keep original filenames (#77)

* keep original filenames

Closes #6

Update CHANGELOG.md

ignore .python-version file for pyenv-virtualenv

* remove keep_original_filenames from CLI

keep_original_filenames is task specific, not invocation specific

* update CHANGELOG.md

* Update CHANGELOG.md

---------

Co-authored-by: Pete Gadomski <pete.gadomski@gmail.com>

.gitignore

@@ -85,7 +85,7 @@ ipython_config.py
 # pyenv
 #   For a library or package, you might want to ignore these files since the code is
 #   intended to run in multiple environments; otherwise, check them in:
-# .python-version
+.python-version
 
 # pipenv
 #   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.

CHANGELOG.md

@@ -12,6 +12,8 @@ and this project adheres to [Semantic Versioning](http://semver.org/spec/v2.0.0.
 - ([#72]) Given that `_get_file` is part of the `AsyncFileSystem` spec, this
   adds the synchronous `get_file` as a way to retrieve files if `_get_file` is
   not found.
+- ([#77](https://github.com/stac-utils/stac-task/pull/77)) Added option `keep_original_filenames` to download routines to
+  support legacy applications dependent on filename specifics.
 
 ## [v0.3.0] - 2023-12-20
 

stactask/asset_io.py

@@ -41,6 +41,7 @@ async def download_item_assets(
     overwrite: bool = False,
     path_template: str = "${collection}/${id}",
     absolute_path: bool = False,
+    keep_original_filenames: bool = False,
     **kwargs: Any,
 ) -> Item:
     _assets = item.assets.keys() if assets is None else assets
@@ -61,8 +62,11 @@ async def download_item_assets(
         href = item.assets[a].href
 
         # local filename
-        ext = os.path.splitext(href)[-1]
-        new_href = os.path.join(path, a + ext)
+        if keep_original_filenames:
+            basename = os.path.basename(href)
+        else:
+            basename = a + os.path.splitext(href)[1]
+        new_href = os.path.join(path, basename)
         if absolute_path:
             new_href = os.path.abspath(new_href)
 

stactask/task.py

@@ -217,33 +217,64 @@ def download_item_assets(
         self,
         item: Item,
         path_template: str = "${collection}/${id}",
+        keep_original_filenames: bool = False,
         **kwargs: Any,
     ) -> Item:
-        """Download provided asset keys for all items in payload. Assets are
-        saved in workdir in a directory named by the Item ID, and the items are
-        updated with the new asset hrefs.
+        """Download provided asset keys for the given item. Assets are
+        saved in workdir in a directory (as specified by path_template), and
+        the items are updated with the new asset hrefs.
 
         Args:
-            assets (Optional[List[str]], optional): List of asset keys to
-                download. Defaults to all assets.
+            item (pystac.Item): STAC Item for which assets need be downloaded.
+            assets (Optional[List[str]]): List of asset keys to download.
+                Defaults to all assets.
+            path_template (Optional[str]): String to be interpolated to specify
+                where to store downloaded files.
+            keep_original_filenames (Optional[bool]): Controls whether original
+                file names should be used, or asset key + extension.
         """
         outdir = str(self._workdir / path_template)
         loop = asyncio.get_event_loop()
         item = loop.run_until_complete(
-            download_item_assets(item, path_template=outdir, **kwargs)
+            download_item_assets(
+                item,
+                path_template=outdir,
+                keep_original_filenames=keep_original_filenames,
+                **kwargs,
+            )
         )
         return item
 
     def download_items_assets(
         self,
         items: Iterable[Item],
         path_template: str = "${collection}/${id}",
+        keep_original_filenames: bool = False,
         **kwargs: Any,
     ) -> List[Item]:
+        """Download provided asset keys for the given items. Assets are
+        saved in workdir in a directory (as specified by path_template), and
+        the items are updated with the new asset hrefs.
+
+        Args:
+            items (List[pystac.Item]): List of STAC Items for which assets need
+                be downloaded.
+            assets (Optional[List[str]]): List of asset keys to download.
+                Defaults to all assets.
+            path_template (Optional[str]): String to be interpolated to specify
+                where to store downloaded files.
+            keep_original_filenames (Optional[bool]): Controls whether original
+                file names should be used, or asset key + extension.
+        """
         outdir = str(self._workdir / path_template)
         loop = asyncio.get_event_loop()
         items = loop.run_until_complete(
-            download_items_assets(items, path_template=outdir, **kwargs)
+            download_items_assets(
+                items,
+                path_template=outdir,
+                keep_original_filenames=keep_original_filenames,
+                **kwargs,
+            )
         )
         return list(items)
 

tests/test_task_download.py

@@ -35,6 +35,21 @@ def test_download_item_asset(tmp_path: Path, item_collection: Dict[str, Any]) ->
     assert filename.is_file() is True
 
 
+def test_download_keep_original_filenames(
+    tmp_path: Path, item_collection: Dict[str, Any]
+) -> None:
+    t = NothingTask(
+        item_collection,
+        workdir=tmp_path / "test-task-download-item-asset",
+    )
+    item = t.download_item_assets(
+        t.items[0], assets=["tileinfo_metadata"], keep_original_filenames=True
+    ).to_dict()
+    fname = item["assets"]["tileinfo_metadata"]["href"]
+    filename = Path(fname)
+    assert filename.name == "tileInfo.json"
+
+
 def test_download_item_asset_local(
     tmp_path: Path, item_collection: Dict[str, Any]
 ) -> None: