Implement content type propagation (#226)

* Implement content type propagation

* add to changelog

* random filenames

* fix for custom endpoint

* check class and live before updating kwargs

* update header level

* Move content type to base client; add docstring

* move other client settings to its own page

* update version and changelog

* Correct docstring error

* fix typo in history

Co-authored-by: Jay Qi <2721979+jayqi@users.noreply.github.com>

* set release date

Co-authored-by: Jay Qi <2721979+jayqi@users.noreply.github.com>

Co-authored-by: Jay Qi <2721979+jayqi@users.noreply.github.com>

HISTORY.md

@@ -1,8 +1,9 @@
 # cloudpathlib Changelog
 
-## v0.7.2 (UNRELEASED)
+## v0.8.0 (2022-05-19)
 
  - Fixed pickling of `CloudPath` objects not working. ([Issue #223](https://github.com/drivendataorg/cloudpathlib/issues/223), [PR #224](https://github.com/drivendataorg/cloudpathlib/pull/224))
+ - Added functionality to [push the MIME (media) type to the content type property on cloud providers by default. ([Issue #222](https://github.com/drivendataorg/cloudpathlib/issues/222), [PR #226](https://github.com/drivendataorg/cloudpathlib/pull/226))
 
 ## v0.7.1 (2022-04-06)
 

cloudpathlib/azure/azblobclient.py

@@ -1,7 +1,8 @@
 from datetime import datetime
+import mimetypes
 import os
 from pathlib import Path, PurePosixPath
-from typing import Any, Dict, Iterable, Optional, Tuple, Union
+from typing import Any, Callable, Dict, Iterable, Optional, Tuple, Union
 
 
 from ..client import Client, register_client_class
@@ -12,7 +13,7 @@
 
 try:
     from azure.core.exceptions import ResourceNotFoundError
-    from azure.storage.blob import BlobServiceClient, BlobProperties
+    from azure.storage.blob import BlobServiceClient, BlobProperties, ContentSettings
 except ModuleNotFoundError:
     implementation_registry["azure"].dependencies_loaded = False
 
@@ -32,6 +33,7 @@ def __init__(
         connection_string: Optional[str] = None,
         blob_service_client: Optional["BlobServiceClient"] = None,
         local_cache_dir: Optional[Union[str, os.PathLike]] = None,
+        content_type_method: Optional[Callable] = mimetypes.guess_type,
     ):
         """Class constructor. Sets up a [`BlobServiceClient`](
         https://docs.microsoft.com/en-us/python/api/azure-storage-blob/azure.storage.blob.blobserviceclient?view=azure-python).
@@ -68,6 +70,8 @@ def __init__(
                 https://docs.microsoft.com/en-us/python/api/azure-storage-blob/azure.storage.blob.blobserviceclient?view=azure-python).
             local_cache_dir (Optional[Union[str, os.PathLike]]): Path to directory to use as cache
                 for downloaded files. If None, will use a temporary directory.
+            content_type_method (Optional[Callable]): Function to call to guess media type (mimetype) when
+                writing a file to the cloud. Defaults to `mimetypes.guess_type`. Must return a tuple (content type, content encoding).
         """
         if connection_string is None:
             connection_string = os.getenv("AZURE_STORAGE_CONNECTION_STRING", None)
@@ -86,14 +90,16 @@ def __init__(
                 "Credentials are required; see docs for options."
             )
 
-        super().__init__(local_cache_dir=local_cache_dir)
+        super().__init__(local_cache_dir=local_cache_dir, content_type_method=content_type_method)
 
     def _get_metadata(self, cloud_path: AzureBlobPath) -> Union["BlobProperties", Dict[str, Any]]:
         blob = self.service_client.get_blob_client(
             container=cloud_path.container, blob=cloud_path.blob
         )
         properties = blob.get_blob_properties()
 
+        properties["content_type"] = properties.content_settings.content_type
+
         return properties
 
     def _download_file(
@@ -220,7 +226,18 @@ def _upload_file(
             container=cloud_path.container, blob=cloud_path.blob
         )
 
