:
     def __init__(
@@ -15,17 +23,11 @@ def __init__(
         :param filename: File name of the input
         :param cut_pdf: Automatically reconstruct pdf with more than 4 pages
         """
-        self.allowed_extensions = [
-            "image/png",
-            "image/jpg",
-            "image/jpeg",
-            "image/webp",
-            "application/pdf",
-        ]
         assert input_type in ["base64", "path", "stream", "dummy"]
         assert 0 < n_pdf_pages <= 3
 
         if input_type == "base64":
+            assert filename, "filename must be set"
             # Only for images
             self.file_object = Inputs.b64_to_stream(file)
             self.input_type = input_type
@@ -53,10 +55,9 @@ def __init__(
             self.filename = ""
             self.filepath = ""
             self.file_extension = ""
-        elif self.file_extension not in self.allowed_extensions:
-            raise Exception(
-                "File type not allowed, must be in {%s}"
-                % ", ".join(self.allowed_extensions)
+        elif self.file_extension not in ALLOWED_EXTENSIONS:
+            raise AssertionError(
+                "File type not allowed, must be in {%s}" % ", ".join(ALLOWED_EXTENSIONS)
             )
 
         if self.file_extension == "application/pdf":
@@ -88,12 +89,12 @@ def load(input_type, filename, filepath, file_extension):
         return file_input
 
     @staticmethod
-    def b64_to_stream(b64_string):
+    def b64_to_stream(b64_string: str):
         """
         :param b64_string: image base 64 string
         :return: stream from base64
         """
-        bytes_object = decodebytes(b64_string.encode("utf-8"))
+        bytes_object = base64.standard_b64decode(b64_string)
         return io.BytesIO(bytes_object)
 
     def count_pdf_pages(self):

tests/test_client.py

@@ -1,5 +1,6 @@
 import pytest
 from mindee import Client, Response, Receipt, Passport
+from mindee.http import HTTPException
 
 
 @pytest.fixture
@@ -61,68 +62,68 @@ def test_parse_license_plate_without_token(empty_client):
 
 
 def test_parse_receipt_with_wrong_filetype(dummy_client):
-    with pytest.raises(Exception):
+    with pytest.raises(AssertionError):
         dummy_client.parse_receipt("./tests/data/expense_receipts/receipt.jpga")
 
 
 def test_parse_invoice_with_wrong_filetype(dummy_client):
-    with pytest.raises(Exception):
+    with pytest.raises(AssertionError):
         dummy_client.parse_invoice("./tests/data/expense_receipts/receipt.jpga")
 
 
 def test_parse_financial_doc_with_wrong_filetype(dummy_client):
-    with pytest.raises(Exception):
+    with pytest.raises(AssertionError):
         dummy_client.parse_financial_document(
             "./tests/data/expense_receipts/receipt.jpga"
         )
 
 
 def test_parse_passport_with_wrong_filetype(dummy_client):
-    with pytest.raises(Exception):
+    with pytest.raises(AssertionError):
         dummy_client.parse_passport("./tests/data/expense_receipts/receipt.jpga")
 
 
 def test_parse_plate_with_wrong_filetype(dummy_client):
-    with pytest.raises(Exception):
+    with pytest.raises(AssertionError):
         dummy_client.parse_license_plate("./tests/data/expense_receipts/receipt.jpga")
 
 
 def test_parse_receipt_with_wrong_token(dummy_client):
-    with pytest.raises(Exception):
+    with pytest.raises(HTTPException):
         dummy_client.parse_receipt("./tests/data/expense_receipts/receipt.jpg")
 
 
 def test_parse_receipt_with_wrong_version(dummy_client):
-    with pytest.raises(Exception):
+    with pytest.raises(HTTPException):
         dummy_client.parse_receipt(
             "./tests/data/expense_receipts/receipt.jpg", version="4000"
         )
 
 
 def test_parse_invoice_with_wrong_token(dummy_client):
-    with pytest.raises(Exception):
+    with pytest.raises(HTTPException):
         dummy_client.parse_invoice("./tests/data/expense_receipts/receipt.jpg")
 
 
 def test_parse_financial_doc_with_wrong_token_jpg(dummy_client):
-    with pytest.raises(Exception):
+    with pytest.raises(HTTPException):
         dummy_client.parse_financial_document(
             "./tests/data/expense_receipts/receipt.jpg"
         )
 
 
 def test_parse_financial_doc_with_wrong_token_pdf(dummy_client):
-    with pytest.raises(Exception):
+    with pytest.raises(HTTPException):
         dummy_client.parse_financial_document("./tests/data/invoices/invoice.pdf")
 
 
 def test_parse_passport_with_wrong_token(dummy_client):
-    with pytest.raises(Exception):
+    with pytest.raises(HTTPException):
         dummy_client.parse_passport("./tests/data/expense_receipts/receipt.jpg")
 
 
 def test_parse_license_plate_with_wrong_token(dummy_client):
-    with pytest.raises(Exception):
+    with pytest.raises(HTTPException):
         dummy_client.parse_license_plate("./tests/data/license_plates/plate.png")
 
 
@@ -147,26 +148,33 @@ def test_response_with_passport_type():
 
 
 def test_request_with_filepath(dummy_client):
-    with pytest.raises(Exception):
+    with pytest.raises(HTTPException):
         dummy_client.parse_receipt(
             "./tests/data/expense_receipts/receipt.jpg", input_type="path"
         )
 
 
 def test_request_with_file(dummy_client):
-    with pytest.raises(Exception):
+    with pytest.raises(HTTPException):
         dummy_client.parse_receipt(
-            open("./tests/data/expense_receipts/receipt.jpg"), input_type="file"
+            open("./tests/data/expense_receipts/receipt.jpg", "rb"), input_type="stream"
         )
 
 
-def test_request_with_base64(dummy_client):
+def test_request_with_base64_no_filename(dummy_client):
     with open("./tests/data/expense_receipts/receipt.txt", "r") as fh:
         b64 = fh.read()
-    with pytest.raises(Exception):
+    with pytest.raises(AssertionError):
         dummy_client.parse_receipt(b64, input_type="base64")
 
 
+def test_request_with_base64(dummy_client):
+    with open("./tests/data/expense_receipts/receipt.txt", "r") as fh:
+        b64 = fh.read()
+    with pytest.raises(HTTPException):
+        dummy_client.parse_receipt(b64, input_type="base64", filename="receipt.txt")
+
+
 def test_request_without_raise_on_error(dummy_client_dont_raise):
     result = dummy_client_dont_raise.parse_receipt(
         "./tests/data/expense_receipts/receipt.jpg", input_type="path"
@@ -186,13 +194,13 @@ def test_request_without_raise_on_error_include_words(dummy_client_dont_raise):
 
 
 def test_request_with_file_wrong_type(dummy_client):
-    with pytest.raises(Exception):
+    with pytest.raises(AssertionError):
         dummy_client.parse_receipt(open("./tests/data/test.txt"), input_type="file")
 
-    with pytest.raises(Exception):
+    with pytest.raises(AssertionError):
         dummy_client.parse_receipt("./tests/data/test.txt", input_type="path")
 
 
-def test_mpdf_reconstruct(dummy_client):
-    with pytest.raises(Exception):
+def test_pdf_reconstruct(dummy_client):
+    with pytest.raises(HTTPException):
         dummy_client.parse_invoice("./tests/data/invoices/invoice_6p.pdf")