scheme-sdk 0.3.5__py3-none-any.whl → 0.3.7__py3-none-any.whl

scheme_sdk/connectors/base/message.py

@@ -2,6 +2,7 @@ from abc import ABC, abstractmethod
 from datetime import datetime
 from dataclasses import dataclass, field
 from typing import Any, Dict, Iterable, Optional
+from html_sanitizer import Sanitizer
 
 from .base import BaseConnector
 
@@ -11,6 +12,11 @@ class MessageConnector(BaseConnector, ABC):
     Abstract base class for all message connectors.
     """
 
+    _sanitizer: Sanitizer
+
+    def __init__(self):
+        self._sanitizer = Sanitizer()
+
     @abstractmethod
     def fetch_conversations(self) -> Iterable[Dict[str, Any]]:
         """
@@ -85,6 +91,7 @@ class MessageConnector(BaseConnector, ABC):
         """
         ...
 
+    @abstractmethod
     def normalize_message(self, raw_message: Dict[str, Any]) -> Dict[str, Any]:
         """
         Transform a platform-specific message into canonical format.
@@ -100,6 +107,7 @@ class MessageConnector(BaseConnector, ABC):
         """
         ...
 
+    @abstractmethod
     def normalize_conversation(
         self, raw_conversation: Dict[str, Any]
     ) -> Dict[str, Any]:
@@ -134,6 +142,18 @@ class MessageConnector(BaseConnector, ABC):
         self._logger.warning(f"search_messages not implemented for {self.platform}")
         return iter([])
 
+    def _sanitize_html(self, html: str) -> str:
+        """
+        Sanitize HTML content.
+
+        Args:
+            html: HTML content to sanitize
+
+        Returns:
+            Sanitized HTML content
+        """
+        return self._sanitizer.sanitize(html)
+
 
 @dataclass
 class Conversation:

scheme_sdk/connectors/outlook.py

@@ -4,6 +4,12 @@ from typing import Any, Dict, List, Optional
 
 import requests
 
+from pprint import pprint as pp
+from dotenv import load_dotenv
+import os
+
+load_dotenv()
+
 from .base import MessageConnector
 
 
@@ -14,6 +20,7 @@ class OutlookConnector(MessageConnector):
     _backoff_cap_seconds = 30
 
     def __init__(self, token: str):
+        super().__init__()
         self.token = token
         self.base = "https://graph.microsoft.com/v1.0"
 
@@ -23,17 +30,27 @@ class OutlookConnector(MessageConnector):
         query: Optional[str] = None,
         since_iso: Optional[str] = None,
     ) -> List[Dict]:
+        """Fetch conversations from Outlook. Filter criteria are ignored if query is provided.
+
+        Args:
+            top (int, optional): The number of conversations to fetch. Defaults to 50.
+            query (Optional[str], optional): The query to search for. Defaults to None.
+            since_iso (Optional[str], optional): The date to search for. Defaults to None.
+
+        Returns:
+            List[Dict]: The conversations.
+        """
         url = f"{self.base}/me/messages"
         page_size = min(max(top, 50), 200)
         params = {
             "$select": "id,conversationId,subject,from,receivedDateTime,hasAttachments,webLink",
-            "$orderby": "receivedDateTime desc",
             "$top": str(page_size),
         }
-        if since_iso:
-            params["$filter"] = f"receivedDateTime ge {since_iso}"
         if query:
-            params["$search"] = query
+            params["$search"] = f'"{query}"'
+        elif since_iso:
+            params["$filter"] = f"receivedDateTime ge {since_iso}"
+            params["$orderby"] = "receivedDateTime desc"
 
         # Dedupe by conversationId, fetching more pages until we have enough.
         threads_by_conv: Dict[str, Dict] = {}
@@ -67,6 +84,15 @@ class OutlookConnector(MessageConnector):
         conversation_id: str,
         since_iso: Optional[str] = None,
     ) -> List[Dict]:
+        """Fetch messages from a specific conversation.
+
+        Args:
+            conversation_id (str): The ID of the conversation.
+            since_iso (Optional[str], optional): The date to search for. Defaults to None.
+
+        Returns:
+            List[Dict]: The messages.
+        """
         url = f"{self.base}/me/messages"
         filter_parts = [f"conversationId eq '{conversation_id}'"]
         if since_iso:
@@ -88,10 +114,18 @@ class OutlookConnector(MessageConnector):
         ]
 
