crypticorn 2.16.0__py3-none-any.whl → 2.17.0__py3-none-any.whl

crypticorn/hive/main.py

@@ -2,7 +2,7 @@ from __future__ import annotations
 import os
 import asyncio
 from pathlib import Path
-from typing import TYPE_CHECKING, Optional
+from typing import TYPE_CHECKING, Optional, Union, Coroutine, Any
 from crypticorn.hive import (
     ApiClient,
     Configuration,
@@ -20,12 +20,38 @@ if TYPE_CHECKING:
     from aiohttp import ClientSession
 
 
+class HiveClient:
+    """
+    A client for interacting with the Crypticorn Hive API.
+    """
+
+    config_class = Configuration
+
+    def __init__(
+        self,
+        config: Configuration,
+        http_client: Optional[ClientSession] = None,
+        is_sync: bool = False,
+    ):
+        self.config = config
+        self.base_client = ApiClient(configuration=self.config)
+        if http_client is not None:
+            self.base_client.rest_client.pool_manager = http_client
+        # Pass sync context to REST client for proper session management
+        self.base_client.rest_client.is_sync = is_sync
+        # Instantiate all the endpoint clients
+        self.models = ModelsApi(self.base_client, is_sync=is_sync)
+        self.data = DataApiWrapper(self.base_client, is_sync=is_sync)
+        self.status = StatusApi(self.base_client, is_sync=is_sync)
+        self.admin = AdminApi(self.base_client, is_sync=is_sync)
+
+
 class DataApiWrapper(DataApi):
     """
     A wrapper for the DataApi class.
     """
 
-    async def download_data(
+    def download_data(
         self,
         model_id: StrictInt,
         folder: Path = Path("data"),
@@ -33,17 +59,86 @@ class DataApiWrapper(DataApi):
         feature_size: Optional[FeatureSize] = None,
         *args,
         **kwargs,
-    ) -> list[Path]:
+    ) -> Union[list[Path], Coroutine[Any, Any, list[Path]]]:
         """
         Download data for model training. All three files (y_train, x_test, x_train) are downloaded and saved under e.g. FOLDER/v1/coin_1/*.feather.
         The folder will be created if it doesn't exist.
+        Works in both sync and async contexts.
 
         :param model_id: Model ID (required) (type: int)
         :param version: Data version. Default is the latest public version. (optional) (type: DataVersion)
         :param feature_size: The number of features in the data. Default is LARGE. (optional) (type: FeatureSize)
         :return: A list of paths to the downloaded files.
         """
-        response = await super().download_data(
+        if self.is_sync:
+            return self._download_data_wrapper_sync(
+                model_id=model_id,
+                folder=folder,
+                version=version,
+                feature_size=feature_size,
+                *args,
+                **kwargs,
+            )
+        else:
+            return self._download_data_wrapper_async(
+                model_id=model_id,
+                folder=folder,
+                version=version,
+                feature_size=feature_size,
+                *args,
+                **kwargs,
+            )
+
+    def _download_data_wrapper_sync(
+        self,
+        model_id: StrictInt,
+        folder: Path = Path("data"),
+        version: Optional[DataVersion] = None,
+        feature_size: Optional[FeatureSize] = None,
+        *args,
+        **kwargs,
+    ) -> list[Path]:
+        """
+        Download data for model training (sync version).
+        """
+        response = self._download_data_sync(
+            model_id=model_id,
+            version=version,
+            feature_size=feature_size,
+            *args,
+            **kwargs,
+        )
+        base_path = f"{folder}/v{response.version.value}/coin_{response.coin.value}/"
+        os.makedirs(base_path, exist_ok=True)
+
+        return [
+            download_file(
+                url=response.links.y_train,
+                dest_path=base_path + "y_train_" + response.target + ".feather",
+            ),
+            download_file(
+                url=response.links.x_test,
+                dest_path=base_path + "x_test_" + response.feature_size + ".feather",
+            ),
+            download_file(
+                url=response.links.x_train,
+                dest_path=base_path + "x_train_" + response.feature_size + ".feather",
+            ),
+        ]
+
+    async def _download_data_wrapper_async(
+        self,
+        model_id: StrictInt,
+        folder: Path = Path("data"),
+        version: Optional[DataVersion] = None,
+        feature_size: Optional[FeatureSize] = None,
+        *args,
+        **kwargs,
+    ) -> list[Path]:
+        """
+        Download data for model training (async version).
+        """
+        response = await self._download_data_async(
             model_id=model_id,
             version=version,
             feature_size=feature_size,
@@ -75,24 +170,3 @@ class DataApiWrapper(DataApi):
                 ),
             ]
         )
-
-
-class HiveClient:
-    """
-    A client for interacting with the Crypticorn Hive API.
-    """
-
-    config_class = Configuration
-
-    def __init__(
-        self, config: Configuration, http_client: Optional[ClientSession] = None
-    ):
-        self.config = config
-        self.base_client = ApiClient(configuration=self.config)
-        if http_client is not None:
-            self.base_client.rest_client.pool_manager = http_client
-        # Instantiate all the endpoint clients
-        self.models = ModelsApi(self.base_client)
-        self.data = DataApiWrapper(self.base_client)
-        self.status = StatusApi(self.base_client)
-        self.admin = AdminApi(self.base_client)

crypticorn/hive/utils.py

@@ -28,7 +28,7 @@ async def download_file(
 
         if file_size < total_size:
             # Download incomplete
-            logger.info(f" resuming download")
+            logger.info(" resuming download")
             resume_header = {"Range": f"bytes={file_size}-"}
             req = requests.get(
                 url,
@@ -45,7 +45,7 @@ async def download_file(
             file_size = 0
     else:
         # File does not exist, starting download
-        logger.info(f" starting download")
+        logger.info(" starting download")
 
     # write dataset to file and show progress bar
     pbar = tqdm.tqdm(