-        blob.upload_blob(Path(local_path).read_bytes(), overwrite=True)  # type: ignore
+        extra_args = {}
+        if self.content_type_method is not None:
+            content_type, content_encoding = self.content_type_method(str(local_path))
+
+            if content_type is not None:
+                extra_args["content_type"] = content_type
+            if content_encoding is not None:
+                extra_args["content_encoding"] = content_encoding
+
+        content_settings = ContentSettings(**extra_args)
+
+        blob.upload_blob(Path(local_path).read_bytes(), overwrite=True, content_settings=content_settings)  # type: ignore
 
         return cloud_path
 

cloudpathlib/client.py

@@ -1,4 +1,5 @@
 import abc
+import mimetypes
 import os
 from pathlib import Path
 from tempfile import TemporaryDirectory
@@ -25,7 +26,11 @@ class Client(abc.ABC, Generic[BoundedCloudPath]):
     _cloud_meta: CloudImplementation
     _default_client = None
 
-    def __init__(self, local_cache_dir: Optional[Union[str, os.PathLike]] = None):
+    def __init__(
+        self,
+        local_cache_dir: Optional[Union[str, os.PathLike]] = None,
+        content_type_method: Optional[Callable] = mimetypes.guess_type,
+    ):
         self._cloud_meta.validate_completeness()
         # setup caching and local versions of file and track if it is a tmp dir
         self._cache_tmp_dir = None
@@ -34,6 +39,7 @@ def __init__(self, local_cache_dir: Optional[Union[str, os.PathLike]] = None):
             local_cache_dir = self._cache_tmp_dir.name
 
         self._local_cache_dir = Path(local_cache_dir)
+        self.content_type_method = content_type_method
 
     def __del__(self) -> None:
         # make sure temp is cleaned up if we created it

cloudpathlib/gs/gsclient.py

@@ -1,7 +1,8 @@
 from datetime import datetime
+import mimetypes
 import os
 from pathlib import Path, PurePosixPath
-from typing import Any, Dict, Iterable, Optional, TYPE_CHECKING, Tuple, Union
+from typing import Any, Callable, Dict, Iterable, Optional, TYPE_CHECKING, Tuple, Union
 
 from ..client import Client, register_client_class
 from ..cloudpath import implementation_registry
@@ -34,6 +35,7 @@ def __init__(
         project: Optional[str] = None,
         storage_client: Optional["StorageClient"] = None,
         local_cache_dir: Optional[Union[str, os.PathLike]] = None,
+        content_type_method: Optional[Callable] = mimetypes.guess_type,
     ):
         """Class constructor. Sets up a [`Storage
         Client`](https://googleapis.dev/python/storage/latest/client.html).
@@ -65,6 +67,8 @@ def __init__(
                 https://googleapis.dev/python/storage/latest/client.html).
             local_cache_dir (Optional[Union[str, os.PathLike]]): Path to directory to use as cache
                 for downloaded files. If None, will use a temporary directory.
+            content_type_method (Optional[Callable]): Function to call to guess media type (mimetype) when
+                writing a file to the cloud. Defaults to `mimetypes.guess_type`. Must return a tuple (content type, content encoding).
         """
         if application_credentials is None:
             application_credentials = os.getenv("GOOGLE_APPLICATION_CREDENTIALS")
@@ -81,7 +85,7 @@ def __init__(
             except DefaultCredentialsError:
                 self.client = StorageClient.create_anonymous_client()
 
-        super().__init__(local_cache_dir=local_cache_dir)
+        super().__init__(local_cache_dir=local_cache_dir, content_type_method=content_type_method)
 
     def _get_metadata(self, cloud_path: GSPath) -> Optional[Dict[str, Any]]:
         bucket = self.client.bucket(cloud_path.bucket)
@@ -94,6 +98,7 @@ def _get_metadata(self, cloud_path: GSPath) -> Optional[Dict[str, Any]]:
                 "etag": blob.etag,
                 "size": blob.size,
                 "updated": blob.updated,
+                "content_type": blob.content_type,
             }
 
     def _download_file(self, cloud_path: GSPath, local_path: Union[str, os.PathLike]) -> Path:
@@ -207,7 +212,12 @@ def _upload_file(self, local_path: Union[str, os.PathLike], cloud_path: GSPath)
         bucket = self.client.bucket(cloud_path.bucket)
         blob = bucket.blob(cloud_path.blob)
 
-        blob.upload_from_filename(str(local_path))
+        extra_args = {}
+        if self.content_type_method is not None:
+            content_type, _ = self.content_type_method(str(local_path))
+            extra_args["content_type"] = content_type
+
+        blob.upload_from_filename(str(local_path), **extra_args)
         return cloud_path
 
 

cloudpathlib/local/localclient.py

@@ -1,10 +1,11 @@
 import atexit
 from hashlib import md5
+import mimetypes
 import os
 from pathlib import Path, PurePosixPath
 import shutil
 from tempfile import TemporaryDirectory
-from typing import Iterable, List, Optional, Tuple, Union
+from typing import Callable, Dict, Iterable, List, Optional, Tuple, Union
 
 from ..client import Client
 from .localpath import LocalPath
@@ -21,14 +22,15 @@ def __init__(
         *args,
         local_cache_dir: Optional[Union[str, os.PathLike]] = None,
         local_storage_dir: Optional[Union[str, os.PathLike]] = None,
+        content_type_method: Optional[Callable] = mimetypes.guess_type,
         **kwargs,
     ):
         # setup caching and local versions of file. use default temp dir if not provided
         if local_storage_dir is None:
             local_storage_dir = self.get_default_storage_dir()
         self._local_storage_dir = Path(local_storage_dir)
 
-        super().__init__(local_cache_dir=local_cache_dir)
+        super().__init__(local_cache_dir=local_cache_dir, content_type_method=content_type_method)
 
     @classmethod
     def get_default_storage_dir(cls) -> Path:
@@ -132,6 +134,17 @@ def _upload_file(
         shutil.copy(local_path, dst)
         return cloud_path
 
+    def _get_metadata(self, cloud_path: "LocalPath") -> Dict:
+        # content_type is the only metadata we test currently
+        if self.content_type_method is None:
+            content_type_method = lambda x: (None, None)
+        else:
+            content_type_method = self.content_type_method
+
+        return {
+            "content_type": content_type_method(str(self._cloud_path_to_local(cloud_path)))[0],
+        }
+
 
 _temp_dirs_to_clean: List[TemporaryDirectory] = []
 

cloudpathlib/s3/s3client.py

@@ -1,6 +1,7 @@
+import mimetypes
 import os
 from pathlib import Path, PurePosixPath
-from typing import Any, Dict, Iterable, Optional, Tuple, Union
+from typing import Any, Callable, Dict, Iterable, Optional, Tuple, Union
 
 
 from ..client import Client, register_client_class
@@ -35,6 +36,7 @@ def __init__(
         local_cache_dir: Optional[Union[str, os.PathLike]] = None,
         endpoint_url: Optional[str] = None,
         boto3_transfer_config: Optional["TransferConfig"] = None,
+        content_type_method: Optional[Callable] = mimetypes.guess_type,
     ):
         """Class constructor. Sets up a boto3 [`Session`](
         https://boto3.amazonaws.com/v1/documentation/api/latest/reference/core/session.html).
@@ -63,6 +65,8 @@ def __init__(
                 Parameterize it to access a customly deployed S3-compatible object store such as MinIO, Ceph or any other.
             boto3_transfer_config (Optional[dict]): Instantiated TransferConfig for managing s3 transfers.
                 (https://boto3.amazonaws.com/v1/documentation/api/latest/reference/customizations/s3.html#boto3.s3.transfer.TransferConfig)
+            content_type_method (Optional[Callable]): Function to call to guess media type (mimetype) when
+                writing a file to the cloud. Defaults to `mimetypes.guess_type`. Must return a tuple (content type, content encoding).
         """
         endpoint_url = endpoint_url or os.getenv("AWS_ENDPOINT_URL")
         if boto3_session is not None:
@@ -93,7 +97,7 @@ def __init__(
 
         self.boto3_transfer_config = boto3_transfer_config
 
-        super().__init__(local_cache_dir=local_cache_dir)
+        super().__init__(local_cache_dir=local_cache_dir, content_type_method=content_type_method)
 
     def _get_metadata(self, cloud_path: S3Path) -> Dict[str, Any]:
         data = self.s3.ObjectSummary(cloud_path.bucket, cloud_path.key).get()
@@ -102,7 +106,7 @@ def _get_metadata(self, cloud_path: S3Path) -> Dict[str, Any]:
             "last_modified": data["LastModified"],
             "size": data["ContentLength"],
             "etag": data["ETag"],
-            "mime": data["ContentType"],
+            "content_type": data["ContentType"],
             "extra": data["Metadata"],
         }
 
