fix: unzip kaggle data (microsoft#464)

* unzip kaggle data

* read local_data_path from .env file

* fix build docs error

* recover azure-identity packages

* optimize code logic

* add error when downloading data from kaggle

---------

Co-authored-by: Xu Yang <peteryang@vip.qq.com>

Author: SunsetWolf

constraints/3.10.txt

@@ -1,3 +1,4 @@
+azure-identity==1.17.1
 dill==0.3.9
 pillow==10.4.0
 psutil==6.1.0

constraints/3.11.txt

@@ -1,3 +1,4 @@
+azure-identity==1.17.1
 dill==0.3.9
 pillow==10.4.0
 psutil==6.1.0

rdagent/app/kaggle/conf.py

@@ -44,7 +44,7 @@ class Config:
     competition: str = ""
     """Kaggle competition name, e.g., 'sf-crime'"""
 
-    local_data_path: str = "/data/userdata/share/kaggle"
+    local_data_path: str = ""
     """Folder storing Kaggle competition data"""
 
     if_action_choosing_based_on_UCB: bool = False

rdagent/core/exception.py

@@ -42,3 +42,9 @@ class ModelEmptyError(Exception):
     """
     Exceptions raised when no model is generated correctly
     """
+
+
+class KaggleError(Exception):
+    """
+    Exceptions raised when calling Kaggle API
+    """

rdagent/scenarios/kaggle/kaggle_crawler.py

@@ -15,6 +15,7 @@
 from selenium.webdriver.common.by import By
 
 from rdagent.app.kaggle.conf import KAGGLE_IMPLEMENT_SETTING
+from rdagent.core.exception import KaggleError
 from rdagent.core.prompts import Prompts
 from rdagent.log import rdagent_logger as logger
 from rdagent.oai.llm_utils import APIBackend
@@ -99,11 +100,28 @@ def kaggle_description_css_selectors() -> tuple[str, str]:
 def download_data(competition: str, local_path: str = KAGGLE_IMPLEMENT_SETTING.local_data_path) -> None:
     data_path = f"{local_path}/{competition}"
     if not Path(data_path).exists():
-        subprocess.run(["kaggle", "competitions", "download", "-c", competition, "-p", data_path])
-
-        # unzip data
-        with zipfile.ZipFile(f"{data_path}/{competition}.zip", "r") as zip_ref:
-            zip_ref.extractall(data_path)
+        try:
+            subprocess.run(
+                ["kaggle", "competitions", "download", "-c", competition, "-p", data_path],
+                check=True,
+                stderr=subprocess.PIPE,
+                stdout=subprocess.PIPE,
+            )
+        except subprocess.CalledProcessError as e:
+            logger.error(f"Download failed: {e}, stderr: {e.stderr}, stdout: {e.stdout}")
+            raise KaggleError(f"Download failed: {e}, stderr: {e.stderr}, stdout: {e.stdout}")
+
+    # unzip data
+    unzip_path = f"{local_path}/{competition}"
+    if not Path(unzip_path).exists():
+        unzip_data(unzip_file_path=f"{data_path}/{competition}.zip", unzip_target_path=unzip_path)
+        for sub_zip_file in Path(unzip_path).rglob("*.zip"):
+            unzip_data(sub_zip_file, unzip_target_path=unzip_path)
+
+
+def unzip_data(unzip_file_path: str, unzip_target_path: str) -> None:
+    with zipfile.ZipFile(unzip_file_path, "r") as zip_ref:
+        zip_ref.extractall(unzip_target_path)
 
 
 def leaderboard_scores(competition: str) -> list[float]: