AtpRemoteBlob: add width and height properties

populated for images, to be used in image embed aspectRatio. for snarfed/bridgy-fed#1571

README.md

@@ -103,8 +103,6 @@ Optional, only used in [com.atproto.repo](https://arroba.readthedocs.io/en/stabl
 
 _Breaking changes:_
 
-* `datastore_storage`:
-  * `DatastoreStorage`: add new `ndb_context_kwargs` constructor kwarg.
 * `repo`:
   * `apply_commit`, `apply_writes`: raise an exception if the repo is inactive.
 * `storage`:
@@ -114,8 +112,12 @@ _Breaking changes:_
 
 _Non-breaking changes:_
 * `datastore_storage`:
-  * `apply_commit`: handle deactivated repos.
-  * `create_repo`: propagate `Repo.status` into `AtpRepo`.
+  * `DatastoreStorage`:
+    * Add new `ndb_context_kwargs` constructor kwarg.
+    * `apply_commit`: handle deactivated repos.
+    * `create_repo`: propagate `Repo.status` into `AtpRepo`.
+  * `AtpRemoteBlob`:
+    * Add `width` and `height` properties, populated for images, to be used in image embed `aspectRatio` ([snarfed/bridgy-fed#1571](https://github.com/snarfed/bridgy-fed/issues/1571)).
 
 
 ### 0.7 - 2024-11-08

arroba/datastore_storage.py

@@ -1,6 +1,7 @@
 """Google Cloud Datastore implementation of repo storage."""
 from datetime import timezone
 from functools import wraps
+from io import BytesIO
 import json
 import logging
 import mimetypes
@@ -15,6 +16,7 @@
 from google.cloud.ndb.exceptions import ContextError
 from lexrpc import ValidationError
 from multiformats import CID, multicodec, multihash
+from PIL import Image, UnidentifiedImageError
 
 from .mst import MST
 from .repo import Repo
@@ -298,6 +300,11 @@ class AtpRemoteBlob(ndb.Model):
     size = ndb.IntegerProperty(required=True)
     mime_type = ndb.StringProperty(required=True, default='application/octet-stream')
 
+    # only populated if mime_type is image/*
+    # used in images.aspectRatio in app.bsky.embed.images
+    width = ndb.IntegerProperty()
+    height = ndb.IntegerProperty()
+
     created = ndb.DateTimeProperty(auto_now_add=True)
     updated = ndb.DateTimeProperty(auto_now=True)
 
@@ -368,6 +375,14 @@ def validate_size(size):
         blob = cls(id=url, cid=cid, size=len(resp.content))
         if mime_type:
             blob.mime_type = mime_type
+
+        if mime_type and mime_type.startswith('image/'):
+            try:
+                with Image.open(BytesIO(resp.content)) as image:
+                    blob.width, blob.height = image.size
+            except UnidentifiedImageError as e:
+                logger.info(e)
+
         blob.put()
 
         # re-validate size in case the server didn't give us Content-Length.

arroba/tests/__init__.py

@@ -4,6 +4,7 @@
 logging.basicConfig()
 if '-v' in sys.argv:
   logging.getLogger().setLevel(logging.DEBUG)
+  logging.getLogger('PIL').setLevel(logging.INFO)
 elif 'discover' in sys.argv or '-q' in sys.argv or '--quiet' in sys.argv:
   logging.disable(logging.CRITICAL + 1)
 

arroba/tests/test_datastore_storage.py

@@ -1,5 +1,6 @@
 """Unit tests for datastore_storage.py."""
 import os
+from pathlib import Path
 from unittest.mock import MagicMock, patch
 
 from google.cloud import ndb
@@ -373,6 +374,22 @@ def test_create_remote_blob_infer_mime_type_from_url(self):
         self.assertEqual(blob, got)
         mock_get.assert_not_called()
 
+    def test_create_remote_blob_image_aspect_ratio(self):
+        image_bytes = Path(__file__).with_name('keyboard.png').read_bytes()
+        cid = CID.decode('bafkreicjoxic5d37v2ae3wfxkjgxtvx5hsp4ejjemaog322z5dvm7kgtvq')
+        mock_get = MagicMock(return_value=requests_response(image_bytes))
+
+        blob = AtpRemoteBlob.get_or_create(url='http://my/blob.png', get_fn=mock_get)
+        mock_get.assert_called_with('http://my/blob.png', stream=True)
+        self.assertEqual({
+            '$type': 'blob',
+            'ref': cid,
+            'mimeType': 'image/png',
+            'size': 9003,
+        }, blob.as_object())
+        self.assertEqual(21, blob.width)
+        self.assertEqual(12, blob.height)
+
     def test_create_remote_blob_default_mime_type(self):
         mock_get = MagicMock(return_value=requests_response('blob contents'))
 

arroba/tests/testutil.py

@@ -54,13 +54,18 @@ def requests_response(body, status=200, headers=None):
     if isinstance(body, (dict, list)):
         resp.headers['content-type'] = 'application/json'
         resp._text = json.dumps(body, indent=2)
-    else:
+        resp._content = resp._text.encode()
+    elif isinstance(body, str):
         resp._text = body
+        resp._content = resp._text.encode()
+    elif isinstance(body, bytes):
+        resp._content = body
+    else:
+        assert False, f'unknown type for body: {type(body)}'
 
     if headers:
         resp.headers.update(headers)
 
-    resp._content = resp._text.encode()
     resp.encoding = 'utf-8'
     resp.status_code = status
     return resp

pyproject.toml

@@ -26,6 +26,7 @@ dependencies = [
     'dnspython>=2.0.0',
     'lexrpc>=0.8',
     'multiformats>=0.3.1',
+    'pillow',
     'pyjwt>=2.0.0',
     'simple-websocket',
 ]

requirements.txt

@@ -46,6 +46,7 @@ MarkupSafe==2.1.3
 multiformats==0.3.1.post4
 multiformats-config==0.3.1
 packaging==23.1
+Pillow==11.0.0
 proto-plus==1.23.0
 protobuf==4.24.3
 pyasn1==0.5.1