implements sdk improvements of stresstest

Author: Johannes Hötter

README.md

@@ -32,7 +32,7 @@ Once you installed the package, you can create a `Client` object from any Python
 ```python
 from kern import Client
 
-user_name = "your-username"
+user_name = "your-username" # this is the email you log in with
 password = "your-password"
 project_id = "your-project-id" # can be found in the URL of the web application
 
@@ -111,6 +111,8 @@ file_path = "my/file/path/data.json"
 upload_was_successful = client.post_file_import(file_path)
 ```
 
+We use Pandas to process the data you upload, so you can also provide `import_file_options` for the file type you use. Currently, you need to provide them as a `\n`-separated string (e.g. `"quoting=1\nsep=';'"`). We'll adapt this in the future to work with dictionaries instead.
+
 Alternatively, you can `kern push <path-to-your-file>` via CLI, given that you have provided the `secrets.json` file in the same directory.
 
 **Make sure that you've selected the correct project beforehand, and fit the data schema of existing records in your project!**

kern/__init__.py

@@ -14,7 +14,7 @@ class Client:
     """Client object which can be used to directly address the Kern AI API.
 
     Args:
-        user_name (str): Your username for the application.
+        user_name (str): Your username (email) for the application.
         password (str): The respective password. Do not share this!
         project_id (str): The link to your project. This can be found in the URL in an active project.
         uri (str, optional): Link to the host of the application. Defaults to "https://app.kern.ai".
@@ -26,6 +26,7 @@ class Client:
     def __init__(
         self, user_name: str, password: str, project_id: str, uri=settings.DEFAULT_URI
     ):
+        msg.info(f"Connecting to {uri}")
         settings.set_base_uri(uri)
         self.session_token = authentication.create_session_token(
             user_name=user_name, password=password
@@ -37,25 +38,33 @@ def __init__(
             raise exceptions.get_api_exception_class(401)
         self.project_id = project_id
 
+        self.get_project_details()
+
     @classmethod
-    def from_secrets_file(cls, path_to_file: str):
+    def from_secrets_file(cls, path_to_file: str, project_id: Optional[str] = None):
         """Creates a Client object from a secrets file.
 
         Args:
             path_to_file (str): Path to the secrets file.
+            project_id (Optional[str], optional): The link to your project. This can be found in the URL in an active project. Defaults to None. In that case, it will read the project id from the file
 
         Returns:
-            Client: kern.Client object.
+            kern.Client: Client object.
         """
         with open(path_to_file, "r") as file:
             content = json.load(file)
+
         uri = content.get("uri")
         if uri is None:
             uri = settings.DEFAULT_URI
+
+        if project_id is None:
+            project_id = content["project_id"]
+
         return cls(
             user_name=content["user_name"],
             password=content["password"],
-            project_id=content["project_id"],
+            project_id=project_id,
             uri=uri,
         )
 
@@ -116,12 +125,9 @@ def get_record_export(
 
         if tokenize:
             tokenize_attributes = []
-            for column in df.columns:
-                if "__confidence" in column:
-                    dtype = type(df[column].iloc[0])
-                    if dtype == list:
-                        attribute = column.split("__")[0]
-                        tokenize_attributes.append(attribute)
+            for attribute in self.get_project_details()["attributes"]:
+                if attribute["data_type"] == "TEXT":
+                    tokenize_attributes.append(attribute["name"])
 
             if len(tokenize_attributes) > 0:
                 tokenizer_package = self.get_project_details()["tokenizer"]
@@ -142,7 +148,7 @@ def get_record_export(
                     )
 
             else:
-                msg.info("No tokenization necessary.")
+                msg.warn("There are no attributes that can be tokenized in this project.")
 
         if download_to is not None:
             df.to_json(download_to, orient="records")

kern/exceptions.py

@@ -5,9 +5,7 @@
 class SDKError(Exception):
     def __init__(self, message: Optional[str] = None):
         if message is None:
-            message = (
-                "Please check the SDK documentation at https://docs.kern.ai/reference."
-            )
+            message = "Please check the SDK documentation at https://github.com/code-kern-ai/kern-python"
         super().__init__(message)
 
 
@@ -20,6 +18,10 @@ class UnauthorizedError(SDKError):
 class NotFoundError(SDKError):
     pass
 
+class UnknownProjectError(SDKError):
+    def __init__(self, project_id: str):
+        super().__init__(message=f"Could not find project '{project_id}'. Please check your input.")
+
 
 # 500 Server Error
 class InternalServerError(SDKError):
@@ -29,10 +31,15 @@ class InternalServerError(SDKError):
 class FileImportError(Exception):
     pass
 
+# mirror this from the rest api class ErrorCodes
+class ErrorCodes:
+    UNRECOGNIZED_USER = "UNRECOGNIZED_USER" # not actively used in SDK
+    PROJECT_NOT_FOUND = "PROJECT_NOT_FOUND"
+
 
 RESPONSE_CODES_API_EXCEPTION_MAP = {
     401: {"*": UnauthorizedError},
-    404: {"*": NotFoundError},
+    404: {"*": NotFoundError, ErrorCodes.PROJECT_NOT_FOUND: UnknownProjectError},
     500: {"*": InternalServerError},
 }