OpenAI API key format validation (microsoft#44)

* Updated to expand acceptable OpenAI API key format * Update to ChromaDB version required for RetrieveChatTest
randombet · Sep 28, 2024 · cd64eb3 · cd64eb3
1 parent 3a683fc
commit cd64eb3
Show file tree

Hide file tree

Showing 4 changed files with 16 additions and 7 deletions.
diff --git a/autogen/oai/openai_utils.py b/autogen/oai/openai_utils.py
@@ -106,15 +106,17 @@ def get_key(config: Dict[str, Any]) -> str:
 
 
 def is_valid_api_key(api_key: str) -> bool:
-    """Determine if input is valid OpenAI API key.
+    """Determine if input is valid OpenAI API key. As of 2024-09-24 there's no official definition of the key structure
+    so we will allow anything starting with "sk-" and having at least 48 alphanumeric (plus underscore and dash) characters.
+    Keys are known to start with "sk-", "sk-proj", "sk-None", and "sk-svcaat"
 
     Args:
         api_key (str): An input string to be validated.
 
     Returns:
         bool: A boolean that indicates if input is valid OpenAI API key.
     """
-    api_key_re = re.compile(r"^sk-([A-Za-z0-9]+(-+[A-Za-z0-9]+)*-)?[A-Za-z0-9]{32,}$")
+    api_key_re = re.compile(r"^sk-[A-Za-z0-9_-]{48,}$")
     return bool(re.fullmatch(api_key_re, api_key))
 
 

diff --git a/setup.py b/setup.py
@@ -49,7 +49,7 @@
 
 retrieve_chat = [
     "protobuf==4.25.3",
-    "chromadb",
+    "chromadb==0.5.3",
     "sentence_transformers",
     "pypdf",
     "ipython",

diff --git a/test/conftest.py b/test/conftest.py
@@ -10,7 +10,7 @@
 skip_redis = False
 skip_docker = False
 reason = "requested to skip"
-MOCK_OPEN_AI_API_KEY = "sk-mockopenaiAPIkeyinexpectedformatfortestingonly"
+MOCK_OPEN_AI_API_KEY = "sk-mockopenaiAPIkeysinexpectedformatsfortestingonly"
 
 
 # Registers command-line options like '--skip-openai' and '--skip-redis' via pytest hook.

diff --git a/test/oai/test_utils.py b/test/oai/test_utils.py
@@ -427,15 +427,22 @@ def test_is_valid_api_key():
     assert not is_valid_api_key("sk-asajsdjsd2")
     assert not is_valid_api_key("FooBar")
     assert not is_valid_api_key("sk-asajsdjsd22372%23kjdfdfdf2329ffUUDSDS")
-    assert is_valid_api_key("sk-asajsdjsd22372X23kjdfdfdf2329ffUUDSDS")
+    assert is_valid_api_key(
+        "sk-1b7n_YEvn7mmiUqkipH-JXk7DvqYoacDoe6Ae8v212T3BlbkFJWFRrGd3ZHN9GpZWX9Zx13gLolZ3NqcMFvb7ov5tzeA"
+    )  # correct format, not real
+    assert is_valid_api_key(
+        "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"
+    )  # correct format, not real
     assert is_valid_api_key("sk-asajsdjsd22372X23kjdfdfdf2329ffUUDSDS1212121221212sssXX")
     assert is_valid_api_key("sk-proj-asajsdjsd22372X23kjdfdfdf2329ffUUDSDS12121212212")
+    assert is_valid_api_key("sk-None-asajsdjsd22372X23kjdfdfdf2329ffUUDSDS12121212212")
+    assert is_valid_api_key("sk-svcaat-asajsdjsd22372X23kjdfdfdf2329ffUUDSDS12121212212")
     assert is_valid_api_key("sk-0-asajsdjsd22372X23kjdfdfdf2329ffUUDSDS12121212212")
     assert is_valid_api_key("sk-aut0gen-asajsdjsd22372X23kjdfdfdf2329ffUUDSDS12121212212")
     assert is_valid_api_key("sk-aut0-gen-asajsdjsd22372X23kjdfdfdf2329ffUUDSDS12121212212")
     assert is_valid_api_key("sk-aut0--gen-asajsdjsd22372X23kjdfdfdf2329ffUUDSDS12121212212")
-    assert not is_valid_api_key("sk-aut0-gen--asajsdjsd22372X23kjdfdfdf2329ffUUDSDS12121212212")
-    assert not is_valid_api_key("sk--aut0-gen-asajsdjsd22372X23kjdfdfdf2329ffUUDSDS12121212212")
+    assert is_valid_api_key("sk-aut0-gen--asajsdjsd22372X23kjdfdfdf2329ffUUDSDS12121212212")
+    assert is_valid_api_key("sk--aut0-gen-asajsdjsd22372X23kjdfdfdf2329ffUUDSDS12121212212")
     assert is_valid_api_key(MOCK_OPEN_AI_API_KEY)