     def normalize_message(self, message: Dict) -> Dict[str, Any]:
+        """Normalize a message from Outlook.
+
+        Args:
+            message (Dict): The message to normalize.
+
+        Returns:
+            Dict[str, Any]: The normalized message.
+        """
         return {
             "title": message["subject"],
             "platform": self.platform,
-            "text": message["body"]["content"],
+            "text": self._sanitize_html(message["body"]["content"]),
             "direct_link": message["webLink"],
             "metadata": {
                 "platform_conversation_id": message["conversationId"],
@@ -106,6 +140,14 @@ class OutlookConnector(MessageConnector):
         }
 
     def normalize_conversation(self, conversation: Dict) -> Dict[str, Any]:
+        """Normalize a conversation from Outlook.
+
+        Args:
+            conversation (Dict): The conversation to normalize.
+
+        Returns:
+            Dict[str, Any]: The normalized conversation.
+        """
         return {
             "id": conversation["conversationId"],
             "title": conversation["subject"],
@@ -177,3 +219,11 @@ class OutlookConnector(MessageConnector):
             next_params = None  # nextLink already includes the query string
 
         return items
+
+
+if __name__ == "__main__":
+    connector = OutlookConnector(os.getenv("OUTLOOK_TOKEN"))
+    conversations = connector.fetch_conversations(top=1)
+    pp(conversations)
+    messages = connector.fetch_messages(conversations[0]["id"])
+    pp(messages)

{scheme_sdk-0.3.5.dist-info → scheme_sdk-0.3.7.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: scheme_sdk
-Version: 0.3.5
+Version: 0.3.7
 Summary: The Scheme SDK provides connectors for ingesting conversations, messages, and files across communication platforms.
 License:                                  Apache License
                                     Version 2.0, January 2004
@@ -203,6 +203,7 @@ License:                                  Apache License
             WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
             See the License for the specific language governing permissions and
             limitations under the License.
+Requires-Dist: html-sanitizer>=2.6.0
 Requires-Dist: requests>=2.32.5
 Requires-Python: >=3.11
 Project-URL: Homepage, https://www.schemebig.com/

{scheme_sdk-0.3.5.dist-info → scheme_sdk-0.3.7.dist-info}/RECORD

@@ -3,11 +3,11 @@ scheme_sdk/connectors/__init__.py,sha256=d7XyCKbx0QklcAFyhTbNxTg0PakQfTs3eQSayYU
 scheme_sdk/connectors/base/__init__.py,sha256=7bTnekIGEB0S8jDT7tBfCIgsW4VXtK_Da_vrrYhByEw,639
 scheme_sdk/connectors/base/base.py,sha256=JtsD-HBdjC_Jl22M9j-ihe0tyXJw6JpgWsZp_ozCwDA,8855
 scheme_sdk/connectors/base/errors.py,sha256=53Dhz5qImXJe0q0cO7IqG4yja0OC0hnxMNMMmgdCtW0,1152
-scheme_sdk/connectors/base/message.py,sha256=pW-snYsHlAQ5-qMFzAQr3CDaNG1Z-HRYB1fdmZww1TU,5423
+scheme_sdk/connectors/base/message.py,sha256=WqKGMsv8YTbCafkA53ybrlgF3FseVKx1M2jjN9O_7kE,5851
 scheme_sdk/connectors/discord.py,sha256=CdtzVlwT0aHcUayhLGno_vmfhdt_WRtGOeV2zj7Ye7I,844
 scheme_sdk/connectors/gmail.py,sha256=wZCj917qujXSE9jexqvkDgdmz02y6QST8q9KNRwEvg0,866
-scheme_sdk/connectors/outlook.py,sha256=5Eli_8P9ZeDKQSq6DlbOlFAmBX5m5QFWQJVPHVoL4bE,6240
+scheme_sdk/connectors/outlook.py,sha256=udANT0wZNOla_u0Tp8ih-HVLeJeGp7P7GHux6l2SMgs,7798
 scheme_sdk/connectors/slack.py,sha256=_g--XxS4_ImT0fs0HX9vNquTkUlaikiylBhsgJzCC-4,1434
-scheme_sdk-0.3.5.dist-info/WHEEL,sha256=e_m4S054HL0hyR3CpOk-b7Q7fDX6BuFkgL5OjAExXas,80
-scheme_sdk-0.3.5.dist-info/METADATA,sha256=UNwkt_Jx_Cgpk5Bd47cOc-r1zVHMkwAWgVNTxt9i8tU,15084
-scheme_sdk-0.3.5.dist-info/RECORD,,
+scheme_sdk-0.3.7.dist-info/WHEEL,sha256=e_m4S054HL0hyR3CpOk-b7Q7fDX6BuFkgL5OjAExXas,80
+scheme_sdk-0.3.7.dist-info/METADATA,sha256=Iu9KqEFvFFKg37iuf2i0h7gyljcTjMPElFf8ZnSgQNY,15121
+scheme_sdk-0.3.7.dist-info/RECORD,,