@@ -250,7 +254,16 @@ def _remove(self, cloud_path: S3Path) -> None:
     def _upload_file(self, local_path: Union[str, os.PathLike], cloud_path: S3Path) -> S3Path:
         obj = self.s3.Object(cloud_path.bucket, cloud_path.key)
 
-        obj.upload_file(str(local_path), Config=self.boto3_transfer_config)
+        extra_args = {}
+
+        if self.content_type_method is not None:
+            content_type, content_encoding = self.content_type_method(str(local_path))
+            if content_type is not None:
+                extra_args["ContentType"] = content_type
+            if content_encoding is not None:
+                extra_args["ContentEncoding"] = content_encoding
+
+        obj.upload_file(str(local_path), Config=self.boto3_transfer_config, ExtraArgs=extra_args)
         return cloud_path
 
 

docs/docs/other_client_settings.md

@@ -0,0 +1,51 @@
+# Other `Client` settings
+
+## Content type guessing (`content_type_method`)
+
+All of the clients support passing a `content_type_method` when they are instantiated.  
+This is a method that is used to guess the [MIME (media) type](https://en.wikipedia.org/wiki/Media_type)
+(often called the "content type") of the file and set that on the cloud provider.
+
+By default, `content_type_method` use the Python built-in 
+[`guess_type`](https://docs.python.org/3/library/mimetypes.html#mimetypes.guess_type)
+to set this content type. This guesses based on the file extension, and may not always get the correct type.
+In these cases, you can set `content_type_method` to your own function that gets the proper type; for example, by 
+reading the file content or by looking it up in a dictionary of filename-to-media-type mappings that you maintain.
+
+If you set a custom method, it should follow the signature of `guess_type` and return a tuple of the form: 
+`(content_type, content_encoding)`; for example, `("text/css", None)`.
+
+If you set `content_type_method` to None, it will do whatever the default of the cloud provider's SDK does. This
+varies from provider to provider.
+
+Here is an example of using a custom `content_type_method`.
+
+```python
+import mimetypes
+from pathlib import Path
+
+from cloudpathlib import S3Client, CloudPath
+
+def my_content_type(path):
+    # do lookup for content types I define; fallback to
+    # guess_type for anything else
+    return {
+        ".potato": ("application/potato", None),
+    }.get(Path(path).suffix, mimetypes.guess_type(path))
+
+
+# create a client with my custom content type
+client = S3Client(content_type_method=my_content_type)
+
+# To use this same method for every cloud path, set our client as the default.
+# This is optional, and you could use client.CloudPath to create paths instead.
+client.set_as_default_client()
+
+# create a cloud path
+cp1 = CloudPath("s3://cloudpathlib-test-bucket/i_am_a.potato")
+cp1.write_text("hello")
+
+# check content type with boto3
+print(client.s3.Object(cp1.bucket, cp1.key).content_type)
+#> application/potato
+```

docs/mkdocs.yml

@@ -20,6 +20,7 @@ nav:
   - Authentication: "authentication.md"
   - Caching: "caching.ipynb"
   - AnyPath: "anypath-polymorphism.md"
+  - Other Client settings: "other_client_settings.md"
   - Testing code that uses cloudpathlib: "testing_mocked_cloudpathlib.ipynb"
   - Integrations: "integrations.md"
   - Changelog: "changelog.md"

setup.py

@@ -60,5 +60,5 @@ def load_requirements(path: Path):
         "Source Code": "https://github.com/drivendataorg/cloudpathlib",
     },
     url="https://github.com/drivendataorg/cloudpathlib",
-    version="0.7.1",
+    version="0.8.0",
 )