Change windows extraction method

ecoset.py

@@ -149,14 +149,20 @@ class Ecoset(datasets.GeneratorBasedBuilder):
                 s3.download_file(urlinfo.netloc, urlinfo.path[1:], zip_path)
             # unzip
             # Expand-Archive -LiteralPath <PathToZipFile> -DestinationPath <PathToDestination>
-            subprocess.call(["unzip", "-n", "-P", password.encode("ascii"), "-o", zip_path, "-d", target_dir], shell=False)
+            if platform.system() in ("Linux", "Darwin"):
+                subprocess.call(["unzip", "-n", "-P", password.encode("ascii"), "-o", zip_path, "-d", target_dir], shell=False)
+            else:
+                subprocess.call(["tar.exe", "-xf", zip_path, "-C", target_dir, "--passphrase", password], shell=False)
+                # tar.exe -xf ecoset_uncompressed.zip -C uncompressed --passphrase ecoset_msjkk
             
             
         # take slow or very slow download depending on platform
-        if platform.system() in ("Linux", "Darwin"):
-            archives = dl_manager.download_custom(_URLS["codeocean"], subprocess_download)
-        else:
-            archives = dl_manager.download_custom(_URLS["codeocean"], s3_zipfile_download)
+        archives = dl_manager.download_custom(_URLS["codeocean"], subprocess_download)
+        # take slow or very slow download depending on platform
+        #if platform.system() in ("Linux", "Darwin"):
+        #    archives = dl_manager.download_custom(_URLS["codeocean"], subprocess_download)
+        #else:
+        #    archives = dl_manager.download_custom(_URLS["codeocean"], s3_zipfile_download)
             
         #archives = dl_manager.download(_URLS["codeocean"])
         print("Ecoset files are stored under: \n", archives)