fix(config): refactor API constants into config file

Author: João Guerreiro

Pipfile

@@ -4,9 +4,7 @@ verify_ssl = true
 name = "pypi"
 
 [packages]
-marshmallow = "~=3.5"
 pygitguardian = {editable = true,path = "."}
-requests = ">=2.21.0"
 
 [dev-packages]
 black = "==19.10b0"
@@ -19,6 +17,3 @@ nose = "*"
 pre-commit = "*"
 vcrpy = "*"
 vcrpy-unittest = "*"
-
-[pipenv]
-allow_prereleases = true

README.md

@@ -33,11 +33,13 @@ pipenv install pygitguardian
 poetry add pygitguardian
 ```
 
-Check [examples/content_scan.py](examples/content_scan.py) for an example usage of the API.
+## Examples
+
+Check [examples/](examples/) for an example usages of the API.
 
 ### Dependencies
 
-Py-gitguardian depends on these excelent libraries:
+Py-gitguardian depends on these excellent libraries:
 
 - `requests` - HTTP client
 - `marshmallow` - Request (de)serialization and input validation

examples/content_scan.py

@@ -7,7 +7,7 @@
 from pygitguardian import GGClient
 
 
-API_KEY = os.getenv("GG_APIKEY")
+API_KEY = os.getenv("GG_API_KEY")
 FILENAME = ".env"
 DOCUMENT = """
     import urllib.request

pygitguardian/client.py

@@ -6,15 +6,11 @@
 from marshmallow import Schema
 from requests import Response, Session, codes
 
+from .config import DEFAULT_API_VERSION, DEFAULT_BASE_URI, DEFAULT_TIMEOUT
 from .models import Detail, ScanResult
 from .schemas import DetailSchema, DocumentSchema, ScanResultSchema
 
 
-_BASE_URI = "https://api.gitguardian.com"
-_API_VERSION = "v1"
-_DEFAULT_TIMEOUT = 20.0  # 20s default timeout
-
-
 class GGClient:
     DETAIL_SCHEMA = DetailSchema()
     DOCUMENT_SCHEMA = DocumentSchema()
@@ -27,7 +23,7 @@ def __init__(
         base_uri: Optional[str] = None,
         session: Optional[requests.Session] = None,
         user_agent: Optional[str] = None,
-        timeout: Optional[float] = _DEFAULT_TIMEOUT,
+        timeout: Optional[float] = DEFAULT_TIMEOUT,
     ) -> "GGClient":
         """
         :param token: APIKey to be added to requests
@@ -43,7 +39,7 @@ def __init__(
             if not base_uri.startswith(("http://", "https://")):
                 raise ValueError("Invalid protocol, prepend with http:// or https://")
         else:
-            base_uri = _BASE_URI
+            base_uri = DEFAULT_BASE_URI
 
         if not isinstance(token, str):
             raise TypeError("Missing token string")
@@ -70,7 +66,7 @@ def request(
         method: str,
         endpoint: str,
         schema: Schema = None,
-        version: str = _API_VERSION,
+        version: str = DEFAULT_API_VERSION,
         many: bool = False,
         **kwargs
     ) -> Tuple[Any, Response]:
@@ -104,7 +100,7 @@ def post(
         endpoint: str,
         data: str = None,
         schema: Schema = None,
-        version: str = _API_VERSION,
+        version: str = DEFAULT_API_VERSION,
         many: bool = False,
         **kwargs
     ) -> Tuple[Any, Response]:
@@ -122,7 +118,7 @@ def get(
         self,
         endpoint: str,
         schema: Schema = None,
-        version: str = _API_VERSION,
+        version: str = DEFAULT_API_VERSION,
         many: bool = False,
         **kwargs
     ) -> Tuple[Any, Response]:
@@ -166,7 +162,7 @@ def multi_content_scan(
         if all(isinstance(doc, dict) for doc in documents):
             request_obj = self.DOCUMENT_SCHEMA.load(documents, many=True)
         else:
-            raise TypeError("documents must be a dict")
+            raise TypeError("each document must be a dict")
 
         obj, resp = self.post(
             endpoint="multiscan",

pygitguardian/config.py

@@ -0,0 +1,5 @@
+DEFAULT_BASE_URI = "https://api.gitguardian.com"
+DEFAULT_API_VERSION = "v1"
+DEFAULT_TIMEOUT = 20.0  # 20s default timeout
+
+DOCUMENT_SIZE_THRESHOLD_BYTES = 1048576  # 1MB

pygitguardian/schemas.py

@@ -13,12 +13,10 @@
     validates,
 )
 
+from .config import DOCUMENT_SIZE_THRESHOLD_BYTES
 from .models import Detail, Match, PolicyBreak, ScanResult
 
 
-MB = 1048576
-
-
 class DocumentSchema(Schema):
     class Meta:
         unknown = EXCLUDE
@@ -32,13 +30,15 @@ def validate_document(self, document: str) -> str:
         validate that document is smaller than scan limit
         """
         encoded = document.encode("utf-8")
-        if len(encoded) > MB:
+        if len(encoded) > DOCUMENT_SIZE_THRESHOLD_BYTES:
             raise ValidationError(
-                "This file exceeds the maximum allowed size of {}B".format(MB)
+                "file exceeds the maximum allowed size of {}B".format(
+                    DOCUMENT_SIZE_THRESHOLD_BYTES
+                )
             )
 
         if "\x00" in document:
-            raise ValidationError("Document has null characters")
+            raise ValidationError("document has null characters")
 
         return document
 

setup.py

@@ -26,7 +26,7 @@ def readme():
     version=__version__,
     packages=find_packages(exclude=["tests"]),
     description=readme(),
-    install_requires=["marshmallow", "requests"],
+    install_requires=["marshmallow>=3.5", "requests>=2"],
     include_package_data=True,
     author="GitGuardian",
     author_email="[email protected]",

tests/test_client.py

@@ -5,7 +5,7 @@
 from vcr_unittest import VCRTestCase
 
 from pygitguardian import GGClient, ScanResultSchema
-from pygitguardian.client import _BASE_URI
+from pygitguardian.config import DEFAULT_BASE_URI
 
 
 FILENAME = ".env"
@@ -225,7 +225,7 @@ def test_client_creation(self):
                     if entry.uri:
                         self.assertEqual(client.base_uri, entry.uri)
                     else:
-                        self.assertEqual(client.base_uri, _BASE_URI)
+                        self.assertEqual(client.base_uri, DEFAULT_BASE_URI)
                     self.assertEqual(client.token, entry.token)
                     self.assertTrue(
                         entry.user_agent in client.session.headers["User-Agent"],