sws-api-client 2.22.2__tar.gz → 2.23.0b2__tar.gz

{sws_api_client-2.22.2 → sws_api_client-2.23.0b2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.5
 Name: sws_api_client
-Version: 2.22.2
+Version: 2.23.0b2
 Author-email: Matteo <matteo.terrinoni@fao.org>, Daniele <daniele.mansillo@fao.org>
 License-File: LICENSE
 Requires-Python: <4,>=3.10

{sws_api_client-2.22.2 → sws_api_client-2.23.0b2}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "sws_api_client"
-version = "2.22.2"
+version = "2.23.0-beta.2"
 description = ""
 authors = [
     { name = "Matteo", email = "matteo.terrinoni@fao.org" },

{sws_api_client-2.22.2 → sws_api_client-2.23.0b2}/sws_api_client/__init__.py

@@ -57,4 +57,4 @@ from .tags import Tags
 from .tasks import PluginPayload, TaskDataset, TaskInfo, TaskManager, TaskResponse
 from .harvester import Harvester
 
-__version__ = "2.22.2"
+__version__ = "2.23.0-beta.2"

{sws_api_client-2.22.2 → sws_api_client-2.23.0b2}/sws_api_client/files.py

@@ -25,10 +25,10 @@ import os
 import warnings
 from datetime import datetime
 from pathlib import Path
-from typing import Dict, List, Literal, Optional, Union
+from typing import Literal
 
 import requests
-from pydantic import BaseModel
+from pydantic import ValidationError
 
 from sws_api_client.case_agnostic_model import CaseAgnosticModel
 from sws_api_client.sws_api_client import SwsApiClient
@@ -39,6 +39,31 @@ logger = logging.getLogger(__name__)
 FileObjectType = Literal["domain", "dataset", "datatable"]
 
 
+DEFAULT_MIME = "application/octet-stream"
+MIME_EXTENSIONS_LOOKUP = {
+    "txt": "text/plain",
+    "csv": "text/csv",
+    "json": "application/json",
+    "xml": "application/xml",
+    "zip": "application/zip",
+    "pdf": "application/pdf",
+    "xls": "application/vnd.ms-excel",
+    "xlsx": "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet",
+    "doc": "application/msword",
+    "docx": "application/vnd.openxmlformats-officedocument.wordprocessingml.document",
+    "py": "text/x-python",
+    "r": "text/x-r",
+    "md": "text/markdown",
+    "ppt": "application/vnd.ms-powerpoint",
+    "pptx": "application/vnd.openxmlformats-officedocument.presentationml.presentation",
+    "png": "image/png",
+    "jpg": "image/jpeg",
+    "jpeg": "image/jpeg",
+    "gif": "image/gif",
+    "svg": "image/svg+xml",
+}
+
+
 class CreatorInfo(CaseAgnosticModel):
     email: str | None = None
     username: str | None = None
@@ -672,7 +697,9 @@ class Files:
 
             # Check if local_path is a directory or ends with /
             # If so, derive filename from remote path
-            if str(local_path).endswith("/") or (local_path_obj.exists() and local_path_obj.is_dir()):
+            if str(local_path).endswith("/") or (
+                local_path_obj.exists() and local_path_obj.is_dir()
+            ):
                 remote_filename = os.path.basename(path)
                 local_path_obj = local_path_obj / remote_filename
 
@@ -681,13 +708,17 @@ class Files:
                 local_path_obj.parent.mkdir(parents=True, exist_ok=True)
 
             # Download content from presigned URL and save to file
-            logger.debug(f"Downloading file from {signed_response.url} to {local_path_obj}")
+            logger.debug(
+                f"Downloading file from {signed_response.url} to {local_path_obj}"
+            )
             resp = requests.get(signed_response.url)
             resp.raise_for_status()
 
             with open(local_path_obj, "wb") as f:
                 f.write(resp.content)
-            logger.info(f"Downloaded '{path}' -> {local_path_obj} ({len(resp.content)} bytes)")
+            logger.info(
+                f"Downloaded '{path}' -> {local_path_obj} ({len(resp.content)} bytes)"
+            )
 
         return signed_response
 
@@ -708,6 +739,63 @@ class Files:
         response = self.sws_client.discoverable.post("session_api", url, data=body)
         return SignedUploadUrlResponse(**response)
 
+    def _extract_file_extension(self, path: Path) -> str:
+        return "".join([c.lstrip(".") for c in path.suffixes])
+
+    def _get_content_type_from_file_extension(self, extension: str) -> str:
+        if extension not in MIME_EXTENSIONS_LOOKUP:
+            logger.warning(
+                "Unable to identify the content type for a file with extension %s. Defaulting to %s",
+                extension,
+                DEFAULT_MIME,
+            )
+        return MIME_EXTENSIONS_LOOKUP.get(extension, DEFAULT_MIME)
+
+    def _file_exists(
+        self, object_type: FileObjectType, object_id: str, *, path: str
+    ) -> bool:
+        try:
+            response = self.list(
+                object_type=object_type, object_id=object_id, path=path
+            )
+            return path in [f.path for f in response.files]
+        except ValidationError:
+            logger.error("Unexpected response")
+            return False
+
+    def upload_file(
+        self,
+        object_type: FileObjectType,
+        object_id: str,
+        *,
+        path: str,
+        local_path: str,
+        new_version: bool = False,
+    ) -> SignedUploadUrlResponse:
+        extension = self._extract_file_extension(Path(local_path))
+        content_type = self._get_content_type_from_file_extension(extension)
+        if not new_version:
+            file_exists = self._file_exists(
+                object_type=object_type, object_id=object_id, path=path
+            )
+            if file_exists:
+                raise FileExistsError(f"File already exists at {path}")
+        upload_response = self.upload_url(
+            object_type=object_type,
+            object_id=object_id,
+            path=path,
+            content_type=content_type,
+        )
+        with open(local_path, "rb") as f:
+            put_response = requests.put(
+                upload_response.url,
+                data=f,
+                headers=upload_response.requiredHeaders or {},
+            )
+        put_response.raise_for_status()
+        logger.info("Uploaded file %s at %s", local_path, path)
+        return upload_response
+
     def create_folder_url(
         self,
         object_type: FileObjectType,
@@ -778,7 +866,6 @@ class Files:
         """Delete a specific S3 object version."""
         url = f"/files/{object_type}/{object_id}/versions"
         body = {"path": path, "versionId": version_id}
-
         response = self.sws_client.discoverable.delete("session_api", url, data=body)
         return DeleteVersionResponse(**response)