Fix canonicalize (#590)

core/ld.py

@@ -547,6 +547,18 @@ schemas = {
             }
         },
     },
+    "schema.org": {
+        "contentType": "application/ld+json",
+        "documentUrl": "https://schema.org/docs/jsonldcontext.json",
+        "contextUrl": None,
+        "document": {
+            "@context": {
+                "schema": "http://schema.org/",
+                "PropertyValue": {"@id": "schema:PropertyValue"},
+                "value": {"@id": "schema:value"},
+            },
+        },
+    },
 }
 
 DATETIME_FORMAT = "%Y-%m-%dT%H:%M:%S.Z"
@@ -592,24 +604,32 @@ def canonicalise(json_data: dict, include_security: bool = False) -> dict:
     """
     if not isinstance(json_data, dict):
         raise ValueError("Pass decoded JSON data into LDDocument")
-    context = [
+
-        "https://www.w3.org/ns/activitystreams",
+    context = json_data.get("@context", [])
-        {
+
-            "blurhash": "toot:blurhash",
+    if not isinstance(context, list):
-            "Emoji": "toot:Emoji",
+        context = [context]
-            "focalPoint": {"@container": "@list", "@id": "toot:focalPoint"},
+
-            "Hashtag": "as:Hashtag",
+    if not context:
-            "manuallyApprovesFollowers": "as:manuallyApprovesFollowers",
+        context.append("https://www.w3.org/ns/activitystreams")
-            "sensitive": "as:sensitive",
+        context.append(
-            "toot": "http://joinmastodon.org/ns#",
+            {
-            "votersCount": "toot:votersCount",
+                "blurhash": "toot:blurhash",
-            "featured": {"@id": "toot:featured", "@type": "@id"},
+                "Emoji": "toot:Emoji",
-        },
+                "focalPoint": {"@container": "@list", "@id": "toot:focalPoint"},
-    ]
+                "Hashtag": "as:Hashtag",
+                "manuallyApprovesFollowers": "as:manuallyApprovesFollowers",
+                "sensitive": "as:sensitive",
+                "toot": "http://joinmastodon.org/ns#",
+                "votersCount": "toot:votersCount",
+                "featured": {"@id": "toot:featured", "@type": "@id"},
+            }
+        )
+
     if include_security:
         context.append("https://w3id.org/security/v1")
-    if "@context" not in json_data:
+
-        json_data["@context"] = context
+    json_data["@context"] = context
 
     return jsonld.compact(jsonld.expand(json_data), context)
 

tests/core/test_ld.py

@@ -2,7 +2,7 @@ import datetime
 
 from dateutil.tz import tzutc
 
-from core.ld import parse_ld_date
+from core.ld import canonicalise, parse_ld_date
 
 
 def test_parse_ld_date():
@@ -41,3 +41,68 @@ def test_parse_ld_date():
         tzinfo=tzutc(),
     )
     assert difference.total_seconds() == 0
+
+
+def test_canonicalise_single_attachment():
+    data = {
+        "@context": [
+            "https://www.w3.org/ns/activitystreams",
+            {
+                "schema": "http://schema.org#",
+                "PropertyValue": "schema:PropertyValue",
+                "value": "schema:value",
+            },
+        ],
+        "attachment": [
+            {
+                "type": "http://schema.org#PropertyValue",
+                "name": "Location",
+                "http://schema.org#value": "Test Location",
+            },
+        ],
+    }
+
+    parsed = canonicalise(data)
+    attachment = parsed["attachment"]
+
+    assert attachment["type"] == "PropertyValue"
+    assert attachment["name"] == "Location"
+    assert attachment["value"] == "Test Location"
+
+
+def test_canonicalise_multiple_attachment():
+    data = {
+        "@context": [
+            "https://www.w3.org/ns/activitystreams",
+            {
+                "schema": "http://schema.org#",
+                "PropertyValue": "schema:PropertyValue",
+                "value": "schema:value",
+            },
+        ],
+        "attachment": [
+            {
+                "type": "http://schema.org#PropertyValue",
+                "name": "Attachment 1",
+                "http://schema.org#value": "Test 1",
+            },
+            {
+                "type": "http://schema.org#PropertyValue",
+                "name": "Attachment 2",
+                "http://schema.org#value": "Test 2",
+            },
+        ],
+    }
+
+    parsed = canonicalise(data)
+    attachment = parsed["attachment"]
+
+    assert len(attachment) == 2
+
+    assert attachment[0]["type"] == "PropertyValue"
+    assert attachment[0]["name"] == "Attachment 1"
+    assert attachment[0]["value"] == "Test 1"
+
+    assert attachment[1]["type"] == "PropertyValue"
+    assert attachment[1]["name"] == "Attachment 2"
+    assert attachment[1]["value"] == "Test 2"

users/models/identity.py

@@ -854,16 +854,14 @@ class Identity(StatorModel):
         self.metadata = []
         for attachment in get_list(document, "attachment"):
             if (
-                attachment["type"] == "http://schema.org#PropertyValue"
+                attachment["type"] == "PropertyValue"
                 and "name" in attachment
-                and "http://schema.org#value" in attachment
+                and "value" in attachment
             ):
                 self.metadata.append(
                     {
-                        "name": attachment.get("name"),
+                        "name": attachment["name"],
-                        "value": FediverseHtmlParser(
+                        "value": FediverseHtmlParser(attachment["value"]).html,
-                            attachment.get("http://schema.org#value")
-                        ).html,
                     }
                 )
         # Now go do webfinger with that info to see if we can get a canonical domain